更便宜地创作。

photorealistic live-action fantasy, entirely first-person rider POV, massive obsidian-crimson dragon, molten-gold eyes, volcanic-black horns, burgundy translucent wings, realistic anatomy and physics,

A luxurious fine-dining restaurant located deep beneath the ocean, surrounded by enormous curved glass windows revealing a vibrant coral reef and schools of fish swimming outside. Guests sit at elegant tables while warm golden pendant lights illuminate the interior. Soft rays of sunlight penetrate the deep blue water above, creating realistic caustic reflections across the walls, tables, and floor. Waiters move naturally between tables carrying plates of food. Subtle bubbles drift past the windows. The camera slowly moves through the restaurant in a smooth cinematic tracking shot, capturing the contrast between the warm elegant interior and the vast blue ocean outside. Photorealistic, believable architecture, realistic water physics, natural human movement, cinematic lighting, high-end film cinematography.

Create an ultra-realistic Indian village documentary sequence shot entirely on an old handheld MiniDV / Hi8-style camcorder, following a young village girl who plays football every day with the local kids. Everything should feel observational and slightly amateur, like a small documentary crew has been following her life for a few days and she has become comfortable with the cameraman. Slightly faded colors, baked-in camcorder sharpness, mild interlacing, tape softness, compression noise, autofocus hunting, exposure pumping, rough handheld movement, imperfect zooms, accidental reframing, operator footsteps, wind hitting the microphone, children shouting in the distance, village ambience, birds, dogs, and natural location audio. No music. No modern cinematic grading. No gimbal. No dramatic sports-commercial photography. The camera should feel physically carried by one person. The setting is a real Indian village football ground: not a professional pitch, just an open dusty field between houses and farmland, uneven patches of grass, stones marking boundaries, homemade goalposts, children in ordinary clothes, a few slippers and bags lying near the side, bicycles nearby, villagers occasionally watching from the edge. Shot 1 — She takes the documentary crew to the ground Start with the cameraman already following the young Indian girl footballer through a narrow village lane. She walks slightly ahead of the camera carrying or lightly kicking her old football, completely natural and comfortable with the documentary crew. She turns around while still walking and says in English with a natural Indian accent: “Come with me. I’ll show you where we play football every day.” The cameraman follows behind her with slightly shaky footsteps. She walks out of the lane and onto the open field. She spreads her hand toward the rough football ground and proudly says: “This is it.” Do not make the reveal grand or cinematic. The charm is that she is incredibly proud of a very ordinary dusty village field. Shot 2 — The joy of playing Cut to the cameraman already standing close to the edge of the game while the children are playing. The same girl runs past the camera chasing the ball with other village kids. The camera operator struggles to follow them, performs a clumsy little pan, briefly loses focus, then catches them again. The cameraman laughs from behind the camera and says in English with a natural Indian accent: “See how much fun this is?” The girl glances toward him while still moving, grins, and immediately runs back into play. Keep this shot energetic and documentary-real: children shouting things like “Pass! Pass!”, feet kicking dust, someone nearly running into the camera, the operator stepping aside, an imperfect zoom toward the ball, then quickly widening again because the action changes direction. Shot 3 — Bicycle-kick goal and celebration Cut to what feels like the final moments of the same village match. The camera is handheld from the sideline, slightly too far away, trying to follow the action. Several kids rush toward the makeshift goal. One player near the side sends a high crossing ball into the middle. The cameraman reacts late and pans quickly. The same girl runs into position, watches the descending ball, turns her body— and performs a completely believable bicycle kick. Her back leaves the ground, one leg swings upward, she connects cleanly with the ball in midair, and the ball flies toward the makeshift goal. The goalkeeper reacts— goal. The children immediately explode in celebration. No slow motion. No heroic cinematic cutaway. Let the old camcorder almost miss the kick because the operator is trying desperately to keep up. The girl jumps back to her feet and starts running directly toward the documentary camera, yelling in English: “YEEEAAAH!” Her teammates chase behind her, screaming and celebrating. She reaches the camera laughing, nearly filling the lens, jumping around with the others as the operator backs away and struggles to keep everyone framed. Kids hug her, slap her hands, shout, jump, and crowd around the camera in exactly the messy way children celebrate after scoring. The lens briefly loses focus because everyone is too close. End inside that chaos — her laughing directly toward the cameraman while the other children are still celebrating around her. Overall feeling: a tiny observational documentary about an ordinary village girl suddenly becoming a football hero for one moment. It should feel warm, real, imperfect, spontaneous and completely unstaged — the camera just happened to be there when she scored the best goal of the match.

SUBJECT ANCHOR 1: <<<image_1>>> — Ash Knight SUBJECT ANCHOR 2: <<<image_2>>> — Obsidian Rival PROP ANCHOR: <<<image_3>>> — Magma Greatsword DURATION: 30 seconds ASPECT RATIO: 16:9 STYLE: Surreal cinematic dark fantasy / science-fantasy CAMERA: Large-format cinema, 24mm / 35mm / 50mm / 85mm FRAME RATE: 24fps with selective 60fps STORY In the middle of an endless desert stands a gigantic circular mirror buried vertically in the sand. Ash Knight arrives from one direction. Obsidian Rival arrives from another. The mirror does not reflect them. Instead, it shows a completely different version of the desert—one thousands of years in the future. --- 00:00–00:05 — THE DESERT Open with an enormous aerial shot of an endless black-and-gold desert at sunset. Wind creates long patterns across the dunes. At the center of the landscape stands a gigantic circular metallic structure half-buried in the sand. The camera descends toward it. Ash Knight appears as a distant silhouette walking across the dunes. --- 00:05–00:09 — THE SECOND ARRIVAL Cut to a 50mm side profile. Ash Knight reaches the structure. Across the enormous circular mirror, Obsidian Rival emerges from the opposite side. Both stop. The mirror stands between them. The camera slowly moves sideways, keeping the entire structure centered. --- 00:09–00:13 — THE MIRROR 85mm close-up. The surface of the mirror is perfectly black. Ash Knight approaches it. His reflection should appear— but it doesn't. Instead, the mirror shows the same location thousands of years later. The desert has disappeared. A massive futuristic city now covers the landscape. --- 00:13–00:17 — THE FUTURE 35mm composition. Obsidian Rival steps closer. The mirror changes again. It shows a completely different future. The city is now empty. Towering structures are covered in sand. A gigantic red moon hangs above the horizon. The two warriors look toward the impossible reflection. --- 00:17–00:21 — THE SWORD 85mm macro. Ash Knight draws the Magma Greatsword. The red lava veins pulse. The mirror responds. A thin red reflection appears across its surface. The reflection of the sword suddenly becomes visible even though the warriors still have no reflections. --- 00:21–00:25 — THE FRACTURE Ash Knight slowly raises the Greatsword toward the mirror. He does not strike. The mirror surface begins developing thin luminous cracks by itself. The cracks spread outward like a massive geometric pattern. The desert wind suddenly stops. Every grain of sand becomes completely still. --- 00:25–00:28 — THE OTHER WORLD The mirror opens like a doorway. Beyond it is not another desert. It is an enormous ocean suspended vertically in the sky. Massive floating structures drift beneath the water. The camera slowly pushes forward. Both warriors remain behind the threshold. --- 00:28–00:30 — FINAL IMAGE Extreme wide shot. The circular mirror now stands open in the middle of the desert. Inside it, the impossible ocean-world stretches endlessly. Ash Knight and Obsidian Rival stand on opposite sides of the opening. The Magma Greatsword glows faintly. A single wave moves across the vertical ocean. The mirror suddenly closes. The desert returns to normal. CUT TO BLACK. No text. No title. No logo. CHARACTER CONTINUITY Preserve the exact reference appearance throughout. Ash Knight: exact armor, magma rune patterns, cape, proportions, silhouette and identity. Obsidian Rival: exact obsidian dragon-scale armor, horned helmet, orange visor, proportions, silhouette and identity. Magma Greatsword: exact blade geometry, jagged edges, handle, forged texture and red lava veins. No redesigns, identity drift, morphing, duplicates, weapon transformation, armor changes or flickering. ENVIRONMENT CONTINUITY One desert. One circular mirror. Same sunset. Same sand. Same weather. Same positions of the characters. The mirror's internal worlds may change, but the physical desert remains consistent. CINEMATOGRAPHY 24mm — vast desert. 35mm — character/environment compositions. 50mm — dramatic character shots. 85mm — mirror and sword details. Slow aerial descent. Lateral tracking. Controlled push-ins. Smooth crane movement. Subtle 60fps during the mirror fracture. Natural motion blur. Realistic sand interaction. Large-format cinematic depth. VISUAL DESIGN Golden desert. Deep blue-black shadows. Warm sunset. Black reflective mirror. Red magma glow. Cold futuristic architecture inside the mirror. Surreal vertical ocean. The color palette should evolve naturally rather than remaining monochromatic. NEGATIVE PROMPT text, subtitles, watermark, logo, modern clothing, cartoon, anime, low-poly, plastic armor, character redesign, identity drift, face morphing, weapon redesign, duplicate characters, extra limbs, malformed hands, inconsistent proportions, flickering, random camera movement, environment reset, inconsistent lighting, excessive destruction, graphic gore, blood, dismemberment, blurry subjects, flat lighting. DIRECTOR'S NOTE: Do not explain the mirror. The audience should discover its rules visually. The first reveal establishes that it shows the future; the second reveal breaks that assumption; the final reveal shows an entirely impossible world. The ending should leave the viewer with a question rather than an answer.

Ultra-realistic AAA live-action dark fantasy film. 30 seconds of completely continuous single-camera, uncut long-take. Maintains strict spatial coordinates and time axis, expressing depth of field and heavy atmosphere comparable to ARRI ALEXA 65. Environment: (Ref 3) huge royal palace. Black/white/gold marble floor, volumetric light from stained glass, huge chandelier. Soldiers in silver, gold, and dark blue military uniforms are crowded on the left and right. Queen: (Ref 1) beautiful queen. Long platinum-blonde hair, golden crown with blue jewels, royal dress of white to ivory with gold embroidery and dark blue. Queen (Monster Form): (Ref 2) huge monstrosity. Retains the crown and platinum hair, has multi-layered fangs, long claws, and bone-membrane organs on the back. Sinner: (Ref 4) male soldier. Brown hair, dirt and wounds, silver-gold armor, dark blue costume, handcuffs and chains. 0.0-4.5s: Camera starts from behind the queen. Queen's back in foreground, sinner about 12m ahead in central aisle, rows of soldiers on left and right. Guard pushes sinner forward, sinner is slammed onto marble on both knees. Camera passes by the queen's right shoulder, continuously advancing toward the sinner to capture the upper chest. <Low impact sound of heavy armor colliding with floor, sharp metallic sound of chains> Sinner looks up and says in an urgent voice. {Your Majesty... please forgive me!} 4.5-9.5s: Camera doesn't stop, passes left side of sinner, draws a smooth arc to capture queen on the throne side from the front. Queen steps forward. Queen says in a quiet, cold voice. {Did you think you could steal the royal treasure and escape?} <Dry metallic sound of soldiers reaching for sword hilts simultaneously> Sinner shouts. {No! I—} Queen takes another step, emotion disappears from face. 9.5-14.4s: Camera approaches queen by about 1m. Queen's skeleton and muscles continuously deform under skin, corners of mouth tear, multi-layered fangs are exposed. Platinum hair flutters, bone-membrane organs expand from back. Body enlarges, transforming into monster form while dress tears. <Skeletal deformation sounds, meat stretching sounds> Monster queen roars. <Unusual roar that makes stained glass resonate> Soldiers on left and right drop to one knee in a wave. Sinner says while backing away. {No way...!} 14.4-19.6s: Sinner turns 180 degrees and begins sprinting at full speed toward camera (palace interior direction). From here, camera continues to retreat along central aisle toward palace interior. Fixed linear relationship: "Camera ← Sinner ← Monster Queen ← Throne". Monster queen begins chase on all fours. Sinner shouts behind while running. {Stop! Don't come!} <Rough breathing, armor sound, heavy impact sound of monster's huge limbs hitting marble> Sinner's chain gets caught on marble decoration, sinner collapses forward. Camera also slows retreat. 19.6-23.4s: Camera almost stops on current central aisle. Monster queen catches up from behind, blocks escape route with right forelimb. Monster's neck extends, huge jaws grab sinner's torso from diagonal rear and lift him up. <Metallic crushing sound, short scream, sound of chain breaking and falling to floor> Monster queen moves jaws and completely swallows sinner. Sinner's body completely disappears from screen. 23.4-30.0s: Monster queen does not move from that predation position (away from throne), begins de-transformation into human form. Bone organs are folded, body shrinks, torn dress is restored with golden magical threads. Camera retreats very slowly by about 1m. <Skeletal contraction sounds, delicate sounds of costume restoration> Queen, fully returned to human form, fixes slightly tilted crown near chains on floor. Surrounding soldiers stand up. A new sinner is pulled out from back of screen. Queen stares at new sinner and says coldly. {Next.} Camera retreats/rises further down central aisle, ending with new sinner in foreground, queen in center, and throne in background. [Continuity & Spatial Constraints] Maintain continuous long-take with a single physical camera for 30 seconds. After 14.4 seconds, camera only retreats toward palace interior. Monster queen always stays behind sinner. After predation, sinner's body remains completely disappeared. Queen remains at the spot where she completed predation and does not return to throne. During de-transformation, camera only retreats maintaining axis on current central aisle, actions like circling queen or moving to front are strictly prohibited.

Create a whimsical, Ghibli-inspired hand-drawn animation showing tiny anime workers harvesting and transporting oversized soybean sprouts inside a large white porcelain bowl on a wooden table. Some workers shovel sprouts, climb a wooden ladder, carry a giant sprout on a stretcher, and load a small wooden cart, while others direct the team and rest from exhaustion. Use warm golden lighting, soft colors, floating dust motes, cozy details, expressive characters, and gentle cinematic camera movements, ending with a wide shot of the whole miniature team working together.

[Style] Photorealistic Wing Chun Film, high-energy Douyin action video, Ramp-mo rhythm transitions, High-Speed Photography, 4K cinematic quality, realistic skin, sweat, and wood textures, no anime feel, no magic effects. [Duration] 20 seconds. [Aspect Ratio] 9:16 vertical screen. [Scene] Traditional old Lingnan martial arts hall, dark wood floors, mottled gray walls, wooden lattice windows; morning sunlight streaming in from the side, tiny dust motes floating in the air. A traditional solid wood wooden dummy is fixed in the center of the frame, with a cylindrical body, three wooden arms, and one slanted wooden leg. [Character] Wing Chun practitioner as per Image 1. Face, features, hairstyle, body proportions, and original clothing must strictly follow the image, maintaining the same person and look throughout. [Audio] No dialogue, no subtitles; emphasize the sound of dummy collisions, footsteps rubbing, short breaths, and the sound of wood splintering at the end. [00:00-00:04] Shot 1: Opening high-speed dummy striking (Impact Hook). The first frame enters high-speed action directly without environmental setup. Close-up side shot: The protagonist is rapidly practicing on the wooden dummy, hands continuously performing Tan Sau, Bong Sau, Pak Sau, palm strikes, and short punches around the centerline; hands switch rapidly between the three wooden arms, causing the arms to vibrate continuously. The camera tracks horizontally close to the protagonist's hands (Close Tracking), with hands flashing past the lens with natural motion blur; each strike produces different wood sounds. Cut to a low-angle full-body shot: The protagonist uses tight small steps to bypass the slanted leg of the dummy, upper body remains stable, feet do not cross, no large jumps. The final palm strike lands heavily in the center, the whole dummy shakes violently, dust falls, but no cracks appear. [00:04-00:08] Shot 2: Sudden deceleration · Precise slow practice (Ramp Down). After the final strike, all sound suddenly goes quiet, transitioning to slow motion. Side-front medium shot: The protagonist slows down, left hand against the upper arm, slowly dissolving outward; right hand delivers a short straight punch along the center of the dummy, retracting quickly after contact. The protagonist continues slowly performing Bong Sau, Pak Sau, and palm strikes, each movement clearly separated, arms short and close to the body, elbows not excessively flared. The camera orbits the protagonist and dummy about 90 degrees (Slow Orbit), showing the palm brushing the rough wood grain, slight vibration of the wooden arms, and sweat dripping down the cheek. The protagonist's gaze is locked on the center of the dummy, shoulders relaxed, lips slightly closed, breathing steady; observing distance and rhythm rather than angry flailing. [00:08-00:12] Shot 3: Slow to fast · Progressive acceleration. Low drum beats enter, the practice speed increases layer by layer. First set: Moderate speed, left hand dissolves arm, right palm strikes center. Second set: Visibly faster, Bong Sau to Pak Sau, followed by two short punches. Third set: Highest speed, left and right hands alternate rapidly, punches and palms crossing along the centerline, combined with small turns and ground-hugging footwork to the side of the dummy. Low-angle Steadicam Orbit from behind to the side-front; foreground wooden arms flash past, background windows and sunlight form slight speed trails. Movement gets faster, but the body doesn't sway, face remains steady, only breathing becomes heavier and gaze more focused. The dummy vibrates constantly with dense wood sounds, remaining intact. [00:12-00:14] Shot 4: Sudden stop · Locking on the dummy (Sudden Silence). After the high-speed combo, hands suddenly stop. All drums, strikes, and camera movement stop simultaneously. Fixed front medium shot: The protagonist stands before the dummy, feet unmoving, left hand retracted to chest, right hand hanging naturally. The dummy is still vibrating slightly, a few wood chips fall. Close-up of face: The protagonist slowly lifts eyes, locking gaze on the center of the dummy; lips tighten slightly, taking a deep breath. The scene remains quiet, pausing for the final one-inch punch. [00:14-00:17] Shot 5: Open palm approach · Clenching fist to store energy (Open Palm to Fist). Side medium-close shot: The protagonist slowly raises the right hand, fingers fully spread, palm facing the center of the dummy. Must clearly show a fully open palm first, not a fist. Body stable, right hand approaches the dummy very slowly; as the palm gets closer, the camera follows. Extreme Close-up: The open palm stops about one inch from the dummy, not touching, fingers naturally extended, palm, knuckles, and rough wood grain clearly visible. After a short pause, the fist clenches: pinky curls first, then ring and middle, index follows, thumb finally presses down. Fingers tighten completely from open palm to vertical fist. During clenching, the arm must not pull back; the distance remains one inch. Muscles and veins in the forearm tighten slightly, shoulders down, body doesn't lean forward. Sound is only skin friction and slow inhalation. [00:17-00:20] Shot 6: One-Inch Punch impact · Dummy shatters (One-Inch Punch). Extreme side close-up: The clenched right fist is one inch from the center. Still for 0.3s. Protagonist exhales sharply, the fist impacts from the close distance instantly; no pulling back, no big swing, moves only a few centimeters, clean hit. The impact must be a short, fast, concentrated explosion: Open palm approaches -> fingers clench -> store energy in place -> sudden impact from one inch. Do not combine into a normal punch. A heavy bang at contact. The dummy vibrates violently, radial cracks appear at the hit point; the thick body shatters into pieces, arms and leg fall off. Super Slow-mo for the shatter: Morning light illuminates flipping chips and heavy wood blocks falling with crack and thud sounds. No fire, smoke, or energy waves. Protagonist's fist remains at the center, body stable, not falling forward or bouncing back. Final 0.6s returns to normal speed: protagonist retracts fist, returns to Man Sau position, calmly watching the debris. Slow Pull-back to a frame of protagonist, shattered dummy, debris, and morning dust. [One-Inch Punch Hard Requirements] Must be split into 5 stages: 1. Fully open palm. 2. Approach slowly. 3. Stop at one inch. 4. Fingers clench sequentially and store energy. 5. Sudden impact. Must see the process of open palm and clenching. No starting with a fist, no pulling back, no skipping the clench, no long-distance punch, no hitting early, must pause after clenching before impact. [Director Constraints] Only one protagonist, strictly follow Image 1. Dummy always has 3 arms, 1 leg. Dummy stays intact until 17s. No boxing swings, hooks, big wind-ups, flying kicks, or acrobatics. No clipping through wood. No extra fingers, missing fingers, or deformations. No magic effects, glowing fists, shockwaves, or manga lines. No glass/foam/rubber texture for wood. No shouting or celebrating. No other characters, dialogue, titles, logos, or watermarks.

Create a 15-second hyper-realistic cinematic video of a stylish young woman skateboarding through a chaotic, crowded modern city during daytime. 0–3 sec: Start with a low-angle close-up of her skateboard wheels rapidly rolling over wet asphalt. Cars, motorcycles, buses, and pedestrians rush past in the background. Natural handheld camera movement. 3–7 sec: Pull back to reveal the girl confidently weaving through a busy city street on her skateboard, narrowly passing between moving vehicles while maintaining perfect balance. Her hair and oversized jacket move naturally with the wind. 7–11 sec: Dynamic tracking shot from the side as she accelerates, jumps onto a low roadside ramp, performs a smooth 180° aerial trick, and lands cleanly. Vehicles continue moving around her, creating controlled chaos. 11–15 sec: Camera swings to the front and moves backward as she skates directly toward the lens through the crowded street. She gives a subtle confident smile, then sharply turns around a car and disappears into the busy city. Visual style: photorealistic, cinematic urban atmosphere, realistic physics, natural skin and hair movement, detailed traffic, motion blur, dramatic depth of field, dynamic camera work, realistic lighting, high-end commercial cinematography, 4K, smooth continuous motion. Important: Keep the girl’s face, hairstyle, clothing, skateboard, and physical appearance consistent throughout the entire 15 seconds. No sudden morphing, duplicate people, distorted limbs, floating objects, or unrealistic vehicle movement.

VISUALS: Global Camera & Style Directives: A high-budget, ultra-realistic cinematic YouTube vlog aesthetic. The footage mimics a modern full-frame mirrorless camera (like a Sony A7S III) paired with a 16mm wide-angle vlogging lens. The visual fidelity features organic handheld camera shake, rolling shutter during fast pans, natural lens flares, and occasional water droplets hitting the front glass. The color grading is moody and highly cinematic, emphasizing the deep, bruising blues and purples of a severe storm contrast against the warm interior lighting of a vehicle. The protagonist is a rugged, energetic male vlogger in a rain jacket, drenched in sweat and rain, vibrating with adrenaline. Scene 1: Close-up, 16mm handheld selfie perspective. The vlogger is sitting in the driver's seat of a rugged 4x4 vehicle. The camera is held at arm's length, capturing his intensely excited face. The vehicle is shaking. Through the rain-streaked windshield behind him, an impossibly massive, pitch-black supercell storm cloud is rapidly rotating in the sky. He yells directly into the lens, pointing a finger out the window. Scene 2: Handheld whip pan. The vlogger aggressively shoves the car door open and steps out into the howling elements, the camera swinging wildly as he moves. The camera rapidly pans from his face out toward the vast, flat Midwestern plains. The autofocus hunts for a split second before locking onto a colossal, mile-wide tornado touching down on the horizon. The scale is epic, terrifying, and beautiful. Scene 3: Wide POV tracking shot. The camera is held out to capture a split-screen view: the vlogger on the left third of the frame, hair whipping violently in the wind, completely awestruck; the massive tornado dominates the right two-thirds of the frame. A massive fork of lightning strikes behind the twister, brilliantly backlighting the funnel in a flash of harsh white light. The vlogger throws his free hand to his head in absolute disbelief. AUDIO & DIALOGUE: Global Audio Directives: Raw, immersive, in-the-moment documentary audio. The sound profile heavily features the harsh, clipping noise of wind hitting a camera microphone (complete with a deadcat wind muff struggling to block the sound). The background is a continuous, deafening, low-frequency rumble that rattles the audio mix. Scene 1 SFX: The heavy, rhythmic thwack-thwack of windshield wipers on maximum speed. The aggressive drumming of heavy hail and rain on a metal car roof. Scene 1 Dialogue: Vlogger (yelling at the top of his lungs over the storm, voice cracking with adrenaline): "Guys! We are directly in the drop zone! It's dropping right now! Look at this!" Scene 2 SFX: The heavy, metallic creak and slam of a car door fighting against extreme wind. An immediate, overwhelming rush of wind noise completely blows out the microphone for a fraction of a secon

@ Your image asset as the main character reference, strictly preserving the girl's facial features, makeup, light green Hanfu, lotus hair ornaments, and full set of accessories, while strictly maintaining the girl's body and waist-hip-chest proportions. @ Your video asset as the movement and camera work reference, the character completely replicates all dance body movements, rhythm, and camera movement from that video.

A close-up aesthetic ASMR cosmetics video showing the step-by-step creation and packaging of a shimmering pink lip gloss.The video opens with a thick, glossy, sheer pink base being poured from a glass beaker into a small clear bowl, followed by a second pour of clear, glittery lip gloss base. A drop of vibrant pink liquid pigment is added right into the center of the shimmering gel. A clear spatula then stirs and mixes the glossy mixture together, blending the shimmering particles and pink pigment into a smooth, thick, rose-gold shimmer gloss.The mixed lip gloss is then filled into a clear, cylindrical lip gloss tube. A applicator wand with a soft plush doe-foot applicator is dipped inside and pulled out, pulling a string of shiny gloss. Finally, the gloss is swatched onto a clear glass plate showing off its smooth, glass-like shine, before ending on a product shot of the finished lip gloss tube sitting on a vanity table next to scattered tiny pink crystals with the text overlay "Would you wear this shade?The lighting is soft, clean, and bright, featuring smooth camera movements, satisfying ASMR sound effects, high clarity, and a glossy luxury aesthetic.

15-Second Gym Vlog — Continuous Dialogue Character: The same young adult woman throughout the entire video, wearing a black T-shirt and sporty gym pants. Keep her face, hairstyle, outfit, and appearance consistent from beginning to end. Scene: A realistic modern gym vlog. From the very first frame, the woman is already talking directly to the camera while walking through the gym. She keeps speaking continuously throughout the entire 15-second video—no silent moments and no voice-over. 0–5 sec: She walks toward the camera in selfie-vlog style, smiling naturally and talking directly to the viewer. Gym equipment and people working out are visible in the background. Dialogue: "Hey guys! I'm at the gym today, and I'm going to show you a little bit of my workout." 5–10 sec: While continuing to talk, she reaches the dumbbell area, picks up a moderate-sized dumbbell with one hand, briefly lifts it in a natural demonstration, and keeps speaking to the camera. Dialogue: "I usually start with some simple exercises like this, just to get warmed up." 10–15 sec: Still holding the dumbbell briefly, she looks at the camera and continues talking with a friendly smile, then places it back naturally. Dialogue: "Alright, let's get started and make this workout a good one!" Important: Continuous talking from start to finish, natural lip-sync matching every word, no narration, no sudden cuts, no change of character or clothing. The same reference character must remain consistent throughout. Realistic gym environment, natural body movement, handheld smartphone vlog style, cinematic 4K quality.

Main Subject: Young Korean woman, 21, natural everyday appearance, short black hair cut into a slightly uneven chin-length bob with wispy bangs, realistic skin texture, minimal makeup, playful and energetic personality. Wearing a faded lavender zip-up hoodie over a white graphic T-shirt, loose black cargo pants, worn white sneakers, a small black shoulder bag, and a simple digital watch. Maintain consistent identity, clothing, hairstyle, and appearance throughout. Location: Small old-fashioned neighborhood arcade in Seoul during a quiet weekday afternoon. Rows of colorful arcade cabinets, CRT screens, coin-operated games, faded Korean game posters, fluorescent ceiling lights, plastic stools, dusty corners, and a few local teenagers playing games. The entrance opens directly onto a narrow residential street. Visual Style: Ultra-realistic documentary realism. Authentic early-2000s atmosphere, spontaneous behavior, imperfect everyday details, natural human reactions. Camera Style: Early-2000s Sony MiniDV camcorder. Heavy handheld shake, imperfect framing, autofocus hunting between bright CRT screens and her face, fluorescent flicker, exposure pumping, faded colors, soft contrast, visible digital noise, slight motion blur, DV compression artifacts, no stabilization. 00:00–00:03 She pushes open the arcade door and steps inside, immediately looking around with curiosity. The camera briefly becomes overexposed from the bright screens before recovering. 00:03–00:06 She inserts a coin into an old fighting-game cabinet and grips the joystick. The CRT screen flickers while autofocus struggles between the screen and her face. 00:06–00:09 She becomes completely focused on the game, rapidly pressing buttons. After losing, she throws her hands up and laughs at herself. 00:09–00:12 She tries another machine, this time a small basketball arcade game, tossing the ball repeatedly while the camera shakes as the operator laughs behind it. 00:12–00:15 She walks toward the exit holding her remaining coins, turns back toward the camera with a playful victory pose, then leaves. The camera follows her into the bright street and abruptly cuts to black. Audio: Natural ambience only—CRT game sounds, coin drops, arcade buttons clicking, joystick movement, teenagers talking, occasional laughter, fluorescent buzzing, footsteps, distant street traffic. No added music. No narration. Goal: A forgotten MiniDV home video from 2004, capturing an ordinary afternoon inside a neighborhood Korean arcade–nostalgic, energetic, imperfect, and convincingly authentic.

Main Subject: Young Korean woman, 24, natural everyday appearance, long dark-brown hair tied into a low messy ponytail with loose strands around her face, realistic skin texture, minimal makeup, relaxed and curious personality. Wearing a faded sage-green windbreaker over a white T-shirt, loose beige cargo pants, worn navy canvas sneakers, a dark baseball cap, and a small canvas backpack. Maintain consistent identity, clothing, hairstyle, and appearance throughout. Location: Small working fishing harbor in Busan just after sunrise. Weathered fishing boats, tangled ropes, blue plastic crates, fishing nets, concrete docks, old warehouses, seagulls, calm seawater, distant hills, and a few local fishermen preparing their boats. No tourists or commercial attractions. Visual Style: Ultra-realistic documentary realism. Genuine candid behavior, natural body language, imperfect everyday moments, authentic coastal atmosphere. Camera Style: Early-2000s Sony MiniDV camcorder. Heavy handheld shake, imperfect framing, autofocus hunting between fishing equipment and her face, exposure pumping between bright sky and shaded docks, faded colors, soft contrast, slight motion blur, DV compression artifacts, mild sensor noise, no stabilization. 00:00–00:03 She walks carefully along the concrete dock carrying a small fishing rod, stepping around coiled ropes. The camera briefly focuses on the ropes before finding her. 00:03–00:06 She sits on the edge of the dock and prepares the fishing line, concentrating as a gentle sea breeze moves her ponytail. 00:06–00:09 She casts the line into the calm water and waits quietly. A seagull lands nearby, making her turn and smile. 00:09–00:12 The fishing rod suddenly moves slightly. She looks surprised and quickly grabs it with both hands, laughing when she realizes it was only the current. 00:12–00:15 She looks toward the camera with an embarrassed smile, shrugs playfully, and returns her attention to the water. The operator pans toward the boats as the recording abruptly cuts to black. Audio: Natural ambience only—seagulls, gentle waves against concrete, ropes creaking, distant boat engines, fishermen talking, wind, footsteps, and water movement. No music. No narration. Goal: A forgotten MiniDV home video from 2004, capturing a quiet morning at an ordinary Busan fishing harbor—raw, peaceful, imperfect, nostalgic, and completely believable.

# Main Prompt - Idol Full-Body Groove (Classroom Ver.) shot: type: r2v aspect_ratio: "9:16" camera: fixed, single continuous shot, no cuts framing: character occupies lower two-thirds of frame, medium shot from waist up, full arm extension should remain within frame character: reference: use attached reference image for full appearance (face, outfit, colors, proportions) art_style: preserve original character art style / cel-shading from reference, do not blend with background style motion: choreography_style: J-pop idol full-body groove - point dance hand accents combined with strong hip/chest movement, wide arm throws, and traveling footwork; energetic and dynamic, not restrained ground_rule: only beat_7 involves a brief hop; all other beats keep at least one foot grounded at all times, weight shifts through full leg and hip engagement rather than staying stiff sequence: - beat_1_2: wide step to the right with full weight transfer, both arms throw outward and up into a diagonal "V" shape, chest pops forward on the accent, hips follow the arm direction - beat_3: arms whip back down and across the body, torso twists with the motion, right hand ends in a sharp point toward the camera - beat_4: quick recover to center, small chest pop plus hip bump to the right on the offbeat - beat_5_6: mirror of beat_1_2 to the left - wide step, arms throw diagonally the other way, hips follow, chest pop - beat_7: energetic accent beat - one clean, controlled hop in place (both feet leave the floor briefly together, land together), arms pump downward on the landing for emphasis; this is the single high-energy peak of the sequence - beat_8: land, feet planted, body drops into a grounded groove - knees bent, torso rolls in a small body wave from chest to hips - beat_9_10: double point combo - right hand points forward with a step forward, immediately followed by left hand pointing forward with weight recovering back, hips keep swaying underneath - beat_11_12: full-body spin-out - a quick 180-degree turn using a pivot step (feet stay grounded, one foot pivots), arms sweep out during the turn, ending facing camera again with both arms opening into a wide "sparkle" gesture - loop: sequence repeats seamlessly from beat_1 tempo: upbeat, driving, high-energy idol-pop tempo; gestures are large and confident, transitions snap rather than drift motion_notes: no slow motion, no speed ramping, natural constant playback speed throughout; beat_7's hop must read as one clean controlled jump-and-land, not repeated bouncing; all other beats stay grounded with movement driven by hip/torso/arm amplitude rather than vertical lift setting: location: after-school classroom, empty of other students time_of_day: late afternoon / early evening, warm light through windows environment_details: rows of desks and chairs as soft silhouettes in the background, chalkboard

**Main Subject:** Young Korean woman, from reference image, 24, naturally attractive, realistic skin, minimal makeup, long dark hair loosely tied back. Wearing a faded oversized mustard-yellow T-shirt, loose dark-gray sweatpants, white socks, and worn house slippers. Thin silver necklace. Preserve her exact identity, facial features, hairstyle, and appearance throughout. **Location:** Old Seoul apartment on a warm weekend afternoon during moving day. Narrow hallway, worn wooden doors, stacked cardboard boxes, plastic bags, old furniture, partially empty rooms, open windows, tangled cables, dusty corners, neighboring apartment buildings visible outside. Warm natural sunlight mixed with slightly dim interior shadows.\n\n**Style:** Ultra-realistic early-2000s Sony MiniDV home video. Candid family recording, imperfect and unstaged. Heavy handheld movement, awkward framing, autofocus hunting, exposure pumping between bright windows and dark rooms, faded colors, soft contrast, DV compression, slight motion blur and microphone noise. No stabilization or modern cinematic camera movement.\n\n**00:00–00:03:** She struggles through the apartment hallway carrying a large cardboard box that blocks most of her view. She nearly bumps into a doorframe, laughs, and sets it down.\n\n**00:03–00:06:** She opens the box and realizes several smaller items have fallen over inside. She kneels down and starts reorganizing them while someone off-camera tells her where the box should go.\n\n**00:06–00:09:** Two people attempt to carry an old wooden chair through the narrow doorway. They get stuck for a moment, laugh, rotate it awkwardly, and finally squeeze it through.\n\n**00:09–00:12:** She sits briefly on an unopened box, exhausted, drinking from a plastic bottle while looking around the half-empty room. Sunlight pours through the window behind her.\n\n**00:12–00:15:** Someone off-camera calls her name. She immediately gets up and grabs another box. As she walks past the camera, she notices it, gives a tired little smile, and keeps working. The camera follows her into the empty room before accidentally pointing toward the bright window and blowing out the exposure.\n\n**Audio:** Natural sound only—cardboard scraping, footsteps, furniture dragging, tape ripping, plastic bags rustling, distant traffic, muffled conversations, elevator sounds, doors closing and occasional laughter. No music, narration, or added sound effects.

AMBULANCE — FORD F-450 MEDIC UNIT "CODE 3" Character: VALENTINA — 34 years old. Paramedic. 11 years on the job. Has seen everything. Still runs toward it. Her unit is Unit 7 — she has named it Siete. 65mm IMAX. Panavision. Grain: starts at zero — the

Close-up of an emotional monologue. The woman in @image_1 sits at a wooden dining table in the dimly lit, warm wooden kitchen of @image_2—wooden cabinets, deeply blurred bowls, plates, and storage shelves behind her, night outside the window. There are no visible light sources in the frame—no chandeliers, lamps, lighting equipment, or fixtures—only the light itself and how it falls on her face are visible; all light sources are outside the frame. The entire film is shot in a single, continuous long take, without any editing, transitions, or camera changes—the same uninterrupted camera from beginning to end. She remains silent throughout, not uttering a single word—the entire emotion is conveyed solely through facial expressions and eye contact. All emotions shift rapidly, quickly transitioning from one to the next, each stage brief—the performance swiftly glides through each emotional phase in a tight rhythm. Camera: The camera position is fixed in place, but it presents the texture of handheld shooting—vivid handheld tremors, slight shaking and vibrations, the lens drifting almost imperceptibly left and right, breathing with the scene. The camera consistently focuses on her face, a close-up, keeping her centered and within the frame—only the subtle hand-held movement around her creates tension. This trembling, intimate feel builds tension, amplifying the emotional outburst. The light moves in circles around her, an invisible beam smoothly circling her along its trajectory, the illumination shifting with each stage: Stage 1 (Fast) – Suppressing bitterness. Her gaze is lowered, tears welling, her lower lip trembling, her chin furrowed, her breathing labored, her lips pressed tightly together. Lighting: Directly facing her, illuminating her entire face evenly and softly from the front, warm spots of light falling on her face, the surrounding background fading into warm darkness. Stage 2 (Fast) – Tears. Tears stream down her cheeks, she abruptly looks up, blinking to force the tears back, quickly wiping her eyes with the back of her hand, her eyes bright and red. Lighting: Shifts to her right side—now only the right half of her face is illuminated, the left half sinking into shadow, a soft, flowing shadow spreading across her face. Stage 3 (Fast) – A burst of laughter. A silent, hysterical laugh bursts forth through her tears, a mixture of sobs and laughter, her head shaking, her expression shifting from pain to relief. The lighting: continues to move behind her—her entire face sinks into shadow, the light coming from behind, the harsh backlight outlining her head with bright edges, illuminating her ears, cheekbones, and wispy strands of hair, leaving her face mostly dark, almost a silhouette with luminous edges. Fourth Stage (Climax)—Breakdown and Relief. She slams her hand on the table—the instant she strikes, the light suddenly blazes, illuminating her entire body. The lighting during the table slam: a strong, direct beam of

Shot 1: Initial anger, 0 to 4 seconds, medium-close shot. A young couple confronts each other in a evening living room. The girl has her arms crossed, eyebrows furrowed with vertical lines between them, a hard stare fixed on the other person, corners of the mouth tight and downturned, cheeks slightly flushed, side-lit, fixed camera. Shot 2: Transition to grievance, 4 to 10 seconds, camera slowly zooms into a facial close-up. The girl's hard gaze softens slightly, inner eyebrows lift gently, her eyes slide away from the other person's face toward the floor, the corners of her mouth are still down but no longer tight, shoulders slump slightly as if choked up, warm light turns cold, extremely slow zoom. Shot 3: Holding back tears, 10 to 15 seconds, extreme facial close-up. The girl purses her lips tightly to endure, lower lip trembles slightly, eyes slowly redden and fill with water, she blinks once to push the tears back, one drop hangs under her eye without falling, head tilts away slightly reluctant to be seen, natural light, fixed camera.

Style: Psychological thriller performance study, unsettling realism, intimate front-facing portrait, fixed camera, subtle natural lighting, raw human micro-expressions, realistic skin texture,

TITLE: The Lost Wind Chime: Memories in the Wind Use the uploaded storyboard and character sheet as the main visual reference. Follow the storyboard naturally. Maintain character, kitten, wind chime, porch, lighting, and environment consistency from the

1. 0–1.3s: Woman laughs with friends on Korean rooftop under string lights; suddenly rigid, eyes wide. 2. 1.3–2.6s: Head snaps down violently, veins darken; friends fall silent, chairs scrape back. 3. 2.6–3.9s: She convulses violently, overturns table, gasps, then goes still. 4. 3.9–5.1s: Head whips up; bloodshot eyes, slack jaw, animalistic snarls. 5. 5.1–6.4s: She sprints low and unnaturally fast between chairs as friends scream and scatter. 6. 6.4–7.6s: Friend knocks over vending drink rack; cans scatter across rooftop. 7. 7.6–8.9s: She rapidly scrambles over overturned table toward friend by railing. 8. 8.9–10.1s: Cornered friend swings plastic chair, knocking her back. 9. 10.1–11.4s: She instantly lunges again; another friend throws cans at her face. 10. 11.4–12.6s: Friend ducks and scrambles along railing, nearly falling; city skyline behind. 11. 12.6–14s: She grabs railing inches behind him; he slips toward stairwell. 12. 14–15s: Stairwell door slams shut; she crashes against it snarling beneath flickering lights. Realistic Korean rooftop horror, cinematic handheld camera, intense motion blur.

I. Generation Goal Generate a complete, continuous 15-second cinematic-grade Chinese Xianxia short film. Overall Style Requirements: Cinematic realistic texture Pure ancient Chinese Xianxia aesthetics Solemnity of epic martial arts challenges clashing with high-level meta-comedy Incorporating silent film-style reaction pacing Incorporating British deadpan humor Incorporating clean-cut pacing of Hong Kong action comedy Using classic three-beat progression and setup-payoff structure Arri Alexa cinematic texture Clear and stable facial micro-details Fine film grain Natural volumetric lighting II. Core Comedy Mechanism The punchline must be simple and direct, immediately understandable to the audience on the first watch. An extremely confident enemy master believes he is delivering a legendary challenge declaration to intimidate the scene. In reality, the two women in Figure 1 and Figure 2 have long been silently predicting the clichéd lines he is about to say. Moreover, the angrier he gets and the more he tries to prove himself, the more accurately he fulfills their predictions step by step. Finally, the master completes the third payoff. III. Character Identity Anchors Character ID A | @Image 1 | Sword Immortal Senior Sister Always maintain the same character identity and appearance: 25–30 year old East Asian female Tall and slender build Oval face Dark almond eyes Black long hair half-tied White jade hairpin fixation White embroidered silk Hanfu Semi-transparent layered wide sleeves Silver waist girdle Jade pendant White cloth boots The only silver long sword Character Temperament: Calm Restrained Deadpan Minimalist reactions, but always accurately catching the humor points Character ID B | @Image 2 | Junior Sister Always maintain the same character identity and appearance: 20–25 year old East Asian female Small and petite build Round and agile face Black hair braided Blue-green linen Hanfu Dark waist belt Wooden hairpin Black cloth shoes The only dark steel sword Character Temperament: Smart Agile Quick reactions Always looking like watching a play already seen before Supporting Character | Elderly Master Exists only as a secondary witness Not much screen time Deadpan from beginning to end Completes the third payoff at the end Supporting Character | Enemy Swordsman Extremely confident Full of aura Believes he is creating pressure In reality, constantly falls into the prediction of his clichéd language by others IV. Environmental DNA and Spatial Requirements All newly uploaded background and location reference images collectively determine the same set of Environmental DNA. Before formal composition, silently integrate and reorganize the following content: Real terrain Architectural language Spatial scale Materials Vegetation Water bodies Weather Clouds Mountain mist Main light direction Reflection relationships Air depth Character's actual walkable path Re-plan a unique, complete, and unified new space for this round. Background Operating Principles The background is always naturally vivid but absolutely neutral in narration. Elements that can continuously exist and move naturally include: Wind Water bodies Vegetation Mountain mist Banners Distant ordinary disciples Reflections Ambient spatial sound However, these elements must never actively create humor or influence plot causality. V. Shot Structure 0–5s | Shot 1 | Wide or Long Shot Visual Task Establish character relationships and complete the first setup. Visual Content An extremely imposing enemy swordsman strides into an open area naturally formed according to this round's reference images. The same elderly master exists quietly a few steps away from the two. The enemy imposing points his yet-to-be-drawn long sword at the senior sister, announcing loudly: "Sword Immortal! I have practiced for ten years, today I shall let you witness the world's number one sword!" Character Reaction The same junior sister is not nervous; instead, she quietly raises three fingers and says softly: "Third sentence." The same senior sister, upon seeing this, only shows a tiny, clearly "guessed it" satisfied nod. Key Requirement This beat must let the audience understand: they are not afraid, but are "predicting clichéd lines." The enemy must be very serious and imposing. Comedy comes from the nonchalant reactions of the two women. 5–10s | Shot 2 | Medium or Cowboy Shot Visual Task Complete the second prediction payoff and let the enemy realize something is wrong. Visual Content Maintain the same two women, same clothing, same enemy, same master, and perfectly consistent geographic space. The enemy's aura is interrupted for the first time, frowning and asking: "What third sentence?" The same junior sister explains very naturally: "Senior sister guessed your third sentence would be 'world's number one'." The enemy's expression clearly tightens, saying annoyedly: "You dare mock me?" The same sword immortal senior sister does not argue with him, but calmly turns to look at the junior sister, only saying: "Fourth sentence, hit again." Pacing Requirement There must be a clean non-dialogue pause here. Let the enemy realize himself: That angry line just now also belongs to the clichés they predicted in advance. Key Requirement Senior sister has very few lines, but the strongest impact. Junior sister's explanation must be natural, like saying a very ordinary thing. The enemy's confidence begins to crack but has not completely collapsed. 10–15s | Shot 3 | Close-up or Extreme Close-up Visual Task Complete the third payoff, with the master delivering the final blow to form the final punchline. Visual Content The same elderly master, who has been silent all along, finally says flatly from behind: "I guess the fifth sentence is drawing the sword." The enemy completely flies into a rage, blurting out: "Absurd!" And instinctively draws his long sword with a sharp "clank." Reaction Design The whole scene is instantly absolutely quiet. The senior sister, junior sister, and the enemy all three simultaneously and very slowly turn their eyes to the master. The same master maintains a completely deadpan expression from start to finish, only nodding very solemnly: "Fifth sentence." Closing Action Extreme close-up: The enemy looks down at the long sword he has already drawn, finally realizing he has personally completed the prediction someone just said. His entire aura begins to deflate rapidly, and then he very slowly lowers the sword again. The same junior sister tries hard to suppress a smirk and asks softly: "Still fighting?" The enemy silences for half a beat, already terrified by his own mouth and actions, only answering lowly: "...Don't speak for now." The same sword immortal senior sister barely pauses, calmly adding one more line: "Smart." Immediately cuts to black precisely. VI. Acting Principles General Principle Everyone's acting must be restrained Not exaggerated No facial contortions No clowning Comedy only comes from: Prediction and fulfillment Deadpan reactions Pacing difference between characters Gradual draining of aura Senior Sister Minimal reaction But every time she speaks, it's like the final cut Stable expression from start to finish Junior Sister The first person to read the clichés Emotion should be a bit more agile But cannot be too showy Master Always like handling an ordinary sect matter The last line must be as calm as a court verdict Precisely because it's not funny at all, it's funnier Enemy Must sincerely believe he is creating pressure Emotional change path: Confidence Interrupted Confused Annoyed Out of control Self-awareness collapse Forced silence VII. Cinematography and Sound Requirements Cinematography Style 16:9 widescreen Stable and restrained cinematic lens language Clear natural parallax Sharp visuals Fine film grain Real atmospheric perspective Arri Alexa cinematic look Camera Principles Lens movement must be stable Focus on observing character reactions No flashy techniques No gaudy effects Shot switching serves the comedy pause and reaction payoff Sound Requirements Native synchronized Mandarin dialogue Precise lip-sync Comedy pauses must be clear Retain real spatial ambient sound Retain sound of fabric, wind, water, footsteps, and slight weapon sounds The sound of drawing the sword must be clear, becoming an important sound beat for the third payoff VIII. Continuity Requirements Must always maintain stability in: Character identity stability Facial stability Hair stability Clothing stability Sword stability Supporting character position stability Geographic space stability Accurate sightline relationships Clear and readable sword-drawing action Realistic silk fabric and hair physics effects Continuous natural parallax in foreground, middle ground, and background Continuous real spatial sound No subtitles generated IX. Finished Film Requirements Strict total duration: 15 seconds 16:9 widescreen Three continuous clear shots Native synchronized Mandarin dialogue Precise lip-sync Comedy pauses must be clear No subtitles generated No watermarks No random text X. Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inc

Seedance 2.0 | 15-second short film "Shadow Play" I. Generation Goal Generate a complete continuous 15-second cinematic Chinese urban fantasy short film titled "Shadow Play". Overall features: cinematic realistic texture, fusion of traditional Chinese shadow puppetry aesthetics, silent film style spatial comedy, restrained/deadpan performance, contrast between warm yellow sodium lamps and cold blue night colors, clear depth of field. A twist in the last 5 seconds redefines the purpose of the entire "battle". II. Core Creativity This is not a real life-and-death duel. For the first 10 seconds, viewers will think two women are engaged in an epic, oriental-style fantasy shadow duel. The final 5 seconds reveal that they are actually using physical actions and projection relationships to perform a "shadow play" for an unseen child upstairs. The twist must be established through: physical actions that remain small, precise, and restrained while the wall shadows are exaggerated; a child's voiceover redefines the purpose; comedy stems from the contrast between "epic momentum" and "children's game needs". III. Workflow & Technical Preferences Organized for Seedance 2.0 multi-modal reference: supports image, audio, and video references; director-level control over performance, lighting, shadows, and camera movement. Focus on: stable character identities, stable shadow geometry, continuity of sword and bicycle, stable wall projection logic, Mandarin lip-sync, and clear cause-effect between physical and shadow actions. IV. Character Identity Anchors Character ID A (@Image1) - Sword Sister: 25-30 years old, East Asian, oval face, fair skin, sharp brown eyes, long black hair, tall/slender, white hanfu with translucent sleeves, silver sword. Temperament: calm, restrained, serious. Character ID B (@Image2) - Bicycle Sister: 25-30 years old, East Asian, ponytail, yellow jacket, blue jeans, white sneakers, same bicycle. Temperament: tacit cooperation, serious, restrained, deadpan humor at the end. V. Environment Setting Scene: narrow alley in an old town after rain. Elements: rough white plaster wall, wet slate road, warm yellow sodium street lamp, overhead wires, bamboo leaf shadows, flowing mist, reflections in puddles, a bicycle. Lighting: warm yellow sodium lamp as main source, cold blue night as ambient base. Shadows on the wall must be clear, huge, and readable. VI. Environmental Movement Principles Background must stay alive but neutral. Allowed movements: flowing mist, slight vibration of wires, swaying bamboo shadows, changing puddle reflections, hair/fabric responding to movement. Prohibited: sudden background shifts, wall deformation, street lamp jumping, bicycle disappearing, shadows detaching from light logic or人物动作, purely special-effect animation. VII. Shot Structure 0-5s | Shot 1 | Wide Fixed + Slow Dolly: Establish spatial relationship and misleading "battle atmosphere". Huge shadows on the white wall look like ancient masters about to fight. 5-10s | Shot 2 | Cowboy/Medium Shot | Smooth Lateral Tracking: Juxtapose "small real actions" with "epic wall battle". Real movements are small while shadows transform into a giant silver dragon and a phoenix-like beast made of wheels. VIII. Acting Principles Overall: restrained, not slapstick. Character A: serious, deadpan, suppressing a smile at the end. Character B: tacit cooperation, low-key funny remark at the end. Child (voiceover): natural, happy, innocent tone. IX. Photography & Visual Control Style: 16:9, 24fps, Arri Alexa cinematic texture, sharp, natural volumetric light, clear depth. Shots: wide dolly, smooth lateral track, close-up to extreme close-up. No aerial shots, shaky cam, or AI morphing. X. Continuity Requirements Must maintain stability of: character identities, faces, costumes, props (sword/bicycle), lighting, alley structure, night color temperature, shadow geometry, and physical relationships of hair/mist/reflections. XI. Final Product Requirements Total duration: 15 seconds, 16:9 aspect ratio, three clean shots, native sync environmental/action sound, accurate Mandarin lip-sync, no subtitles, watermarks, or random text. XII. Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, glitching cuts, disappearing props, unstable shadow geometry, fake shadow projection, shadow detached from light source, bicycle disappearing, sword morphing, wall deformation, exaggerated acting, slapstick overacting, random fantasy effects, random CGI creatures, modern interface elements

Seedance 2.0 Fast | 15s Chinese Xianxia Suspense Emotional Short Film I. Generation Goal Generate a continuous 15-second Xianxia cinematic film. Style: Realistic texture, suspenseful, restrained emotional temperature, "physical evidence reversal" as core narrative tool. II. Core Theme Redefining "Trust" through high observation: trust is not a verbal declaration but remembering habits. The Senior Sister identifies a forged sword tassel knot because she knows the Junior Sister's specific way of tying it. III. Characters Character A (Image 1): Senior Sister, 25-30, calm, observant, firm judgment. Character B (Image 2): Junior Sister, 20-25, round face, green linen hanfu, moved by the sister's observation. IV. Timeline 0-5s: Enemy throws a green sword tassel as evidence against Junior Sister. All look at her. Senior Sister picks it up silently. 5-10s: Master asks if she trusts her. Senior Sister examines the knot and says "This isn't hers," explaining the specific tying habit. Junior Sister is shocked she knows. 10-15s: Senior Sister returns the forgery, stating she trusts the past seven years of observation, not just words. Junior Sister smiles with relief. Master turns his gaze to the enemy.

# Main Prompt - A: Clockwise Heart (Classroom Ver.) shot: type: r2v aspect_ratio: "9:16" camera: fixed, single continuous shot, no cuts framing: character occupies lower two-thirds of frame, medium shot from waist up character: reference: use attached reference image for full appearance (face, outfit, colors, proportions) art_style: preserve original character art style / cel-shading from reference, do not blend with background style motion: sequence: - beat_1_2: both arms raised, sweep in a slow clockwise circular motion like clock hands, starting from 12 o'clock position - beat_3: arms cross at chest height, hands open - beat_4: arms draw outward and down into a heart shape formed above the head, hold for half a beat - beat_5_6: small step-touch side to side (left-right), shoulders swaying gently in sync - loop: sequence repeats seamlessly from beat_1 tempo: moderate pop tempo, sharp but not rushed transitions motion_notes: no slow motion, no speed ramping, natural constant playback speed throughout setting: location: after-school classroom, empty of other students time_of_day: late afternoon / early evening, golden hour light through windows environment_details: rows of desks and chairs as soft silhouettes in the background, chalkboard faintly visible, window frames casting long soft shadows across the floor atmosphere: quiet, nostalgic, gentle contrast between the character's lively motion and the stillness of the empty room background: style: simple, uncluttered classroom silhouette, low detail so it doesn't compete with character motion rendering: keep visually separate from character line/shading style; desks/windows rendered with minimal linework, mostly shape and light, not full detail motion: static background, only ambient light shift (e.g. faint dust motes in the light beam), no moving elements lighting: warm golden-hour light streaming through windows from one side, soft rim light on character, gentle long shadows cast by desks expression: light, playful smile, eyes forward toward camera during heart pose

The dancer in Image 1 is performing a sexy and cool, intense soul dance. It starts with gigolo-style movements and transitions into cool soul dancing. The dancer is moving stylishly, alluringly, and happily to soulful music. A 15-second high-quality multi-cut music video. Use your best creativity for shot division, shot size, angles, and camera work. Set the background to match the person and the dance. Show me the best 15-second soul dance MV. However, since there are only 15 seconds, basically just keep dancing.

An emotionally charged, avant-garde Japanese anime opening sequence with playful, surreal, graphic visual direction. A cute anime girl moves through an endlessly transforming visual world. Multiple variations of the same girl appear, disappear, overlap, split apart, and replace one another, as if different versions of her personality and identity are competing inside the same opening sequence. Full-color Japanese 2D animation, expressive hand-drawn character animation, clean cel shading, highly polished sakuga, 24fps. [Core Visual Goal] Prioritize flatness, graphic design, symbolic clarity, and emotional coherence. The sequence should feel like a moving 2D graphic-design composition rather than a cinematic world. Every visual element must relate directly to the girl, her emotions, her body, her memories, or the story's central symbolic ideas. Do not introduce random decorative motifs or unrelated objects. Do not use imagery just because it looks stylish. Every repeated shape, symbol, prop, or visual metaphor must feel meaningfully connected to the same character and emotional theme. [Overall Visual Direction] Pop, surreal, stylish, cute, strange, slightly unsettling. Very flat visual design. Strong 2D composition over three-dimensional realism. Large areas of flat color. Minimal depth. Minimal volumetric lighting. Minimal environmental realism. Graphic staging over cinematic staging. The screen behaves like a moving poster, a motion-graphics illustration, or a transforming printed design. Backgrounds are simplified into flat planes, cutout-like layers, symbolic spaces, and graphic fields. When depth appears, it should be stylized and brief, used only for emphasis. Flat images may suddenly bend, stretch, fold, rotate, or transform, but they should still feel fundamentally graphic and planar. [Motif Control] Use only motifs directly derived from the main character and her emotional world. Allowed motifs should come from: her eyes, hair, hands, mouth, silhouette, clothing, accessories, shadows, reflections, mirrors, frames, ribbons, simple body fragments, personal symbolic objects, and abstract shapes that clearly evolve from these elements. Motifs may also come from a small number of central emotional symbols, but only if they are consistently repeated and clearly connected to the girl's identity. Avoid unrelated decorative symbols such as random stars, random flowers, random animals, random cosmic imagery, random surreal props, or arbitrary visual clutter unless they are specifically established as core motifs. Limit the motif vocabulary. Reuse the same small symbolic set across the entire sequence. The world should feel unified by recurring, meaningful visual symbols. [Character Direction] A cute

Landscape iPhone video, one unbroken 30-second take, no cuts. LOCATION / TIME Hongdae, Seoul, blue hour. The pavement is still wet from rain. Neon signage is just starting to come on. Use a long, straight stretch of shopfront-lined sidewalk with enough clear depth for the performer and camera to travel roughly 40 meters continuously. SOUNDTRACK This is a music video. The track is the entire audio. The music is NOT playing inside the scene. It is a finished studio recording laid directly over the footage in the edit, exactly like a professional music-video soundtrack. The music must be: - Clean - Loud - Full-range - Mastered - Immediate and front-of-mix - Filling the entire frequency spectrum, including deep sub frequencies Music style: K-pop dance-break instrumental at around 104 BPM. Musical elements: - Hard 808 sub - Tight, snappy claps - Sparse minor-key synth stabs - Layered female vocal ad-libs - Stacked vocal hooks with no intelligible words Energy progression: The track begins sparse and restrained, then transitions into a full drop that carries the entire second half. IMPORTANT AUDIO RESTRICTIONS: NO street ambience. NO traffic noise. NO footsteps. NO crowd noise. NO voices. NO wind. NO environmental sound. NO room reverb. NO outdoor echo on the music. The soundtrack must NOT sound like: - A phone speaker - A car stereo - A busking speaker - Music playing from a nearby store - Music heard from across the street No muffling. No distance effect. No bandpass filtering. The music must remain studio-clean, full-frequency, loud, and immediate throughout. THE PERFORMER A Korean woman in her early twenties. Appearance: - High ponytail - Oversized varsity jacket - Cropped tee underneath - Wide black trousers - Chunky sneakers She is a professional working music-video dancer and immediately reads as one through her movement quality. She travels forward for the ENTIRE 30-second take. She dances for the ENTIRE 30-second take. The forward walk itself is choreography. Walking is never used as a break between dance phrases. Movement vocabulary: - Hip-hop bounce with strong, visible knee action - Chest isolations - Rib isolations - Shoulder pops - Waacking arm circles - Sharp wrist snaps - K-pop point-choreography accents - Quick heel-toe footwork incorporated into her traveling steps Movement density: Maintain roughly 4 to 6 clearly visible choreographic accents per second. The movement remains sustained throughout. She NEVER drops into ordinary walking. Every accent should feel rhythmically connected to the soundtrack. Her movement amplitude follows the music: - Smaller and more contained during the sparse opening - Increasingly confident as the track builds - Biggest and most explosive during the final drop Performance behavior: She occasionally half-sings along to the track without producing audible diegetic vocals. She catches the camera lens, grins, breaks eye contact naturally, then reconnects with the lens. She feels confident, playful, spontaneous, and extremely comfortable performing directly to camera. THE CAMERA Handheld iPhone footage. Landscape orientation. One continuous 30-second shot. NO cuts. The camera operator walks backward ahead of the performer for the entire take, matching her forward pace. Camera height: Approximately chest height. Framing: Keep her centered and framed from head to sneakers for most of the take. Use a wide iPhone lens perspective. She must remain large enough in the frame that: - Her face stays clearly resolved - Her eyes remain readable - Her hands remain clearly visible - Her dance details remain easy to understand Camera behavior should feel authentically handheld: - Mild natural hand movement - Slight rolling-shutter wobble - Small framing imperfections - Autofocus hunts once or twice - Exposure subtly adjusts as she passes bright storefronts Wet pavement and asphalt reflect the neon signage and surrounding lights. DO NOT use: - Gimbal smoothness - Orbiting camera movement - Zooms - Camera rotation - Slow motion - Drone shots - Stabilized cinematic tracking CROWD BUILD 0-4 SECONDS The performer dances forward alone. Ordinary pedestrians are already moving naturally through the background and around her. Nobody appears to know what is about to happen. 4-9 SECONDS One guy who is already visible behind her begins catching the groove. He starts copying her shoulder choreography approximately half a beat late. His participation should initially feel accidental rather than staged. 9-15 SECONDS Three or four additional people gradually join. Each dancer must peel naturally out of the existing pedestrian flow. Entrances can come from: - The left side of the sidewalk - A visible shopfront - Further behind the performer - Other positions already established within the shot Nobody should suddenly materialize or enter from an impossible location. Each person was already somewhere logically present within the continuous environment. 15-21 SECONDS The group grows to approximately 12 to 15 dancers behind her. They remain loosely packed rather than forming a perfect formation. They copy her accents with slight ripple delays. The timing differences should make the sequence feel spontaneous and contagious rather than perfectly rehearsed. 21-30 SECONDS The number of dancers now HOLDS. Do not keep adding more people. Instead, increase the energy. The group gradually tightens into a loose wedge formation behind the lead performer. Choreographic intensity reaches its peak. The lead performer remains the unmistakable visual focus. FINAL 2 SECONDS She throws her hardest, sharpest choreographic accent directly toward the camera lens. Everyone behind her lands the same beat with her. The movement hits hard and feels satisfying. The shot does NOT freeze. The dancers remain physically alive after the accent. Bodies continue breathing and settling naturally as the 30-second frame ends. BACKGROUND PEDESTRIANS Some pedestrians NEVER join the choreography. They continue walking normally around the dancers. Their reactions vary naturally: - Slightly annoyed - Curious - Amused - Indifferent Some carefully edge around the growing dance group while continuing toward their destination. This contrast between dancers and uninvolved pedestrians is essential for making the moment feel real. PHYSICAL REALISM Grounded human biomechanics throughout. Every movement must show: - Visible weight transfer - Proper foot planting - Real momentum - Believable balance - Natural joint articulation - Correct body mechanics No sliding feet. No floating. No impossible limb motion. No rubbery joints. No unnatural acceleration. No teleporting people. No duplicated pedestrians. The lead performer stays the clear focus throughout the entire take. Maintain enough proximity and image clarity that her face, hands, expressions, clothing, and choreography stay clearly resolved even with the wide lens.

A confident white female rapper @[Image 1](image_1). Double bun braided hair styled high. Wearing a pink Zipp Republic jersey, layered gold chains and cross necklace, gold star-shaped drop earrings, and a small cross tattoo under the left eye. Standing firmly beside the rear quarter panel of a car, swaying weight from heel to toe to the beat. Expressing rhythm with both hands, using cutting motions, pointing, and gesturing as if stacking bars in the air. Vocal Profile: A white woman speaking in a London accent with a Nigerian lilt. Low chest-voice range, crisp consonants, and vocal delivery that strongly punches the end of each bar. Dropping volume to a low murmur between bars. Signature Tic: The moment the hook line starts, thrusting both hands forward to make a double point. However, the moment a bar lands perfectly, dropping shoulders back and raising the chin slightly half a beat before the next line begins. Eye Contact: Staring straight into the lens, blinking slowly and intentionally. In transitions, momentarily dropping the gaze to the man sitting on the car before returning to the lens. The character's appearance must match the reference image 100%. Do not change anything other than the character's appearance. @[Image 2](image_2) - Already image-referenced. White male, early to mid-30s. Short tight black hair, full black beard connected to a trimmed mustache, wide nose, thick lips, solid wide-shouldered build, and prominent tattoos on both forearms. Wearing a red Zipp Republic 2.0 jersey with white outlined graphics and '2.0' numbering on the chest. Layered gold Cuban link chains, black jeans, black low-cut trainers, and a gold watch on the wrist. Sitting on the trunk lid of a red car, placing both feet firmly on the rear bumper for balance. The upper body sways loosely in a relaxed state to the beat, nodding the head low with every snare hit before lifting it slightly. Tapping a thigh lightly with one hand to keep rhythm. Psychological Engine: Possessing a calm sense of intimidation and composure, not flaunting the performance but surrendering to the groove itself. Signature Tic: The chains catching the light with every shoulder roll, while the chin remains stable during nodding. The moment the rapper hits a punchline, turning the face toward her, holding that gaze for one beat before returning to his own rhythm. Eye Contact: Calm, half-closed eyes with a slow, relaxed gaze. Moving sight naturally between the rapper and the lens. The character's appearance must match the reference image 100%. Do not change any other elements. The characters appearing are Western/white only. Both perform singing/rapping with a Nigerian accent.

@Image1=MAINIMG @Image2=PIPIMG MAIN="MAINIMG unique subject"; MAIN_FACE="Main screen character facial identity"; MAIN_HAIR="Main screen character hairstyle base"; MAIN_BODY="Main screen character body proportions"; MAIN_C0="Main screen character original outfit base"; MAIN_ACC="Main screen character key accessory logic" PIPCHAR="PIPIMG unique subject"; PIP_FACE="Lower left character facial identity"; PIP_HAIR="Lower left character hairstyle base"; PIP_BODY="Lower left character body proportions"; PIP_C0="Lower left character original outfit base"; PIP_ACC="Lower left character key identification logic" WALL="Background wall layer"; FLOOR="Floor layer"; CHAR="Main screen character layer"; PIP="Lower left observation window layer" subject: "Read MAINIMG and PIPIMG only once and lock the two subjects respectively. MAIN only belongs to the main screen character; PIPCHAR only belongs to the lower left PIP window character. Both are independent characters, no confusion, no premature swapping, no mistaking PIPCHAR for MAIN, and no mistaking MAIN for PIPCHAR. For most of the video, Main screen = MAIN, lower left PIP = PIPCHAR, with a clear position swap only in the final high-speed transition." vocal: "Use the same original Japanese female lead vocal throughout; lyrics are not pre-written, freely improvised Japanese short phrases by the generation side; clear Japanese articulation and lip-sync; strong rhythmic hooks + short phrases + sustained notes; no quoting existing songs, no reading prompt words, no ah/oh/woo loops. No male lead vocals, no male-female switching, no changing singer due to character swaps." music: "150BPM intense Japanese dance music/Electronic House/High-speed Club; huge four-beat kick, tight snare/claps, rolling low frequency, sidechain synth bass, bright supersaw chord wall, metallic percussion, vocal chops, arpeggios, riser effects, impact. No slow intro. 0-13s each segment has clear downbeats and transition feel; 13-15s enter high-density chopped finale, and complete MAIN and PIPCHAR position swap at the end. The lead vocal remains the same female voice, no changing to male, no changing voice identity due to role swap." VOCAL_IDENTITY_LOCK: "0.00-15.00s always use the same original Japanese female lead vocal. Sound is an independent audio layer, does not change with visual character identity, appearance, gender performance, PIP content, or final role swap. Even if PIPCHAR enters the main screen, the female lead vocal must be completely continuous. No male lead, no male-female switching, no dual lead, no changing singer due to character swap." priority: "1 Main screen MAIN recognition established, 2 lower left PIPCHAR recognition established, 3 same female lead vocal throughout, 4 main screen continuous Japanese singing/dancing, 5 main screen sync 3-layer transition, 6 lower left PIP stability, 7 CHAR description change clearly visible, 8 camera, 9 final swap successful." MAIN_IDENTITY_DEFINITION: "MAIN consistency refers to character identity recognition, not identical pixel-level appearance. The goal is not a pixel-perfect copy of the source image, but for the audience to recognize it is the same MAIN regardless of style changes." MAIN_IDENTITY_LOCK: "MAIN fixed attributes: facial bone structure and feature relative positions; eye shape and expression tone; front hair structure and hairstyle base; head-to-body ratio, physique, body proportions; sense of age and overall temperament; key accessory logic. These form the core of MAIN recognition and must be maintained." MAIN_MUTABLE_ATTRIBUTES: "MAIN attributes allowed to change drastically: drawing style and medium; line language; coloring and shading; material representation; 2D/3D/Print/Craft processing; clothing color; clothing material; clothing details and decorations; stage-style modification of the same clothing base; local accessory enhancement or weakening; lighting effects and outline emphasis; degree of abstraction. As long as fixed attributes remain, these changes are valid." MAIN_FORBIDDEN_DRIFT: "MAIN Forbidden: turning into someone else; facial/feature reconstruction; hairstyle base disappearing; physique, age, or gender drifting significantly; unfamiliar clothing completely detached from original identity; key accessory logic disappearing." PIP_IDENTITY_DEFINITION: "PIPCHAR must also maintain consistent character identity. No matter how the angle, movement, or eating state changes in the lower left PIP, it must always be seen as the same PIPCHAR." PIP_IDENTITY_LOCK: "PIPCHAR fixed attributes: facial bone structure and feature relative positions; hairstyle base; body proportions; sense of age and temperament; clothing base and key identification points. The character in the lower left PIP window must always be PIPCHAR herself." identity_repeat: "MAIN and PIPCHAR must always be clearly distinguished. 0-13s main screen = MAIN, lower left PIP = PIPCHAR. Only the final swap segment allows swapping positions. The audio layer does not participate in this swap." MAIN_SCREEN_MODEL: "The main screen consists of three synchronized layers: WALL = back wall/facade; FLOOR = floor/stage floor/ground; CHAR = MAIN character body. The main screen uses a synchronized switching logic where WALL/FLOOR/CHAR enter new segments together at each major transition point." SYNC_RULE: "The main screen uses synchronized three-layer transitions, not complex asynchronous ones. Goal is execution priority, with each segment clearly changing scene." CHAR_RULE: "Main screen CHAR only refers to MAIN body. All main screen character changes only apply to MAIN: face/head/hair/neck/torso/arms/hands/waist/legs/feet/clothing/accessories. Must still be seen as the same MAIN after change." CHAR_TRANSFORMATION_REQUIRED: "Each MAIN description change must be immediately visible. Just changing color temperature, brightness, slight filters, or weak textures does not count as a change. Each time at least two of the following must change: line language, coloring, material, medium, clothing color, or clothing details." WALL_RULE: "WALL is only responsible for background walls/screens/posters/displays/installations. Mysterious large faces, huge eyes, abstract portraits, posters, screens, graphic character impressions, and huge MAIN close-ups are allowed as wall art, but only as wall images, not independent characters." FLOOR_RULE: "FLOOR is only for the ground. Allows huge facial images, huge eyes, abstract portrait LED floors, and fire/water/grass/neon grid/liquid metal/flower field/ice/prism/pixel ground/metaverse data platforms. These must stay on the ground and not become independent characters or change the MAIN body surface." BACKGROUND_EXTRA_RULE: "WALL and FLOOR can absorb two environmental languages: 1 Metaverse/virtual city/holographic data architecture/transparent UI space; 2 Pixel art/8-bit/16-bit/blocky game landscapes. These only apply to the environment layer and must not change MAIN identity." goal: "15-second high-density experimental MV. 0-13s main screen is a synchronized 3-layer MV of MAIN, lower left PIP is a strange additional video of PIPCHAR eating a burger on a black background. 13-15s final 2 seconds enter high-speed finale transition, finally completing the position swap between MAIN and PIPCHAR characters, but the lead vocal always remains the same female voice." dance: "MAIN performs high-energy club choreography at 150BPM: heel-toe fast steps, fast footwork, side shifts, hip beats, shoulder hits, chest pops, arm waves, locking, body rolls, diagonal moves, half-turns, rotations, bounces, high/low position changes, short jumps, rhythmic head turns. MAIN continues clear Japanese singing and dancing from 0.00-15.00s, never standing still, staying center stage until the final swap." dance_timeline: "0-1.5 heel-toe advance+shoulder hit+arm wave | 1.5-3 diagonal move+chest pop+half-turn bounce | 3-4.5 low side step+upward arm sweep+small spin | 4.5-6 fast two-step+torso twist+backward bounce | 6-7.5 side step+arm lock+forward rhythm | 7.5-9 high/low change+fast turn+hip beat | 9-10.5 running step+controlled spin+arm sweep | 10.5-12 reverse step+body roll+jump landing | 12-13 fast cross step+270 turn+move end | 13-13.5 high-speed A | 13.5-14 high-speed B | 14-14.5 high-speed C | 14.5-15 ultimate burst" MAIN_AUTONOMOUS_STYLE: "When MAIN description changes, the generation side must autonomously invent a new visual style distinctly different from the current one. Don't follow a fixed list. Freely change line language, color systems, textures, 2D/3D, print methods, animation media, craft feel, abstraction, clothing colors, and stage details. Goal is for the audience to see ‘the same MAIN described by another visual culture’." MAIN_STYLE_RANGE: "Directions include but not limited to: comic line drawing, high-contrast American graphic, pop art, retro hand-drawn animation, sticker-style, collectible model, plush, clay, paper-cut, collage, picture book, fashion illustration, 3D cel-shaded, paper pulp craft, pixelated, low-poly, mosaic, stained glass, neon, holographic, blueprint, graffiti, Art Deco, off-register print, silkscreen, etc." COSTUME_RULE: "MAIN clothing uses ‘fixed base, variable expression’ principle. Must keep original silhouette, core identification structure, and wearing logic. Changes allowed: colors, materials, decorations, stage enhancements, pattern density, sheen/reflection, layer additions/subtractions. Forbidden: completely switching to a stranger's clothing system or unrelated identity." COSTUME_CHANGE_REQUIRED: "MAIN clothing must have immediately visible stage changes. After 3s, each stage must clearly change color system, material, or details; can't just rely on lighting/temp/filters. Even if visual medium changes, the clothing itself must independently change." costume_vector: 'C0=original; C1=keep silhouette, change primary colors to bright stage colors, add highlight fabrics and decoration; C2=keep base, change to strong graphic two/three-tone blocks, add geometric borders; C3=keep base, change to deep black/dark main + high contrast bright borders, sharpen decorations; C4=keep base, change to pearlescent/iridescent/translucent future materials, different colors from C3; C5=clear white-gold finale outfit.' COSTUME_VISIBILITY_RULE: 'C1, C2, C3, C4, C5 must be clearly different. Failure to judge stage by clothing is considered a clothing change failure.' costume_timeline: '0-3 C0 | 3-6 C1 | 6-9 C2 | 9-12 C3 | 12-13 C4 | 13-15 C5 finale leaning' MAIN_STAGE_TRANSITIONS: 'Main screen 3-layer sync transitions advance by segment. WALL/FLOOR/CHAR enter a new overall segment together at each switch.' main_stage_timeline: '0-1.5 Stage1: WALL=pop art wall+abstract graphics+high contrast colors; FLOOR=huge stranger face LED floor; CHAR=close to original high-detail, light stage mod; Mood=Opening grab. 1.5-3 Stage2: WALL=poster collage/ad visual/torn edges; FLOOR=huge eyes/iris dynamic floor; CHAR=clear comic line or high contrast graphic; Mood=First major style change. 3-4.5 Stage3: WALL=folding screen style/gold ground/traditional painting; FLOOR=fire surface/heatwaves/cracks; CHAR=retro hand-drawn animation or picture book style; Mood=Traditional+Heat. 4.5-6 Stage4: WALL=large display wall/CRT matrix/projection; FLOOR=transparent water/ripples/mirror; CHAR=sticker or graphic poster style; Mood=Digital+Fluid. 6-7.5 Stage5: WALL=MAIN huge close-up face art; FLOOR=real grassland/blowing grass; CHAR=collectible model or 3D cel-shaded; Mood=Strongest MAIN identity emphasis. 7.5-9 Stage6: WALL=metaverse city/holographic buildings/UI/data arch; FLOOR=metaverse electronic platform/grid; CHAR=holographic/neon/future stage; Mood=Tech Nightclub. 9-10.5 Stage7: WALL=large pixel art wall/8-bit city; FLOOR=pixel ground/bricks/water; CHAR=pixel-leaning or low-poly but clearly MAIN; Mood=Gaming electronic feel. 10.5-12 Stage8: WALL=mysterious giant face/abstract portrait; FLOOR=liquid metal surface; CHAR=mosaic/stained glass/craft; Mood=Heterogeneous art. 12-13 Stage9: WALL=graffiti/misprint/neon sign mix; FLOOR=flower field to ice prism; CHAR=mixed media, finale outfit push; Mood=Finale build-up. 13-13.5 Burst A: high-frequency sync switch, compress space, enhance speed. 13.5-14 Burst B: continue sync switch, prepare for PIP intrusion. 14-14.5 Burst C: boundary destruction, screen erosion, swap starts. 14.5-15 Finale: complete MAIN and PIPCHAR swap, main screen subject becomes PIPCHAR, MAIN enters lower left PIP.' camera_lock: 'Main screen camera must move significantly in 3D space, not fixed station panning/tilt/digital zoom. Each segment must change X/Y/Z coordinates with visible parallax. No stationary shaking to fake orbiting, no digital zoom only.' orbit_rule: 'Each orbit must complete at least 90-degree change, focus segments 120-180. Clockwise/Counter-clockwise must be clear. Vertical orbits must rise from low to high over the subject.' frame_rotation: 'Screen rotation must be continuous 30-90 degrees or more, overlapping with orbit/push/pull, not random shaking.' camera_timeline: '0-1.5 front-left low clockwise 120 to right, knee to chest height | 1.5-3 continue clockwise 90 to right-back, high speed graze past shoulder to re-catch face | 3-4.5 back-low vertical orbit, over head to high-angle then descend forward | 4.5-6 front-right low counter-clockwise 150 to left-back, radius shrinks then expands | 6-7.5 high-angle counter-clockwise fast descend to ground-level graze | 7.5-9 ground sprint: dash from far to MAIN, sharp pull up at feet, over shoulder then 90 degree reverse orbit | 9-10.5 giant spiral: 120 orbit + rise + radius contraction + 60 roll | 10.5-12 super-high dive past shoulder to back then 180 flip back to front | 12-13 ultra-close to ultra-wide reverse flyover + huge WALL/FLOOR then high-speed push-in | 13-13.5 fast clockwise 90 orbit + 45 roll | 13.5-14 fast counter-clockwise 120 orbit + dive | 14-14.5 spiral 120 + 90 screen roll + screen tear | 14.5-15 complete swap from broken screen and relock new protagonist' PIP_RULE: 'Add a small PIP window in the lower left. Not a split screen, but an independent small PiP, approx 1/16 area (width/height 22-25%). Keep it small and clear, don't interfere with main screen. Fixed position, can have white line/glow border.' PIP_BACKGROUND: 'PIP environment fixed to pure black or dark void. No complex background or floor patterns. Goal is clear view of PIPCHAR eating a burger.' PIP_SUBJECT: 'Display PIPCHAR in the PIP window, not MAIN. This is an independent curious video. For 0-13s, PIP character must be PIPCHAR.' PIP_PURPOSE: 'Show PIPCHAR eating a burger from various angles on a black background. Atmosphere slightly comical/mysterious, like a random insert, but visually clear and cute.' PIP_ACTION_CORE: 'PIPCHAR holds a burger or thick meat patty and clearly eats: looking at it, lifting, biting, chewing, swallowing, biting again. Must clearly see eating, not just holding.' PIP_HAMBURG_RULE: 'Burger must be clearly visible (bun, patty, toppings). Key point is PIPCHAR seriously eating. No other food, no eating air, no just posing.' PIP_ANGLE_RULE: 'PIP shows PIPCHAR with eating action. Can slowly turn, rotate, or small continuous changes while keeping eating visible. Show front, 45, side, back-turn, half-body close-up, etc.' PIP_MOTION_STYLE: 'Minor movements only: turning, head tilting, lifting burger, biting, chewing. No major dancing. Focus is multi-angle continuous eating.' PIP_AUDIO_RULE: 'PIPCHAR does not sing or provide vocals. Does not take over song. Performs silent or low-presence eating. No impact on the female lead vocal.' PIP_TIMELINE: '0.8-2.8 front and left 45 eating, first clear bite | 2.8-4.8 left side and left-back 45, chewing while turning | 4.8-6.8 back and right-back 45, looking back while holding burger | 6.8-8.8 right side and right 45, another clear bite | 8.8-10.8 upper body close-up turn, highlight face and burger, clear chewing | 10.8-12.8 small continuous turntable-style observations, different angles continuous eating | 12.8-13.8 faster eating, sensing main screen abnormality | 13.8-15.0 PIP boundary sucked into main screen and expanded, PIPCHAR bursts out of lower left window to become main screen character.' PIP_STYLE_RELATION: 'Even if main screen changes are intense, prioritize PIPCHAR identification and eating visibility. Style can follow finale atmosphere but black background and burger are core.' PIP_MAIN_SEPARATION: '0-13.8s MAIN continues dancing/singing in main screen. PIP is just an additional layer. Both exist simultaneously without PIP interfering with MAIN facial visibility.' SWAP_RULE: 'Final swap is the core experimental event. 13.8-15.0 a clear position swap must occur: PIPCHAR breaks PIP boundary, expands via high-speed transition to main center; MAIN is compressed, shrunk, pulled into the lower left PIP window. After swap, main subject = PIPCHAR, lower left PIP = MAIN. Vocal identity remains unchanged.' SWAP_VISUAL_GRAMMAR: 'Use screen tearing, mirror flips, PiP expansion, window burst, layer penetration, character through-screen, rotate/scale, glitch, etc. Swap is not a simple cut but a visual spectacle.' SWAP_FINAL_STATE: 'By 14.7-15.0, new main subject must be PIPCHAR (optionally with burger); MAIN must be in the PIP window. Swap must be completed. Female lead vocal remains unchanged.' PIP_FAILURE: 'Failed if PIP window is too large, not black background, not PIPCHAR, no eating, or no final swap.' typo: 'Minimal dynamic text, single kanji or short English only. Candidates: Dance, Light, Sound, Instant / GO, UP'

A vibrant K-pop stage performance under bright purple and pink LED lights. Three young East Asian women idols with long dark hair stand in a line on stage, wearing headset microphones and stylish crop tops. Characters and outfits: Left idol: lime-green sleeveless top with white “BOYS LIE SPORT” text and heart logo. Center idol: baby-pink spaghetti-strap crop top with white piping. Right idol: navy-and-green horizontal striped collared crop top. Sequence — approximately 17 seconds: 0–2s: The three idols begin by dancing lightly and naturally in sync to the beat. The center idol flips her long hair and turns slightly while dancing. The right idol dances with a playful expression. Do not have them cover their faces yet. Soft stage lighting and large LED screens show a big red “5” and heart graphics in the background. 2–6s: While continuing their dance movement, all three suddenly bring both hands up to cover their mouths, reacting with genuine surprise and laughter. Their shoulders shake as they giggle and they gently bounce in place. The center idol laughs hardest, with her long hair naturally swaying. 6–12s: They continue the synchronized cover-mouth-and-laugh motion while still subtly dancing to the beat. Keep their bodies moving naturally rather than freezing in place. Their expressions are joyful, playful, and genuinely amused. Medium camera framing captures all three idols clearly. 12–15s: The dancing continues at 12 seconds as they smoothly transition from covering their mouths into making double peace signs with both hands raised near their faces. They keep moving rhythmically while smiling and laughing. The center idol briefly turns and then faces forward again. 15–17s: Final pose while the dance energy continues subtly — all three hold double peace signs near their faces, beaming directly at the camera with bright, cheerful expressions. Soft stage haze and colorful LED lighting create a high-energy yet adorable K-pop concert atmosphere. Visual style: high-quality live concert footage, sharp details, natural skin texture, realistic facial expressions, dynamic but soft purple-and-pink stage lighting, subtle handheld camera shake for realism, natural hair movement, believable body motion, polished K-pop performance cinematography, cute, energetic, wholesome atmosphere. Important motion requirements: The idols must dance first before putting their hands on their faces. Do not start with their hands covering their mouths. The dancing continues throughout the sequence, including after the 12-second mark. No frozen poses before the final moment. No text overlays.

15-second cinematic fantasy dark-pop music video, 16:9, 1080p/4K-quality look, ultra-photorealistic, premium high-fashion MV finish.

A woman with long black hair dancing in a cozy room. When she forms frames, triangles, or squares with her hands, overlays appear inside them. It starts as an anime-style avatar, transitions to a Spider-Verse-inspired comic book style in the middle, and finally smoothy returns to live-action.

Generate a women's clothing display video. Character identity and clothing should strictly refer to Image 1, maintaining consistency in the character's face shape, features, hairstyle, body proportions, and clothing style. Scene and environment should strictly refer to Image 2, using Image 2 as a background reference, trying to keep the spatial structure, setting, color tone, light direction, and overall atmosphere consistent, so that the character blends naturally into the scene. Actions should refer to the provided black-and-white depth video, extracting only the character's body posture, dance movements, rhythm, limb trajectories, and motion relationships in the shot. The black-and-white depth video is only for motion control reference; do not refer to its black-and-white visual effects, character identity, appearance, clothing, background, original scene, materials, or colors. The final video should show: the character from Image 1, in the real scene from Image 2, naturally performing a women's clothing display according to the movements in the depth video. Keep character identity, clothing, and background stable, with smooth and natural movements, avoiding facial drift, clothing changes, background changes, limb distortion, and character flickering.

Create a 15-second photorealistic Chinese Brown Sugar Boba Milk commercial using @image1 only as a visual reference for the adult Chinese female model, milk-tea shop, transparent cup, brown-sugar syrup, black tapioca pearls, fresh milk, lighting, and scene progression.

Create a 15-second ultra-premium shampoo advertisement featuring a beautiful young woman with long, naturally thick, glossy hair. The entire film should look like a high-budget international beauty campaign, with sophisticated cinematography, realistic hair physics, elegant lighting, premium production design, and photorealistic detail. 0–3 SEC — OPENING Extreme close-up of the girl standing in a luxurious modern bathroom. Warm morning light softly illuminates her face and wet hair. She gently runs her fingers through her damp hair as tiny water droplets glisten on individual strands. Camera slowly moves toward her hair with an elegant macro transition. 3–6 SEC — SHAMPOO APPLICATION Close-up of her hand dispensing a rich, luxurious pearlescent shampoo into her palm. Cut to a beautiful close-up as she massages the shampoo gently into her wet hair. Rich creamy lather forms naturally between her fingers. Detailed hair strands, realistic foam, water droplets, and soft skin texture. 6–10 SEC — TRANSFORMATION Cinematic slow-motion shot. She rinses her hair under crystal-clear water. As she lifts her head, her freshly washed hair falls naturally around her shoulders. She turns toward the camera and slowly runs her fingers through her hair. Her hair appears silky, smooth, hydrated, voluminous, and intensely glossy, with natural movement and realistic individual strands. Soft sunlight creates elegant highlights across the flowing hair. 10–13 SEC — HERO BEAUTY SHOT The girl walks slowly through a luxurious minimalist setting. She turns her head gently, allowing her long, silky hair to sweep naturally through the air in beautiful slow motion. Camera performs a smooth cinematic orbit around her. Confident expression. Natural beauty. Premium fashion-editorial aesthetic. 13–15 SEC — PRODUCT REVEAL Match-cut from her flowing hair to the shampoo bottle standing on a polished marble surface. The bottle is surrounded by subtle water droplets with a soft luxury light glowing behind it. Elegant text appears: [BRAND NAME] UNLOCK YOUR HAIR'S NATURAL BRILLIANCE. Final subtle light sweep across the logo. STYLE: Luxury beauty commercial, international cosmetics campaign, photorealistic, 4K/8K, cinematic slow motion, premium studio lighting, realistic wet hair, physically accurate water and foam, natural skin texture, elegant camera movement, shallow depth of field, sophisticated color grading, high-end product cinematography. AVOID: Cheap CGI, excessive effects, unrealistic hair movement, plastic-looking skin, distorted hands, warped facial features, messy backgrounds, excessive text, oversaturation, artificial-looking foam, cartoon aesthetics.

Create a polished 20-second, 16:9 photorealistic automotive design-to-road commercial using the uploaded reference board as strict vehicle/storyboard

15 seconds | 16:9 | live-action fashion runway footage\nReal outdoor fashion event in Times Square, New York City.\nRaw professional camera footage, observational realism.\nNo commercial-style fantasy treatment.\n\nREFERENCE CONTROL\n@Image1 and @Image2 = exact dress reference.\n\nThe model wears the exact couture gown shown in the references.\n\nPreserve the dress construction precisely:\nstrapless fitted metallic champagne-gold bodice,\nstructured waist,\nextremely voluminous floor-length ball-gown silhouette,\nirregular overlapping sculptural tiers,\ndense silver, charcoal-black and champagne embellishment,\nintricate beadwork and embroidery,\ndark heavily decorated layered edges.\n\nDo not simplify or redesign the gown.\nMaintain the exact dress throughout every shot.\n\nSETTING\nExclusive open-air couture runway in Times Square, New York City during late afternoon.\n\nTimes Square architecture, illuminated digital billboards, storefronts, city streets, trees and a seated fashion audience surrounding a long outdoor runway.\n\nSoft overcast daylight.\nReal New York City fashion event atmosphere.\n\nMODEL\nAdult high-fashion runway model.\nTall, slender proportions.\nMinimal makeup, understated hair.\nSerious neutral runway expression.\nControlled professional walk.\n\n00–04s — ENTRANCE\nLong telephoto shot from the photographers’ end.\n\nThe model enters the runway and walks directly toward camera.\n\nThe heavy layered skirt begins moving naturally with her steps.\nAudience members watch and occasionally raise phones.\n\n04–08s — WALK\nMedium frontal tracking shot.\n\nCamera moves backward as she approaches.\n\nHer upper body remains composed while the enormous skirt sways with realistic weight and delayed momentum.\n\nMetallic embroidery catches subtle daylight.\n\n08–12s — DETAIL\nLow three-quarter side angle as she passes close to camera.\n\nShow the dimensional beadwork, embroidery and layered construction.\n\nCamera pans naturally with her.\nForeground spectators briefly partially obscure the gown.\n\n12–15s — RUNWAY END\nTelephoto full-body shot.\n\nShe reaches the runway end, pauses and performs one restrained couture turn.\n\nHer body rotates first; the heavy skirt and train follow with delayed momentum.\n\nPhotographers fire irregular flashes.\n\nCAMERA\nReal fashion-week camera coverage.\n\nTelephoto and handheld professional camera footage.\nNatural operator micro-movement.\nMinor reframing.\nReal autofocus adjustments.\nNatural motion blur and depth of field.\n\nNo floating camera.\nNo impossible movements.\nNo slow motion.\n\nLIGHTING\nNatural outdoor daylight.\nSoft overcast illumination.\nPhysically accurate reflections from metallic embellishments.\nNo artificial glow or cinematic spotlighting.\n\nPHYSICS\nThe gown is heavy and behaves accordingly.\n\nLayers sway, overlap, collide and settle naturally.\nThe train remains in contact with the runway.\nFabric follows footsteps, body rotation and light wind.

[Style] Hollywood Live-Action Automotive Sci-Fi short film, First-person POV AR/MR Interface, Hard-Surface Mechanical Transformation, Photorealistic, 4K, 30fps, realistic metal, inertia, weight, and conservation of part mass, no anime feel, no toy plastic feel. [Duration] 30 seconds [Aspect Ratio] 16:9 Landscape [Scenes] Walnut computer desk in a pure black studio with a desktop holographic interface; white polygonal virtual car showroom; European Gothic bell tower city bridge at rainy dusk; high-altitude snow mountain road, ice surface, rock peaks, low clouds, and powder snow. [Characters] Character 1: POV operator showing only tattooed arms and hands, face not shown; Character 2: The same Rosso Corsa red Ferrari 12Cilindri and its red armored robot form; Character 3: The same black unmanned attack helicopter attacking from off-screen and later clearly appearing in shot. [Absolute Sequence Lock] The entire film is in strict forward order, prohibiting flashbacks, cold opens, and repeated events: Start from POV holographic car selection; the white space on the desk is always a virtual car showroom in an AR/MR volumetric window, not a physical miniature manufacturing cell; after the vehicle in the virtual showroom completes digital loading, the camera rushes into the holographic space, and the showroom and vehicle transition into a full-scale real showroom and real car; while the red Ferrari drives in its complete car form, the helicopter first strafes continuously off-screen with 'du-du-du' sounds, the impact tossing the car into the air, and the car only then initiates mechanical transformation to respond; the same helicopter continues the pursuit, and the robot counterattacks to hit and destroy it; only after clearly seeing the burning enemy aircraft and confirming the threat is cleared can the robot change back into the same red Ferrari; finally, snow mountain driving, aerial title shot, and pulling back to the tablet at the beginning. Never let the car transform actively without cause, and never let the helicopter become a new attacker only after the robot is completed. [Vehicle and Robot Consistency Lock] Only one Ferrari 12Cilindri throughout the film: fixed front V12 long hood, rearward cockpit, low two-door Berlinetta silhouette; fixed high-gloss Rosso Corsa red paint, black roof and glass, slender cool white LED headlights, four round red taillights, black forged wheels, yellow wheel center Prancing Horse badge, yellow brake calipers, yellow fender shield badge, black carbon fiber front lip/side skirts/diffuser. No shot can turn into a black, gray, orange car, or other Ferrari models. Strictly four tires throughout. The robot's internal skeleton is fixed as gunmetal titanium, and the external car shell armor always remains the same Rosso Corsa red; the long red engine hood forms the breastplate, the headlights become cool white chest lights, the two doors become shoulder wings, the two front wheels are fixed behind the left and right shoulders, the two rear wheels are fixed on the outside of the left and right calves and heels, the windshield and black roof become the upper back, and the red rear fenders and taillights become the back plate. Each part moves along clear hinges, hydraulic rods, or slide rails, detached once, rotated once, locked once; prohibition of copying, interspersing, dissolving, and adding parts out of thin air. [Shots 00:00-00:30] Shot 1: POV Holographic Selection and VR Expansion. Shot 2: Digital loading and transition to full-scale real car. Shot 3: Real showroom to city bridge transition. Shot 4: Off-screen helicopter strafing tossing car into air. Shot 5: Mid-air transformation to robot. Shot 6: Robot dodging strafing and locking target. Shot 7: Robot destroys helicopter with arm cannon. Shot 8: Threat confirmation and weapon retraction. Shot 9: Reverse transformation back to car. Shot 10: High-speed snow mountain driving. Shot 11: Aerial view with title 'OVER//DRIVE'. Shot 12: Zoom out to show the video playing on a tablet on the walnut desk. [Negative Constraints] No flashbacks, no toy plastic feel, no other car models, no missing parts, no organic melting, no ghost parts, no robotic limbs appearing out of nowhere, no helicopter exploding before being hit, no random UI text, no anime style, no low-detail CGI.

A cinematic, stylish sequence of a confident young woman with short curly brown hair, bold red lipstick, and a statement necklace. She wears a sharp grey pinstripe oversized blazer and matching trousers with bright red pointed high heels.\nStarts with a close-up of her red heels stepping out of a black luxury car onto a wet city street at dusk. She walks confidently through revolving glass doors into a modern office building, presses the elevator button, checks her tablet in the mirrored elevator, then enters a boardroom. She presents charts and graphs on a large screen to a group of executives in suits, gestures confidently while speaking, signs a document, shakes hands with a senior executive, checks her elegant wristwatch with a burgundy leather strap, and finally walks across a glass-walled rooftop terrace overlooking the New York City skyline at golden hour, wind lightly moving her hair, powerful and composed expression.\nMoody cinematic lighting, wet reflections, sharp fashion photography style, high-end commercial aesthetic, 16:9.

FORMAT:16:9 widescreen, 15 seconds, realistic influencer-style lifestyle video, cinematic 4K, natural lighting, premium fashion aesthetic, fast-paced editing, authentic creator energy. 0:00–0:03 — HOOK Influencer looks into the camera while holding her bag. Dialogue: “Okay, these are the things I’m obsessed with right now.” 0:03–0:05 — BAG Close-up of the bag as she picks it up. Dialogue: “First, this bag. I literally take it everywhere.” 0:05–0:07 — PERFUME Macro shot of perfume as she sprays it on her wrist. Dialogue: “And this perfume? My current favorite.” 0:07–0:09 — SUNGLASSES She puts on her sunglasses and looks confidently at the camera. Dialogue: “These sunglasses make every outfit better.” 0:09–0:11 — SHOES Close-up of her shoes as she starts walking. Dialogue: “And these shoes are way too good.” 0:11–0:13 — PHONE She checks her phone while walking through a stylish café or city street. Dialogue: “Obviously, my phone never leaves my hand.” 0:13–0:15 — FINAL SHOT Quick montage of all five items. She smiles and walks away. Dialogue: “Yeah… I’m obsessed with all of them.” STYLE: Natural influencer delivery, conversational tone, subtle facial expressions, accurate lip-sync, clean dialogue audio, upbeat background music underneath, quick 1–2 second cuts, smooth whip transitions, handheld camera movement, shallow depth of field, realistic product textures, premium but authentic social-media aesthetic.

PROJECT 15-second vertical 16:9 fashion UGC GRWM video. Super casual real smartphone home-video footage of a young woman getting ready for a spontaneous weekend night out. The video feels like an authentic memory captured on her phone and later posted as a GRWM. Natural, imperfect, intimate, stylish, and believable. No commercial polish. MAIN SUBJECT Young Korean woman in her early 20s, natural everyday appearance, realistic skin texture, minimal makeup, warm expressive personality, long dark wavy hair with loose natural strands. She wears a fitted white ribbed tank top and relaxed low-rise blue jeans at the beginning of the video. Final outfit: oversized black leather jacket over a fitted white top, dark charcoal mini skirt, sheer black tights, black knee-high boots, small black shoulder bag, and minimal silver jewelry. Maintain identical facial identity, skin tone, body proportions, hairstyle, and appearance throughout the entire video. Clothing and accessories must remain physically consistent once introduced. LOCATION Small lived-in bedroom with a full-length mirror, unmade bed, clothing casually placed on a chair, open wardrobe, small vanity, cosmetics, jewelry tray, handbag, shoes, and warm apartment lighting. Soft blue evening light enters through the window while a warm bedside lamp illuminates the room. The environment should feel personal and naturally messy rather than staged. No studio environment. No luxury showroom. No commercial set. VISUAL STYLE Super casual real smartphone home video footage. Authentic social-media GRWM aesthetic. Realistic smartphone dynamic range, natural skin texture, imperfect exposure, subtle digital sharpening, ordinary indoor noise, realistic motion blur, believable shadows, and natural color reproduction. The footage should feel spontaneously recorded at home by a real person. No beauty filters. No cinematic color grading. No artificial film grain. No CGI appearance. No overly perfect skin. No fashion-commercial polish. CAMERA STYLE Super casual real smartphone home video footage captured by the creator herself. Slight authentic handheld shake, imperfect framing, occasional awkward camera positioning, small arm movements, natural walking bounce, subtle autofocus hunting, realistic exposure adaptation between window light and warm room lighting, slight smartphone sharpening, normal frame rate, natural motion blur, and occasional focus shifts. The phone is frequently repositioned between her hand, the bed, and the mirror. Use casual selfie footage, handheld rear-camera footage, mirror recording, close-up phone footage, and spontaneous room-level framing. Some compositions should be slightly off-center or imperfect. No professional stabilization. No gimbal. No cinematic tracking. No dramatic lens changes. No artificial depth of field. No polished fashion-film cinematography. PERFORMANCE Natural creator behavior. She is not acting like a model. She casually talks to the phone, makes small spontaneous reactions, checks her outfit, fixes her hair, laughs at herself, and moves naturally around the bedroom. Dialogue: “Get ready with me for tonight. I had no idea what to wear, but I think this might actually work.” Natural conversational delivery with realistic pauses, breathing, blinking, and accurate lip synchronization. ⸻ STORYBOARD 00:00–00:03 — “WHAT DO I WEAR?” Selfie-style handheld phone footage. She sits on the edge of her bed surrounded by two or three clothing options. She looks into the phone and says: “Get ready with me for tonight.” She quickly holds up a jacket toward the camera, makes an unsure expression, then looks toward the clothes on the bed. The framing is slightly crooked as she adjusts the phone in her hand. She laughs quietly. 00:03–00:06 — OUTFIT CHANGE Quick natural handheld transition. She is now wearing the white top, dark charcoal mini skirt, sheer black tights, and oversized black leather jacket. She stands in front of the full-length mirror while holding the phone. She pulls the jacket into position, looks at herself, then looks at the phone screen. She says: “I think this might actually work.” The camera briefly loses focus as she moves closer to the mirror before recovering naturally. 00:06–00:09 — ACCESSORIES Close handheld phone shot from the vanity. She picks up small silver earrings from the jewelry tray. Cut naturally to a mirror angle as she puts them on. She quickly runs her fingers through her hair, allowing loose strands to fall naturally around her face. She grabs the small black shoulder bag from the bed. Natural jewelry movement and realistic hand interaction. 00:09–00:12 — SHOES + FINAL CHECK Low casual phone angle. She sits on the edge of the bed and pulls on her black knee-high boots. The phone is slightly tilted, creating an imperfect home-video composition. She stands up, grabs her bag, and looks at herself in the mirror. She turns slightly from side to side to check the outfit. Natural fabric movement, realistic leather reflections, believable body mechanics. 00:12–00:15 — FINAL GRWM MOMENT Handheld mirror shot. She steps backward and gives the phone a quick full-outfit view. She adjusts the shoulder bag and lightly fixes her hair. She looks directly into the phone camera, smiles naturally, and says: “Okay… I’m actually obsessed.” She gives a small laugh, picks up her keys, and walks toward the bedroom door. The phone moves slightly with her hand. Recording ends naturally mid-motion. ⸻ AUDIO Authentic smartphone microphone recording. Natural bedroom ambience, subtle room tone, clothing movement, jewelry sounds, footsteps, bag movement, keys lightly jingling, and realistic household sounds. Her voice should sound naturally recorded in the room rather than professionally recorded. Very subtle background music may be present at low volume, similar to casual social-media GRWM content, but natural voice and room ambience remain dominant. No cinematic sound design. No exaggerated transitions. PHYSICAL REALISM Realistic human anatomy and hand proportions. Natural walking, sitting, standing, turning, and dressing movements. Correct hand-to-clothing interactions. Realistic leather jacket movement. Natural skirt and fabric folds. Realistic tights and boot interaction. Natural hair physics with individual strands moving during head movement. Realistic jewelry reflections and movement. Correct mirror reflections. No duplicated objects. No warped fingers. No impossible reflections. No floating clothing. No unnatural body mechanics. CONTINUITY LOCK Maintain the exact same woman throughout the entire video. Preserve facial identity, skin texture, hairstyle, body proportions, and natural appearance. Preserve the bedroom layout and lighting direction. Once the final outfit appears, maintain the exact same jacket, top, skirt, tights, boots, jewelry, and handbag. No spontaneous wardrobe changes. No changing accessories. No hairstyle changes. GENERATION REQUIREMENTS 15 seconds. 9:16 vertical. Super casual real smartphone home-video aesthetic. Authentic GRWM behavior. Realistic handheld phone movement. Natural imperfect framing. Realistic autofocus hunting. Natural exposure adaptation. Normal frame-rate motion. Accurate lip synchronization. Consistent identity. Consistent wardrobe. Realistic hair, fabric, leather.

WOMAN = a 26-year-old British woman, natural relatable UGC style, warm fair skin that reddens easily in the sun, light freckles across nose and cheeks, shoulder-length wavy light-brown hair loose, soft brows, blue-grey eyes, everyday face with minimal makeup, wearing a strappy pastel summer top with sunglasses pushed up on her head. Authentic girl-next-door energy. Appearance only. CITY = a sunny city street on a blazing hot summer afternoon, bright pavements, shopfronts, street trees, heat haze shimmering, an intensely bright sun high in a clear blue sky, harsh direct sunlight, strong warm highlights and glare. Environment only. DELURE = a sleek modern sunscreen bottle, clean minimal pastel-and-white packaging, the brand name "DELURE" printed clearly and legibly across the front in a modern sans-serif font, premium skincare look. Product only. SCENE: a warm, authentic UGC-style sunscreen advert. A young British woman walks through a sunny city on a blazing hot day; the harsh sun is reddening her cheeks. She stops, takes a DELURE sunscreen bottle from her bag, and applies the cream to her cheeks; the redness fades and her skin calms back to normal, healthy and happy. She smiles, relieved. The advert ends on a clean product shot of the DELURE bottle. TECHNICAL: 16:9 (or 9:16 vertical if preferred for UGC), bright natural handheld UGC-style camera, warm sunny summer color grade, crisp and clean, realistic self-filmed advert feel, shallow depth of field on the product shots, photorealistic. CUTS: CUT 1 (0-3s): Zoom straight into the blazing bright sun high in the clear blue sky, intense glare and lens flare, heat haze. Then tilt down to the woman walking along the sunny city street, squinting slightly against the heat, fanning herself. CUT 2 (3-6s): Camera zooms in on her FACE — warm and sunlit, her cheeks visibly flushed and reddened from the harsh sun, a slightly uncomfortable expression. Close on the reddened cheeks. CUT 3 (6-9s): She stops, reaches into her bag, and pulls out the DELURE sunscreen bottle. She squeezes a little cream onto her fingertips and gently applies it to her reddened cheeks, smoothing it in. CUT 4 (9-12s): Close-up on her CHEEKS — the redness visibly fading and calming, her skin returning to its normal healthy warm-fair tone, soothed and even. She relaxes, the discomfort gone, and smiles softly with relief. CUT 5 (12-15s): Camera shifts to a clean hero product shot of the DELURE sunscreen bottle held up in the sunlight (or resting on a bright surface), the brand name "DELURE" sharp and clearly legible, sunlight catching the packaging. A polished final beat. RULES: References are appearance only, do not recreate. Keep the woman's face and identity consistent across every cut. The brand name "DELURE" must be spelled correctly

Photorealistic live-action video, shot on cinema camera, natural skin texture, realistic lighting, 16:9 horizontal. Setting: A cozy indoor room with warm ambient lighting, posters on wall, hanging lamp, modern furniture. Subject: A young blonde American woman with shoulder-length golden blonde hair, fair skin, wearing a gray hooded jacket. She faces the camera from waist up. The video has TWO distinct phases — do NOT repeat the same motion throughout. === FIRST HALF (seconds 0–5): Horizontal spring stretch === - Both hands at face height. Thumb and index finger on each hand form L-shaped viewfinder corners facing inward. - Purely horizontal motion: hands close together in front of her face, then pull outward left and right to shoulder width, then pull back to center. Repeat this lateral spring cycle 2–3 times. - Smooth, elastic sideways pull. Forearms stay level. Elbows move outward as hands spread. === SECOND HALF (seconds 5–10): Varied finger-frame gestures === - Transition naturally into different finger-frame poses. She rotates and flips her wrists so the thumb-index frame orientation changes. - Always keep thumb and index finger extended upright forming the corner of a viewfinder on BOTH hands. - End with hands at a relaxed wide finger-frame pose. NOT animated, NOT cartoon, NOT 3D render, NOT anime — real filmed footage. Stable camera, no zoom, no cuts.

[Speech rule — read first] The woman speaks JAPANESE ONLY. She never speaks Chinese, English, or any other language. She speaks almost continuously for the whole clip. She says exactly this one sentence, broken into four phrases, in this order and nothing else: 「Seedance 2.5なら」 「黒板に日本語の文字を」 「書くことができます。」 「とても便利ですね」 She delivers it at the unhurried pace of a teacher writing on a board: her voice slows to the speed of her own chalk, so each phrase is spoken while her hand is forming that same phrase. Voice and chalk advance together. Only a short natural breath separates one phrase from the next — no long pause anywhere. Total speech fills roughly 13 of the 15 seconds. The only silence is the final 0.8 seconds, where she has finished writing and simply smiles and nods. Forbidden at all times: ad-libbed dialogue, filler words, repeating a phrase, reading the sentence twice, humming, laughing, sighing, and every wordless vocalisation such as "んっ" "うん" "ふふ" "ah" "mm" "hm". She says the four phrases once each and nothing more. Pronounce "Seedance 2.5" as English "seedance two point five". [Reference] @Image1 = the keyframe: a woman in a white coat standing beside a completely empty green chalkboard, holding a slim wooden pointer at her side. Strictly preserve @Image1's face, hair, glasses, white coat, the chalkboard's position and size, the easel, the background shelves, and the framing. Nothing moves in the frame except her arm, her upper body, her face, and the chalk marks she draws. [One-line summary] 15 seconds. A cheerful presenter gives a short lecture, writing four lines of Japanese on a green chalkboard with the tip of a wooden pointer while speaking the same words aloud. Pixar-style 3D animation. Locked-off static camera. [Global setup] Environment: warm indoor study room, soft warm ceiling light, shallow depth of field, background shelves stay blurred. Highly realistic physical texture on the chalk dust and the wood of the easel. Visual style: Pixar-style 3D animation, warm color grade, soft shadows. Camera language: medium shot, eye level, straight-on. The camera is completely locked off — no pan, no tilt, no zoom, no push-in, no handheld shake. One camera behavior only: none. Character: as in @Image1. Keep the glasses, the hair silhouette, and the coat identical. Keep believable skin shading and fine surface detail — do not make her look plastic. Acting core: a warm, confident teacher who enjoys showing something off. *** WRITING RULE — SHE IS THE ONE WRITING, AND SHE WRITES AS SHE SPEAKS. *** Every chalk mark is produced by her own hand movement, in time with her voice. For each phrase: 1. Her shoulder and elbow carry the pointer tip to the start of the line. Her torso turns slightly toward the board and her eyes go to that point first. 2. The tip presses lightly to the board and travels along the path of each stroke. The chalk line appears directly behind the moving tip, following it exactly, never running ahead of it, never appearing anywhere the tip has not touched. 3. Characters are formed one stroke at a time, in correct Japanese stroke order, left to right. Faint chalk dust puffs at the tip and a soft scraping sound follows it. 4. Between strokes the tip lifts a few centimetres, then sets down at the start of the next. 5. The syllables she speaks land together with the characters she is forming — when her voice reaches a word, her hand is writing that word. Voice and chalk stay in step. 6. At the end of each line she lowers the pointer slightly, takes one small breath, glances at the camera, then moves to the line below. Her arm leads and the mark follows — the timing of the arm and the timing of the mark must match frame for frame. Marks NEVER fade in as a whole word. Marks NEVER appear while her arm is at rest or while the tip is away from the board. No invisible hand. No magic writing. The arm motion is fluid and human: the wrist flexes, the elbow opens and closes, the shoulder carries the reach across the board, her body weight shifts slightly as she reaches. Once a line is finished it stays on the board unchanged for the rest of the shot. Japanese script rule (important): Render every kanji with its correct stroke structure and correct number of strokes — 黒 板 日 本 語 文 字 書 便 利. They must be legible as real Japanese characters, not invented or Chinese-simplified lookalikes. Write them at a generous size so the strokes stay separated. "Seedance 2.5" is written in Latin letters and digits exactly as spelled: S-e-e-d-a-n-c-e, space, 2, period, 5. Chalk colors — only two, never any other: YELLOW chalk for "Seedance 2.5" only. WHITE chalk for everything else. Layout rule (important): Four lines, left-aligned to a common left margin, stacked evenly down the board from top to bottom, filling the board by the end. Large, cheerful hand-lettered chalk style with slightly uneven strokes. Exact content, top to bottom: Seedance 2.5なら 黒板に日本語の文字を 書くことができます。 とても便利ですね Never centre a line. Never write small. Each line is written as large as it can be while still fitting: a completed line spans almost the full writable width of the board, from the left margin nearly to the right edge of the board, and the four lines together fill the board from top to bottom. Characters are wide and generously spaced so that dense kanji such as 黒 板 語 書 便 利 keep their strokes separated and legible. If a line would not fit, make the characters narrower rather than shorter in height — never shrink a line so that it occupies only part of the board's width. Do not add any other word, symbol, arrow, underline or drawing. Forbidden: no subtitles, no captions, no on-screen UI, no watermark, no background music, no camera movement, no cutting to another shot, no extra people, no second hand entering the frame, no stick of chalk held in her free hand, no distortion of lines already written. [Timestamp storyboard] 0.0-3.6s LINE 1 Action: she raises the pointer to the upper-left of the empty board and writes "Seedance 2.5なら" in YELLOW chalk — first the Latin letters and digits, then the kana なら — the marks trailing the moving tip. Speech (in step with the writing): 「Seedance 2.5なら」 Intent: a small proud lift in her eyebrows on the brand name. This is the thing she is showing off. Camera: locked. No subtitles. No BGM. Japanese only. 3.6-7.2s LINE 2 Action: she drops to the next line and writes "黒板に日本語の文字を" in WHITE chalk, stroke by stroke. Her eyes follow her own tip. Speech (in step with the writing): 「黒板に日本語の文字を」 Intent: matter-of-fact, the setup half of the sentence. Keep the arm steady and even. Camera: locked. No subtitles. No BGM. Japanese only. 7.2-11.0s LINE 3 Action: she drops to the third line and writes "書くことができます。" in WHITE chalk, stroke by stroke, ending with a firm dot for the 。 Speech (in step with the writing): 「書くことができます。」 Intent: this is the payoff of the sentence — her voice firms up slightly on ます, and the final dot is pressed with a small decisive tap. Camera: locked. No subtitles. No BGM. Japanese only. 11.0-15.0s LINE 4 — the close Action: she writes "とても便利ですね" in WHITE chalk on the bottom line, then lowers the pointer to her side, turns her face fully to the camera, breaks into an open warm smile and gives one clear nod. She holds that pose for the final 0.8 seconds in silence. Speech (in step with the writing, finishing by 14.2s): 「とても便利ですね」 Intent: the nod is the close. Warm, a little playful, as if sharing a good discovery. Camera: locked. No subtitles. No BGM. Japanese only. [Closing constraints — restated for the whole clip]

Create a 15-second vertical 9:16 UGC-style phone case advertisement that feels authentic, premium, and social-media ready. A confident young woman in her mid-20s stands in a bright, modern apartment with soft natural daylight. She holds her smartphone toward the camera, showing a sleek premium phone case with a clean minimalist design. She smiles naturally and speaks directly to the camera as if recommending the case to a friend. Her appearance, outfit, phone, and phone case remain identical throughout the entire video. She speaks directly to the camera: “I’ve been using this case for a few weeks, and I honestly love it. It’s slim, feels amazing in my hand, and still gives my phone the protection I need. Plus, it looks so clean. Definitely one of my favorite phone accessories.” Storyboard: Scene 1: Close-up of the phone and case in her hand as she introduces it, rotating the phone slightly to reveal the back, edges, and camera protection. Scene 2: She holds the phone naturally in one hand, showing how slim the case looks while pressing the buttons and running her fingers across the smooth textured surface. Scene 3: She casually places the phone on a table and picks it back up, demonstrating the everyday practicality and secure grip of the case. Scene 4: She takes a quick mirror selfie, then turns the phone toward the camera to show how the case complements the phone’s original design. Scene 5: Final hero shot holding the phone beside her face with a clean, premium apartment background. She smiles confidently at the camera and says, “Definitely recommend.” The video should have realistic smartphone camera quality, natural handheld movement, genuine facial expressions, subtle autofocus adjustments, realistic lighting, smooth transitions, accurate lip sync, and a polished yet authentic UGC aesthetic. The phone and case must remain physically consistent throughout every shot. The case geometry, color, material, buttons, camera cutouts, edges, and proportions remain unchanged. The case stays sharply detailed and clearly visible whenever shown, with no warped branding, distorted geometry, or changing product design.

MODE: REALISTIC MOTION: MEDIUM DURATION: 15-18 SEC REF: STORYBOARD GRID Real woman at vanity getting ready. Thick tortoiseshell square frames, wavy dark brown hair, houndstooth blazer, dark burgundy nails, natural glass skin with visible pores. Warm Hollywood vanity bulb lighting. Beauty products on table. Shallow DOF. Kodak Portra 400. CUT 1-MEDIUM CLOSE-UP, SLIGHT LOW ANGLE: Leans forward, one eyebrow raises, direct eye contact into camera. Jaw slightly tense. She says "Okay be honest." CUT 2-OVER-THE-SHOULDER, HIGH ANGLE: At mirror dabbing lip tint on lower lip, slowly turns toward camera with knowing smile. She says "You thought I was real." CUT 3-EXTREME CLOSE-UP, EYE LEVEL: Dark nails tracing cheekbone, touch under eye, head tilts. Glasses catching light. She says "The face. The eyes. The voice. CUT 4-MEDIUM SHOT, STRAIGHT ON: Sets lip tint down on vanity. Turns to camera. Completely blank expression. Total stillness. She says "None of it exists." CUT 5-MEDIUM CLOSE-UP, SLIGHT HIGH ANGLE: Picks up lip tint tube, twists cap open, confident smirk, slight head tilt. She says "Pure Al." CUT 6-TIGHT CLOSE-UP, EYE LEVEL: Slow eyebrow raise, corners of mouth curl into smirk. HARD FREEZE on expression. She says break your brain-" "And here's the part that will Natural lip sync throughout. Micro expressions. Real blinking. Hair movement naturally shifting. Feels like a real creator talking to camera. --no robotic movement, no puppet face, no plastic skin, no uncanny valley, -no grain, no shaky camera, no raw footage, no different characters, no inconsistent face, no Al look

CAMERA: Handheld DV 16mm vlog footage. CHASE films herself at arm’s length and occasionally places the camera on a bench, gym bag, or near the stretching mat. Keep subtle hand movement, drifting composition, autofocus hunting, rushed reframing, uneven zooms, and brief accidental face cropping. The camera itself is never visible. LOOK: Warm analog tape texture with gentle grain, slightly reduced sharpness, soft halos around overhead lights, small exposure shifts, low contrast, and realistic skin tones. STYLE: An intimate end-of-workout diary filmed in a mostly empty gym. The mood is casual and unpolished: tired breathing, small pauses, quiet humor, relaxed movements, and natural smiles. CHARACTER: CHASE — a beautiful white Brazilian Instagram fitness model and lifestyle creator in her 20s. Long wavy chestnut-brown hair tied in a loose ponytail, expressive brown eyes, glowing skin, and a slim athletic figure. She wears a fitted long-sleeve workout top, high-waisted leggings, white socks, sneakers, and a towel loosely resting around her shoulders. SETTING: A quiet boutique gym late in the evening. Stretching mats, a mirrored wall, dumbbell racks, a workout bag, and a protein shake nearby. Warm overhead lighting with a few darker corners in the background. SCENES: CHASE drops onto the mat and catches her breath. “Okay… that session is officially over.” She stretches forward over one leg and smiles at the camera. “I definitely pushed that last set too far.” She rolls one shoulder and slowly switches sides. “Tomorrow is going to be interesting.” Close-up as she reaches into her gym bag and pulls out a protein shake. She takes a sip, pauses, and looks pleasantly surprised. “Wait… this one is actually good.” She raises both arms above her head and le both arms above her head and leans gently to the side. “I’d call today a solid eight and a half.” She rests her arms on one knee, still slightly out of breath. “Exhausted, but in the best way.” She stands, throws the bag over her shoulder, and gives the camera a small wave. “I’m calling it. See you at the next workout.”

A beautiful young woman named Aria is standing in a bright, modern kitchen wearing a stylish cream outfit. She is holding a premium portable blender filled with colorful fresh fruits. She looks directly into the camera with a warm smile and speaks naturally with perfect English lip sync. Character Dialogue (Voice): Hi everyone! I'm Aria. If you love fresh smoothies as much as I do, you're going to love this portable blender. It's USB rechargeable, incredibly powerful, and perfect for home, work, the gym, or travel. Just add your favorite fruits, press one button, and enjoy a fresh, healthy smoothie anytime, anywhere. Make every sip count!" Actions: Aria smiles and shows the blender to the camera. She adds strawberries, kiwi, mango, and spinach into the blender. She presses the power button and the blender starts blending smoothly. Close-up shots of the spinning blades and vibrant smoothie. She pours the smoothie into a glass, takes a sip, smiles confidently, and gives a thumbs-up. Final cinematic close-up of the blender with the brand logo. Ending Text on Screen: "Fresh Smoothies. Anytime. Anywhere." Style: Ultra-realistic, luxury commercial, cinematic camera movement, premium product advertisement, 4K UHD, soft natural lighting, realistic facial expressions, synchronized English voice, accurate lip sync, shallow depth of field, clean modern kitchen, vibrant colors, professional advertising quality, no subtitles, high-end brand aesthetic.

Duration: 15 seconds Aspect Ratio: 16:9 Style: Authentic UGC / iPhone selfie-vlog, handheld, natural light, TikTok/Reels aesthetic. Product Reference: Use the uploaded gourmet burger image as the only product reference. Preserve the bun shape, patty thickness, cheese melt, lettuce, tomato, sauces, and proportions exactly in every shot. Character Description Name: Hana A young Japanese woman (Image 1) in her early 20s with natural beauty, long dark hair in a loose ponytail, oversized cream sweatshirt, minimal makeup, bright smile, friendly lifestyle-vlogger personality. Shot Breakdown SHOT 1 (0–2s) — Selfie showing the burger box. Dialogue: "Burger night!" SHOT 2 (2–4s) — Opens the box. SHOT 3 (4–6s) — Quick zoom on the burger. SHOT 4 (6–8s) — Hands lifting the burger with cheese stretching naturally. SHOT 5 (8–10s) — Bite reaction. Dialogue: "Okay... that's incredible." SHOT 6 (10–12s) — Casual close-up b-roll while reaching for fries. SHOT 7 (12–14s) — Toasting the burger toward the camera. Dialogue: "You need this." SHOT 8 (14–15s) — Freeze frame with overlay: "burger cravings = solved 🍔" Look & Feel Warm apartment lighting, genuine phone footage, slight grain, natural autofocus breathing, handheld imperfections, fast jump cuts. Negative Prompt cinematic grading, commercial production, CGI burger, fake cheese, distorted hands, warped food, perfect stabilization, studio lighting, text glitches, logo distortion.

A photorealistic smartphone selfie vlog that looks exactly like a real mobile phone recording. A young woman (image = her face and hair) visits a cozy thrift store on a sunny afternoon, filming everything herself with a handheld smartphone in selfie mode. The video features natural hand shake, realistic autofocus, slight exposure changes, authentic smartphone stabilization, and true-to-life colors. Smiling at the camera, she says, "Today's thrift store challenge—I'm only buying the first cute thing I find!" She walks through colorful racks of vintage clothes, shelves of accessories, plush toys, and home décor until she spots a random adorable oversized hat. Laughing, she immediately tries it on in front of a mirror, making funny poses and bursting into genuine laughter at how silly she looks. She decides to buy it, walks out of the store holding a small shopping bag, proudly shows her surprise find to the camera, smiles brightly, waves, and says, "See you in the next vlog. Bye!" before reaching toward the phone to stop the recording. The video should feel completely real with natural human movement, consistent facial features, realistic hand interactions, authentic thrift store lighting, no beauty filters, no CGI, no AI-plastic appearance, no subtitles, no logos, no watermark, and no background music—only real store ambience, footsteps, quiet conversations, clothing rustling, and natural environmental sounds.

A aesthetic, warm, first-person POV cooking scene featuring a young woman preparing Japanese Omurice: Prep: Close-up of raw eggs cracked into a white bowl, followed by a cube of butter melting in a hot cast-iron skillet.
 Whisking & Interaction: A young woman with blonde hair and a light blue oversized hoodie whisking eggs with chopsticks in a modern white-tiled kitchen. A hand gently tucks a strand of hair behind her ear as she looks up and smiles.
 Cooking: She pours the eggs into the pan, swirling to form a smooth, soft omelet, then carefully places it over a molded bed of fried rice.
 The Reveal: A knife slice opens the omelet down the middle, revealing a runny, soft-scrambled interior drizzled with ketchup as she smiles next to the finished plate.

Concept: A creative mixed-media video blending live-action culinary footage with dynamic 2D line-art animations using a stylus interaction. Scene Breakdown 00:00 - 00:01 | The Prep Visual: A hand taps a light-blue stylus onto a slice of bread soaking in egg mixture. Effect: The real-life scene instantly transforms into a stylized black-and-yellow 2D cartoon sketch. 00:01 - 00:04 | The Cook Visual: Top-down view of the cartoon bread cooking in a cast-iron skillet with melting butter. Effect: A stylus tap converts the animated toast into golden-brown, real-life French toast sizzling in the pan. 00:04 - 00:08 | The Plating Visual: Cartoon-drawn French toast on a dark plate, accented with animated steam and powdered sugar sifted overhead. Effect: A final stylus touch morphs the animated plate into a gourmet dish complete with powdered sugar, fresh berries, mint leaves, and maple syrup pouring from a boat.

Create a 10-second cinematic 3D animated food video titled “MAKING MANGO JUICE”. Use the uploaded storyboard as the STRICT visual and shot-sequence reference. IMPORTANT: Follow all 8 storyboard shots in the exact order. Do not add, remove, merge, or rearrange shots. Each shot lasts exactly 1.25 seconds. Total video duration: exactly 10 seconds. Maintain EXACT character consistency throughout the entire video: The same young woman appears in every character shot. She has dark brown hair in a casual ponytail, expressive brown eyes, a friendly youthful face, a yellow short-sleeve T-shirt, and a dark olive-green apron. Do not change her face, hairstyle, clothing, body proportions, or character design between shots. Maintain the same warm, cozy kitchen, lighting direction, color palette, props, counter, cutting board, mangoes, blender, and glass throughout the sequence. VISUAL STYLE: High-quality cinematic 3D animated film, polished stylized realism, detailed food textures, realistic mango pulp, natural liquid physics, expressive character animation, warm natural kitchen lighting, soft shadows, cinematic composition, smooth camera movement, realistic materials, vibrant but natural colors. SHOT 1 — 0.00–1.25s WIDE ESTABLISHING SHOT. The woman stands behind the kitchen counter surrounded by fresh ripe mangoes. She looks toward the camera, smiles warmly, and gives a small friendly wave. A clean empty drinking glass is visible on the counter. Camera makes a very subtle cinematic push-in. Natural character movement and realistic hand motion. SHOT 2 — 1.25–2.50s MEDIUM SHOT. Cut smoothly to the same woman. She confidently raises one ripe mango beside her face and looks toward the camera with a cheerful expression. Mango slices and pieces are arranged on the cutting board in front of her. Her other hand rests naturally near the cutting board. Keep her face and outfit identical to Shot 1. SHOT 3 — 2.50–3.75s CLOSE-UP. Cut to her hands preparing the mango. She holds the mango firmly on the wooden cutting board and makes a clean controlled knife cut. Show the knife moving naturally through the mango flesh. Capture detailed golden mango texture and small natural juice movement. No exaggerated or dangerous knife motion. SHOT 4 — 3.75–5.00s EXTREME CLOSE-UP. Fresh golden mango cubes fall naturally into a transparent blender jar. Several mango pieces drop from above and bounce naturally inside the blender. Use realistic gravity and food physics. Camera follows the falling mango pieces with a subtle downward movement. SHOT 5 — 5.00–6.25s MEDIUM CLOSE-UP. The same woman pours water or milk from a transparent measuring jug into the blender containing the mango pieces. Show a continuous realistic liquid stream. The woman looks focused and cheerful. Keep the blender, mango pieces, kitchen and lighting consistent. SHOT 6 — 6.25–7.50s PERFECT OVERHEAD TOP-DOWN SHOT. Camera looks directly down

A cinematic product demo of a luxury smartwatch. Open with macro shots highlighting the display, premium materials, smooth camera movement, dramatic lighting, and product features before showing the final use case. Ultra-realistic commercial style, 4K. A cinematic product demo of a luxury smartwatch. Open with the finished commercial showing the smartwatch in use during a workout. After revealing the outcome, transition into the workflow and close-up feature shots with premium lighting, smooth cinematic movement, ultra-realistic commercial style, 4K.

Behind-the-scenes phone footage from inside a large practical-effects soundstage, shot vertically on a handheld modern smartphone. >> Scene context: a film crew is shooting the sinking of a miniature RMS Titanic. A gigantic practical ocean tank set, dressed as the North Atlantic, contains a 1:24 scale Titanic complete with illuminated portholes, four funnels, lifeboats, deck railings, rigging, cargo cranes, and miniature passengers. Above the tank, a scaffold gantry carries a row of huge dump tanks loaded with practical seawater; two techs stand on the gantry at the release levers. >> First frame: already loaded and ready — the ship and ocean tank fill mid-ground screen-right, dump tanks looming above the set; in the foreground screen-left, within 1 meter of the phone, three crew in black EFFECTS CREW t-shirts stand behind a low containment wall, one on a cinema camera on rails, another on a headset with a hand raised. A huge blue chroma-key wall with orange cross markers rises behind the horizon. A technocrane hangs above the set. >> Action and physics: a shout of “Action!” — the dump tank gates bang open in sequence and a massive surge of real water crashes across the ocean set with real weight and turbulence, the first wave breaking over the bow before racing along the decks, tearing away lifeboats, snapping miniature railings flat, then slamming into the ship: lower decks disappear beneath the waterline, the bow plunging under the surface, the stern rising higher under the hydraulic gimbal, funnels collapsing one after another as the rigging whips, the hull splitting apart at midships before the stern slides beneath the churn; the water bursts over the containment lip and a fine spray cloud rolls across the stage floor, whiting out the frame as the crew shield their faces and stumble back. >> Camera: operator standing then backing away, phone in one hand at chest height, fast handheld pan following the action from bow to stern, framing loose and reactive, hard jolt as the spray cloud hits, autofocus hunting through the whiteout, water droplets clinging to the lens in the final beat, 84° wide diagonal FOV, camera about 2 meters from the containment wall. >> Lighting: overhead LED soft panel grid as primary source, top-front on the camera side, bright daylight-balanced, exposure locked for the reflective water so foreground crew read darker; the spray cloud glows backlit under the panels. >> Audio: on-set ambient — a loud “Action!”, the metallic clang of dump tank gates, the dense roar of cascading water, snapping miniature steel, the groan of the hull breaking apart, crew shouting off-camera: “Hold your marks! Keep rolling! Camera two, stay on the stern!” >> Realistic amateur smartphone capture, natural handheld shake and rolling shutter, real practical effects, documentary behind-the-scenes footage, no cinematic composition, no film grain, no artificial mot

Handheld home-video style montage, 7 continuous shots with varied angles (avoid a single camera angle or one-take). Filmed one-handed on a phone with slight handheld shake, natural outdoor daylight, subtle film grain, photorealistic realism. A woman Image1 repots a houseplant alone on a cozy apartment balcony. Image1 is used only for her face and hairstyle. She wears a faded denim short-sleeve shirt over a white tank top with a dirt-smudged canvas gardening apron. The narrow balcony has potted plants, terracotta pots, potting soil, gardening tools, string lights, and warm midday sunlight. She is the only person in the video. The plant goes from being root-bound in a cracked plastic pot to freshly repotted and watered. Dialogue is natural spoken Korean. Shot 1 (0–2s): She taps and squeezes the old plastic pot to loosen the plant. "자, 나와라~" Shot 2 (2–4s): Close overhead as she removes the plant, exposing tightly coiled roots and gently loosening them. Shot 3 (4–6s): She places it into a terracotta pot, fills it with fresh soil by hand, and says, "훨씬 낫다." Shot 4 (6–8s): She brushes dirt from the leaves, straightens the plant, and says, "There we go." Shot 5 (8–10s): She wipes her dirty hands on her apron and smiles with a small approving nod. Shot 6 (10–13s): She waters the plant with a small metal watering can. Soil darkens as it absorbs the water. "쭉쭉 마셔." Shot 7 (13–15s): She steps back, leans on the balcony railing, admires the plant, wipes her brow, and exhales contentedly. Audio: No music—only natural sounds of tapping plastic, crumbling soil, rustling roots, patting soil, fabric movement, pouring water, light breeze, birds, and a quiet satisfied breath. No subtitles, text, logos, or watermarks. Do not recreate or copy the reference image itself.

Create a cinematic 3D architectural visualization of a traditional Korean Hanok house on a technical blueprint background. Show a curved dark Giwa tiled roof, exposed wooden beams, raised stone foundation, Hanji sliding doors, and a wooden balcony. Add blue architectural schematics, measurements, floor plans, and cross-section overlays. Use a smooth camera rotation from side view to front view, with a colorful rainbow scanning light effect highlighting the roof and wooden joints. Warm interior lighting creates a peaceful, elegant atmosphere while showcasing the structure and design details.

CAMERA / LOOK: Handheld mini DV camcorder footage filmed by the subject herself. Slight hand shake, occasional focus hunting, imperfect framing, natural zoom adjustments, soft tape-like image quality, subtle grain, realistic auto-exposure shifts from bright kitchen morning light. Natural skin tones, mild motion blur, authentic consumer camcorder aesthetic. STYLE: Cozy coffee-prep vlog with gentle ASMR elements. Relaxed pacing, minimal dialogue, candid moments. Focus on satisfying sounds: bean grinder whirring, portafilter tamping, steam wand hissing, cup clinking, milk frothing. SUBJECT: Young man in his mid-20s, plain t-shirt, hair slightly tousled, minimal accessories. Calm, focused energy while making his morning coffee. SETTING: Small kitchen counter with an espresso machine on a bright morning. Natural daylight, coffee beans and a mug nearby, quiet atmosphere. STORYBOARD: → (3s, propped medium shot) Places camera on the counter, switches on the machine. "Morning coffee, the proper way." → (3s, overhead shot) Grinds fresh coffee beans, fine grounds falling into the portafilter. → (3s, close-up) Tamps the grounds down firmly and evenly. → (3s, handheld shot) Locks the portafilter into the machine. "Here we go." → (3s, detail shot) Espresso streams slowly into a small cup. No dialogue. → (3s, medium shot) Pours cold milk into a small steel pitcher. "Time for the milk." → (3s, macro shot) Steam wand hissing as it froths the milk. → (3s, propped shot) Pours frothed milk carefully over the espresso, forming light layers. → (3s, warm ending shot) Holds the finished cup, takes a small sip, satisfied smile. "That's exactly what I needed." → (3s, final shot) Reaches toward camera, still holding the cup. "See you later." Hand covers lens as recording ends. AUDIO NOTES: Natural kitchen ambience — grinder whirring, tamping, steam hissing, milk pouring should be clearly audible. Dialogue quiet and casual. REALISM NOTES: Authentic body language, natural blinking, genuine focused smiles, occasional careful pauses while pouring, imperfect framing, focus breathing, bright morning light shifts. Should resemble a genuine personal coffee vlog on a consumer camcorder, not a commercial or AI-generated production.

A tranquil animated forest picnic by the riverside, where every ingredient comes straight from nature. Hands scoop silverfish from a clear running stream, rinse fresh green vegetable leaves, and slice fish fillets and vegetables on a wooden cutting board. Spices are sprinkled into a cast-iron pot over an open campfire—fish, greens, and mushrooms simmer together into a rich, steaming soup in a clay pot. Steamed rice is served on a rustic wooden table alongside a glass of fresh coconut juice, warm sunlight filtering softly through the treetops. Cinematic animation style, vivid colors, soft natural lighting, ultra-detailed, peaceful atmosphere, 4K.

Live-action + flat 2D sticker composite, POV cooking vlog, vertical 9:16, 10 seconds, 8K, light handheld micro-shake. Realistic kitchen + funny flat sticker character. Scene: Home kitchen from first-person perspective. Beef and greens are frying in a black pan, oil sizzles, steam rises. White tiles, sauce bottles, sink on the right, side daylight. Character: Little chibi raccoon Solka as a flat 2D sticker: gray-blue fur, dark mask around eyes, striped tail, big red bow, round black glasses, yellow dress, red shoes. Thick outline, paper texture, colored pencil style. Sitting on a small stool by the stove. 00:00–00:03 — Salty Avalanche Real hand stirs meat with a spatula. Solka smiles cunningly and pours a whole jar of salt into the pan. Salt falls like a waterfall, forming a white mound. SFX: sizzling, salt pouring. 00:03–00:05 — Oops Hand takes the jar away, next to her head the spatula makes a comical "BONK!". Glasses slide down, tail stands on end, Solka jumps. SFX: bonk, cartoon spring. 00:05–00:08 — Salty Payback Solka tastes the oversalted food. Cheeks puff up, eyes become spirals, cartoon tears spray from eyes. SFX: crunch, pause, exaggerated crying. 00:08–00:10 — Salty K.O. She swallows, eyes turn into Xs, falls off the stool like a paper sticker. Stars around head, final freeze-frame. SFX: croak, soft thud, funny spirit sound. Negative prompt: 3D raccoon, realistic fur, redesign, extra limbs, deformed hands, morphing, flicker, blur, fake salt physics, gore, subtitles, watermark, UI.

Live-Action + 2D Anime Sticker Composite, First-Person Cooking POV, Realistic Kitchen Texture vs Flat Cartoon Style, HD, Static Camera, 【Duration】 15 seconds 【Scene】 Real Home Kitchen First-Person View: Black skillet on a gas stove cooking sliced mushrooms and butter. Background features white subway tiles, hanging copper pans, ceramic jars, a wooden cutting board, and bright natural light from a window. 【Character】 Q-Version Anime Sticker Character: Flat 2D sticker style, thick outlines. Small girl with light blue hair, a green leaf hat, brown cape, and white apron. Simple dot eyes and rosy cheeks. Pure 2D flat quality, unaffected by real lighting, portraying a cheerful but accident-prone helper. Shot 1: Cooperative Stirring. First-person POV: Real hands use a wooden spoon to stir mushrooms in the pan. The 2D girl stands on the stove, cheerfully holding the spoon handle to help stir. Sound Effects: Sizzling butter. Shot 2: Pepper Seasoning. Real Real hands bring in a tall wooden pepper grinder. The girl hugs the grinder to assist as pepper dust falls over the food. The real hands then exit the frame, leaving her on the stovetop. Sound Effects: Grinding clicks, sizzling. Shot 3: Pepper Sneeze Launch. Floating pepper dust surrounds the girl. She inhales it, her face scrunches up, and she unleashes a massive cartoon sneeze. The sheer force, accompanied by flying cartoon leaves, launches her backward through the air. Sound Effects: Large "Ah-choo," wind whoosh. Shot 4: Dizzy Bowl Landing. The girl flies off-screen and lands in a beige ceramic mixing bowl on the counter. Her leaf hat flies off. She pops her head up with swirling spiral eyes, her tongue hanging out, and cartoon dizzy rings circling her head. Sound Effects: Clatter, cartoon dizzy twinkling.

A 4-view character reference sheet of a stylish 20-year-old girl with long wavy black hair, trendy sunglasses pushed up on her head, wearing a yellow crop top and high-waist jeans, holding a smartphone. 3D animated cartoon style, CGI animated feature film quality, NOT photorealistic. Views: FULL BODY FRONT, FULL BODY REAR, FRONT CLOSE-UP, PROFILE CLOSE-UP. Neutral grey background, clean studio lighting, no text or watermarks. High quality, 4K, 16:9 ratio. A 4-view character reference sheet of a fluffy reddish-brown squirrel with a huge bushy tail, big expressive eyes, mischievous grin. 3D animated cartoon style, CGI animated feature film quality, NOT photorealistic. Views: FULL BODY FRONT, FULL BODY REAR, FRONT CLOSE-UP, PROFILE CLOSE-UP. Neutral grey background, clean studio lighting, no text or watermarks. High quality, 4K, 16:9 ratio. A cinematic animated short film with feature-film-quality 3D animation, expressive stylized characters, a scenic golden-hour mountain viewpoint with soft clouds and rolling hills in the background, warm sunset light, fast dynamic cinematic camera cuts with playful zoom-ins, upbeat comedic orchestral score with plucky rhythm, polished CGI animated feature film quality rendering, NOT photorealistic, exaggerated expressive facial animation, kinetic comedic timing. SET LAYOUT: Girl remains fixed center-frame holding her phone up for selfies throughout. Squirrel appears and disappears from a fixed tree branch to her left edge of frame. A stylish 20-year-old girl with long wavy black hair holds up her phone, adjusting her angle, practicing a perfect smile against the scenic mountain backdrop. Girl (to herself, confident): "Okay... perfect lighting..." She snaps the photo. A fluffy reddish-brown squirrel leaps into frame at the last second, pulling a ridiculous cross-eyed face right next to her head. Girl (checking phone, annoyed): "Ugh—again?! She resets, retries her pose with an exaggerated forced smile, clearly getting frustrated. Girl (through gritted teeth): "One good picture. That's all I need." Continue seamlessly from the previous scene. Maintain the exact same characters, clothing, facial features, lighting, mountain environment, and CGI animated feature film quality 3D animation style, NOT photorealistic. SET LAYOUT: Girl remains fixed center-frame. Squirrel's photobombs escalate in speed and silliness, still entering from the same fixed branch area, ending seated calmly on her shoulder for the final shot. She snaps again — this time the squirrel photobombs with an even more exaggerated pose, tongue out, tail puffed dramatically. Girl (exasperated shout): "SERIOUSLY?!" Rapid montage of three quick snap attempts in fast succession, each interrupted by an increasingly ridiculous squirrel face — cross-eyed, mid-jump, cheeks stuffed with an acorn. The girl finally throws her hands up, defeated, sitting down on a rock with a sigh. Girl (muttering): "Fine. You win." The squirrel calmly hops up and settles onto her shoulder, striking a proud little pose. She glances at it, surprised, then can't help but smile genuinely. She holds up the phone one last time — girl and squirrel both grinning naturally at the camera — and snaps the photo. It's clearly the best shot yet. Girl (delighted, laughing): "...Okay, THIS one's perfect." Camera quick zoom into the phone screen showing the adorable photo. Cut to black.

1-10s Create the first half of an original medieval political-fantasy television opening-title sequence. The story concerns royal succession, rival noble houses, a frozen northern frontier, mounted knights, ancient dragons, fire, and winter. This is an opening-title sequence, not a trailer. Use slow ceremonial reveals, tactile materials, controlled camera movement, deliberate pauses, and readable title cards. Use completely original characters, original heraldry, original architecture, and an original dragon design. Visual language: pale winter stone, black iron, weathered parchment, cold blue frost, deep pine green, dark crimson pigment, red sealing wax, rough wool, black volcanic glass, dragon scale, smoke, snow, and restrained orange firelight. Do not use any existing franchise names, characters, places, house symbols, logos, official maps, recognizable costumes, title designs, or musical themes. No invented writing, no runes, no extra text. 0–3.0s: Cold dawn light travels slowly across a frost-covered limestone relief. Shallow carved lines form a map of mountain passes, frozen rivers, raised roads, fortress walls, and a walled city. A thin line of dark red mineral pigment fills one route toward the city. The red line glows like banked fire beneath the frost, then becomes completely still. Reveal only: A SERIES BY MARA VANE The letters are carved into pale limestone, monumental serif typography, flat, centered, correctly spelled, and clearly readable. Hold the credit clearly. 3.0–6.0s: Small pieces of forged black iron slide across dark charcoal wool. A broken lance, a plain knight’s visor, a round shield fragment, and a small dragon-scale shard move separately, then lock into an original circular royal war seal. Two miniature armored riders cross paths once behind the seal, suggesting a medieval battle without becoming a full battle scene. The iron stops moving completely before the credit appears. Reveal only: STARRING CAEL DRAKE The letters are burnished silver on a clean black surface. No crest, no recognizable emblem, and no other text. 6.0–10.0s: An oval slab of black volcanic glass reflects the flicker of a brazier. Frost retreats from the surface and reveals the partial silhouette of an armored knight wearing a heavy wool cloak and a plain steel helm. The face remains hidden. Smoke gathers above the reflection. A huge original dragon wing shadow passes once across the smoke, followed by a brief reflection of orange fire and blue-white frost. Reveal only: CREATED BY ELIAS ROWE The text sits inside the clean dark reflection, perfectly readable and motionless. At the final second, hold the exact final state: the same knight silhouette, the same brazier, the same dragon-wing shadow, the same frost line, the same camera position, and the complete credit still visible. Do not begin the next transition. AUDIO — SOUND EFFECTS ONLY: No music, no melody, no choir, no chanting, no drumbeat, no horn, no bass drone, no orchestral score, and no trailer impacts. Use only cold wind, frost cracking, carved stone friction, red wax dripping, iron sliding on wool, chainmail movement, one restrained metal strike, brazier fire, armored breathing, smoke movement, black glass friction, and one heavy dragon wingbeat. All sounds must have visible physical sources. 11-20 Extend the accepted first clip by one further clip. Use the actual final frame of the first clip as the only continuity source. Do not restart the opening, repeat the previous credits, or redesign the materials. The planned handoff state is the black volcanic-glass reflection with the armored knight, brazier fire, blue-white frost line, dragon-wing shadow, and the fully readable credit: CREATED BY ELIAS ROWE If the actual final frame differs, preserve the actual final frame and its real camera position, lighting, object placement, and motion state. 0–1.2s: Continue the accepted final frame exactly. The credit remains visible for a quiet moment. The brazier flickers once, frost glistens, and the knight remains motionless. No new text. 1.2–2.5s: The credit fades away naturally. The camera slowly pulls back from the black glass. The dark reflection becomes the surface beneath a miniature mountain capital. The same frost, iron, stone, smoke, and firelight continue without a visual reset. 2.5–5.5s: No text. A stone model of the mountain capital is gradually covered by the shadow of an approaching eclipse. Snow sweeps across the outer walls. A line of armored knights advances over a narrow causeway, shields raised. Their horses breathe white vapor. A large original dragon circles above the city. Its shadow covers the knights, while controlled orange fire catches along one section of the fortress wall. The battle remains monumental and restrained: no graphic gore, no dismemberment, no explosions, no modern weapons, and no chaotic camera shake. 5.5–6.5s: The dragon passes through smoke. One dark dragon scale breaks loose, glowing faintly with orange heat on one side and blue frost on the other. It falls into the last remaining strip of light and becomes a brilliant white-gold rimefire sun against black volcanic glass. 6.5–10.0s: Pale limestone and black-iron letterforms rise slowly around the glowing sun and reveal exactly: THE RIMEFIRE CROWN A single warm light passes across the title while a thin blue frost edge remains on the letters. Every object then becomes completely still. Hold the final title until the end. No extra credits, subtitles, logos, runes, invented writing, recognizable franchise symbols, or additional characters. AUDIO — SOUND EFFECTS ONLY: Begin with the exact wind, brazier fire, armored breathing, and glass ambience from the accepted first clip. Do not introduce music or a new audio bed. Use only fabric movement, stone friction, snow blowing across walls, synchronized horse hoofbeats, leather reins, chainmail, shield impact, horse breathing, dragon wing displacement, fortress fire, snow hissing against flame, ice cracking, falling scale impact, limestone scraping, black-iron contact, and the final sound of objects settling into silence. No music, no melody, no rhythm section, no chanting, no choir, no horn, no sustained musical tone, no bass drone, and no trailer-style impact.

Style: 2 seconds of realistic life-feeling mobile phone live shooting → Hard cut for position matching → 8 seconds of static beautiful poster (2.5D parallax + player UI), forming a strong contrast between the two textures. Character: Protagonist @ Image 1 —————— [Shot 1 00:00-00:02 | Realistic Home · Spinning Record] Medium shot, fixed camera, no camera movement. Indoor corridor, background is a creamy-white wall and a dark brown wooden door (hinges clearly visible), dim indoor warm lighting, the picture has realistic mobile phone shooting noise and slight exposure fluctuations. The protagonist wears a white top with black floral print, thin-framed glasses, in a natural home state, standing on the right half of the frame facing the camera, expression calm and restrained, eyes looking straight at the camera, body almost motionless. 00:00-00:00.6 She lifts her left hand towards the left side of her body, her left index finger through the center hole of a black record, holding the record to the **middle-left of the frame** (not in the center, but obviously to the left), the record surface completely faces the camera, like a disc suspended on the left side of the frame. 00:00.6-00:02 She extends her right hand, fingertips touch the edge of the record, her wrist flicks hard—the record immediately spins at high speed around her left index finger. The right hand retracts immediately after the flick, the record continues to spin alone, the surface blurred into circular motion, and the edge reflections into a spinning light arc. The record stays at that position on the left, spinning at high speed, size and position unchanged. 00:02 Hard cut (position matching). At the frame where the record spins fastest, the screen cuts directly—the next frame is a completely different scene, and at the position of the spinning record on the left, it perfectly connects to the rotating vinyl record control in the player UI, with position, size, and rotation direction perfectly aligned. Visually, it looks like the same record is spinning, but the world around it has changed. —————— [Shot 2 00:02-00:10 | Record Cover Poster · 2.5D Parallax] Close-up character shot. This section is not a live video, but a refined static poster given 2.5D parallax dynamics—no physical camera movement, only a very slow pseudo-push-in feel: slight relative displacement between the character layer and background layer, with foreground falling leaves moving fastest, the character following, and the background slowest, creating stereoscopic depth. The visual rhythm is extremely slow and soothing. The right half of the frame features the same protagonist, but transformed into a poster image: exquisite makeup (translucent white skin, cream eyeshadow, upturned eyeliner, full hydrated lips), wearing a diamond-studded thin-strap gown, a string of pearls at the collarbone, long hair in a loose high bun with a few stray strands. She has no physical movement, just smiling slightly at the camera, eyes gentle and slightly melancholic, serene and beautiful, like a frozen magazine cover. Background: Autumn golden ginkgo avenue, sunlight through leaves casting light spots, distant blurred silhouettes of pedestrians and benches, shallow depth of field turning the background into a flowing warm golden glow. [Particle Effects] The entire screen is covered with a golden falling leaf effect, leaves slowly spinning down from the top—leaves close to the lens have obvious depth-of-field blur, passing large and blurry in front of the screen; distant leaves are small and clear, strengthening stereoscopic depth. [UI Layer (fixed on screen, doesn't move with parallax)] - Top left: Golden serif large text 'FASHION', small text 'SELF-PORTRAIT' below, next line year '1901'. - **Middle-left of the frame (matching the record position from the previous shot)**: A 3D vinyl record-shaped player control, rotating slowly clockwise, with her own circular avatar embedded in the center. - Bottom left: Minimalist music player progress bar, with play/pause, previous, and next buttons, the bar slowly advancing over time. - Left side also has light yellow and light gray magazine-style color blocks as layout decoration. - No lyrics or sentimental text appears on the screen. [Overall Tone] Shot 1 is realistic dim indoor warm tone; Shot 2 is rich autumn warm golden tone, strong sunlight, translucent skin, commercial poster quality. [Sound Effects] Shot 1 is realistic indoor ambient sound + slight plastic friction and wind from the spinning record; sounds cut at the moment of the hard cut; Shot 2 cuts directly into full sentimental music (warm texture with vinyl noise), background is very light wind and leaf friction.

P1 [0:00–0:01.25] Use the provided storyboard frame, character reference image, and provided Higgsfield logo reference image. Maintain exact character appearance: pink hair, olive cargo pants, white crop top, green cap, tattoos on arms. Keep the Higgsfield logo and “Higgsfield” text naturally visible throughout the scene as environmental branding. Real photography, low-angle shot, young woman with pink hair and olive cargo pants crouching low, black ink explosion erupting from the ground around her fist impact, high-speed photography, dramatic studio lighting, ink splatter physics, hyper-detailed skin texture, Canon R5, 85mm f/1.4. P2 [0:01.25–0:02.50] Photorealistic, motion blur, athletic woman with pink hair executing a spinning back kick, black ink trails streaking through the air, cinematic lighting, shallow depth of field, gritty concrete environment, sports photography style, Nikon Z9. Maintain exact character appearance and keep Higgsfield branding visible. P3 [0:02.50–0:03.75] Photorealistic low-angle upward shot, woman with pink hair leaping vertically, massive black ink vortex surrounding her, dramatic rim lighting, rising pressure distortion, high-speed shutter, realistic fabric movement, dynamic pose, 24mm wide lens. P4 [0:03.75–0:05.00] Macro photography of a swirling black ink whirlpool on a flat surface, spiral formation, high-contrast black and white tones, studio lighting, ultra-sharp focus, realistic fluid dynamics. Character visible at the edge or reflected in the ink vortex. P5 [0:05.00–0:06.25] Photorealistic dynamic action shot, woman with pink hair dodging low, black ink ribbons whipping around her body, motion blur, hard dramatic lighting, gritty urban backdrop, athletic wear, cinematic color grade. P6 [0:06.25–0:07.50] Real photography, woman with pink hair sprinting toward camera, black ink streaks rushing past her, wind-blown hair, sharp focus on face, motion-blurred background, dramatic backlight silhouette. P7 [0:07.50–0:08.75] Cinematic photorealistic shot, woman with pink hair bursting through a shattering surface, fragmented ink-soaked debris suspended in air, dynamic landing pose, explosive backlighting, dust and ink particles, deep focus, IMAX-style drama. P8 [0:08.75–0:10.00] Photorealistic cinematic medium shot, young woman with pink hair standing upright facing camera with direct eye contact, black ink dripping from both fists, ink-splattered clothes, defiant expression, dramatic backlight, ink mist settling around her feet, shallow depth of field, gritty atmosphere, Canon R5, 85mm f/1.4. Global Instruction: Use the provided Higgsfield logo reference image exactly as supplied. Do not alter the logo design or the “Higgsfield” text. Keep the branding clearly visible throughout every shot and until the final frame.

Panel 01 (00:00 - 00:01) 1980s retro anime style. Wide establishing shot. Static camera. A vibrant retro-futuristic city skyline at sunset. The sky is a gradient of pastel purple

@[Image1] MORTHEA — "Harbinger's Wake" | Opening Titles. 15-second montage, 30 beats @ 0.5s each, ethereal overcranked slow-motion. Art Direction & Vibe: Premium key art. Unified style: high-end dark fantasy anime-realism, sharp graphic linework, cold

Y2K digital nostalgic-style poster, with a background featuring the Windows XP default wallpaper “Bliss”—green grass and blue sky—combined with a stretched pixel grid, and multiple retro system pop-up windows layered across the top and bottom. The windows are stacked out of alignment, with screen blur and pixelated edges. The main visual is a front-facing, medium-shot Y2K cool girl holding a shiny digital-camera-style object, while retro digital devices decorated with cartoon stickers and mobile-phone frames are mixed together and layered over parts of the subject. The Chinese-English main title “FUTURE” is positioned in the lower-left corner, set in 3D chrome pixel-heavy sans-serif lettering with a light-blue gradient outline and glass highlights. The subtitle “by Wenye Bot” appears in small grayish-purple type tiled across the bottom. The overall palette is dominated by blue-green tones, blending watery reflective highlights with clean digital grain. The composition features multiple misaligned scattered layers, conveying a sci-fi-cute, shiny, vivid cyber-subcultural aesthetic, with a Y2K chrome-material filter and enhanced bubble reflections.

DURATION: 6s STYLE Ultra-cinematic dark studio logo intro, photorealistic stone typography realism, premium game studio opening sequence, black void environment, ARRI Alexa 65, anamorphic lens, realistic gravity simulation, impact destruction physics,

Vibe: High-Fashion Editorial meets Cyberpunk Action. The Hook: The viewer thinks they are watching a standard fashion lookbook or K-Pop teaser, but the clothing materials suddenly reveal their true, dangerous purpose. 15-Second Storyboard Timestamp Visual Action Audio / Sound Design 0:00 - 0:03 The Setup: Extreme close-up on the holographic vinyl choker and titanium lock from @image1. The lock suddenly spins like a combination dial and glows deep purple. Upbeat, glossy pop music playing softly. A sudden, sharp, mechanical CLICK. 0:03 - 0:06 The Shift: The camera snaps back. The character flashes her perfect "Bright Smile" directly at the lens. Suddenly, the image glitches heavily. Her expression drops instantly into the cold, deadpan "Vogue Stare." The pop music violently stutters, pitches down, and drops into a heavy, aggressive dark synth-bass. 0:06 - 0:11 The Twist: She spins into her "Dynamic Pose." The unexpected element: The sheer organza sleeves aren't just fabric—they ignite with iridescent light, acting as a kinetic energy shield that deflects an unseen laser blast off-screen. High-frequency hum of energy. A loud CRACK of a deflected laser blast. 0:11 - 0:15 The Climax: Dropping into "Stage Focus," she grabs the heavy chain hanging from her distressed denim cutouts. She yanks it free—it superheats into a glowing plasma whip. She lashes the chain directly at the camera lens. Sound of grinding metal, a roaring crackle of energy, and the sound of shattering glass. Cut to silence.

Create a cinematic anime-style celebration for achieving Top my X weekly views in Japan.

Stop-Motion Chip Stacking — Chips stack themselves into the logo shape or brand initials, then explode apart — playful, highly shareable.

15 seconds, 16:9, flashy transformation scene in high-quality anime movie style. Only one character appearing. @[img1] is a three-view drawing showing the same person from the front, side, and back, not three people. The center front image is used as the starting composition, and the side/back views are used as design consistency references. Maintain the character's face, purple-blue eyes, dark blue hair longer than the waist, hair length, and physique throughout. Accurately reproduce the black glossy futuristic bodysuit and blue glowing lines from Image 1 before the transformation. @[img2] is also treated as a three-view drawing for the same person's costume. After transformation, accurately reproduce the white and silver idol costume from Image 2, featuring blue crystal decorations, an off-the-shoulder white bodice, dark blue pleated skirt, blue glowing trim, transparent iridescent sleeves and tails, a large bow on the back, dark blue long boots, and a blue-black microphone. The setting is the blue-white glowing futuristic lab from @[img3]. Maintain the central cylindrical glass chamber, ceiling and floor glowing rings, left and right control terminals, and symmetrical interior structure. Keep the character in anime style and unify the lab with the same anime movie style light and texture naturally. (Detailed breakdown of transformation from 0.0s to 15.0s follows). The entire video should be a smooth and physically continuous transformation. The person is always one person. Do not generate the three-view diagrams as multiple people. Do not change the face or hairstyle. No sudden costume swaps, extra limbs, deformed fingers, duplicated microphones, changing into another person, distorted background structure, text, subtitles, logos, or watermarks. No dialogue or singing. Synchronize futuristic mechanical drive sounds, energy rising sounds, crystal sounds, and deep bass at transformation completion with the video.

A 15-second character showcase clip presented through 15 different camera angles featuring armor, face, weapons, and body silhouette. It adopts an epic villain-style cinematic 3D animation texture, combined with various camera stability (such as handheld shake, smooth movement, etc.) and camera techniques, focusing on character details. Includes shots of the sword gathering white energy. The environmental wind effects are vivid with motion blur, using focus-shifting techniques. Ends with a cinematic full-body shot of the character. Frame rate is 24fps with a non-smooth 'stepped' effect (drawing every 2 or 3 frames). Includes character entrance, a low growl, and breathing sounds behind a mask; no background music.

Duration: 14 Seconds | Aspect Ratio: 16:9 STYLE: Ultra-photorealistic REAL LIVE-ACTION, AAA Hollywood supernatural assassin action, premium realistic Semi-CGI VFX, ARRI ALEXA 65, IMAX, Panavision anamorphic. Fast controlled choreography, brief micro slow-motion ONLY for phase-dodge. CHARACTER: @Image1 = KAIA. Preserve exact face, hairstyle, body proportions and ORIGINAL CLOTHING. TWIN CURVED DAGGERS. Personality: COLD, FOCUSED, EFFICIENT, UNHURRIED—never frantic. LOCATION: Night atop a snowbound mountain fortress: stone battlements, frost-slick walkways, watchtowers, torch braziers under a full moon. Sentries patrol separately. Fortress stays quiet—NO alarm, crowd, or large battle. 00:00–00:03 — SILENT HUNT Camera already moving, low handheld pursuit behind Kaia as she crosses a frosted walkway. Two sentries patrol ahead along the battlement. She waits, motionless in shadow, until one turns his back. WHOOM— short burst of SWIRLING SNOW-ASH. Kaia's REAL BODY physically accelerates past him, trailing translucent frost-grey afterimages. She stops at blade range behind him. SHK-SHK— TWO precise dagger strikes. She catches and quietly lowers him. The second sentry begins turning—Kaia is already gone. 00:03–00:06 — BLIND SPOT Camera slides around a watchtower as the second sentry scans the dark, breath visible in the cold. Kaia silently emerges from his blind spot, pins his weapon arm— SHK-SHK. TWO precise strikes. She quietly lowers him and looks toward three sentries near the gatehouse, expression unreadable. 00:06–00:08 — PHASE-DODGE Kaia crosses silently behind the gatehouse sentries. One unexpectedly notices and swings at her FROM BEHIND. Camera rushes toward the incoming blade—MICRO SLOW-MOTION. Just before impact, Kaia's physical body visibly DISSOLVES into DRIFTING SNOW-ASH and pale moonlit vapor. The blade passes harmlessly THROUGH her swirling form. TIME SNAP—WHOOSH! The ash sweeps around the attacker as camera performs a curved whip-pan. Kaia visibly REFORMS directly behind him—feet → torso → arms → face → TWIN DAGGERS. He turns too late. CROSS-SLASH. ONE precise counter. Tiny camera impact bump. Kaia catches and silently lowers him. 00:08–00:11 — ASSASSIN CHAIN Two sentries remain. Kaia disappears behind a brazier instead of charging. One passes—snow-ash curls behind him. Kaia emerges: TWO STRIKES, catches him, then slips back into shadow. Final sentry sees a fading translucent afterimage and follows it, sword raised. Camera rotates around him—the REAL Kaia is already in his blind spot. SHK-SHK. ONE controlled exchange. She catches and quietly lowers the FINAL SENTRY. ALL ENEMIES ARE DEFEATED. NONE REMAIN. 00:11–00:14 — SILENT AFTERMATH Absolute quiet. Camera tracks backward along the battlement, slower than before. All defeated sentries lie silently along Kaia's infiltration path. Braziers still burn; gates remain intact; NO alarm. Kaia calmly wipes her TWIN DAGGERS and sheathes them. CLICK. Her eyes shift toward distant torchlight. Snow-ash curls around her feet. WHOOM— Kaia silently bursts into the storm. Camera rushes after her but catches only translucent frost-grey afterimages fading into the blizzard, then holds on empty, drifting snow. END. CORE ASSASSIN LOGIC OBSERVE → BLIND SPOT → SILENT APPROACH → PRECISE STRIKE → CONTROL FALL → DISENGAGE → NEXT TARGET. Never frontal brawling or prolonged blade exchanges. PHASE-DODGE: Incoming attack → micro slow-motion → blade almost connects → Kaia visibly dissolves into snow-ash → attack passes through → ash travels around attacker → Kaia visibly reforms at blind spot → precise counter. Use ONLY when directly attacked; do not spam. VFX / CAMERA / AUDIO Speed VFX: Drifting snow-ash + pale moonlit vapor + translucent frost-grey afterimages. During normal bursts, REAL Kaia always physically leads; VFX trails behind. Phase-dodge requires visible dissolve → ash travel → physical reformation. NO lightning, electrical arcs, or teleportation. Camera: NEVER static. Low pursuit, over-shoulder stalking, watchtower reveals, close reactions, reactive whip-pans, curved tracking. FAST during eliminations, immediately CALM afterward: QUIET → FAST → QUIET → FAST → QUIET. Audio: Wind, distant howling, torch crackle, soft footsteps on snow, fabric movement, subtle ash WHOOSH, dagger draw/impact, controlled body movement. No loud battle music or alarms. NEGATIVE: No frontal mass battle, prolonged blade exchange, running fight, lightning/electricity, teleportation, excessive phase-dodge, explosions/destruction, static camera, robotic movement, air-gap dagger hits, surviving enemies, face/body/clothing/dagger drift, broken anatomy, full-3D/game look, subtitles, logo, watermark.

15-second high-octane Chinese martial arts 2D animated film, mature-oriented Guoman, 2D hand-drawn cel animation texture, realistic adult body proportions, clear outlines, Level 2-3 hard-edged shadows, cinematic lighting and shadows, partial integration of ink splash/streaks, high-speed brushes, smear frames, air blade cuts, and exaggerated yet logically consistent high-speed afterimages. Not Xianxia, not a game skill showcase, the core is a realistic martial arts chase under extreme speed differences. Setting: 'Duanyun Temple' at dusk just after heavy rain, a massive abandoned ancient temple nestled among mountains and cliffs. Features a three-story wooden main hall, giant dark brown wooden pillars, dark blue-black wet tiles, half-collapsed corridors, stone courtyards, a bell tower, broken railings, a stone bridge over a cliff, ancient pines, and tattered Buddhist banners. The western dark clouds split to reveal an orange-red sunset, with warm golden slanted light illuminating the cold bluish-gray temple. Wet stone slabs and tiles reflect characters and special effects. The air is filled with mist, dust, falling leaves, and broken tiles. Character A: 'Bai Jin', a white-haired thread-suspending guest, adult male, slender and lean, long silvery-white hair, black narrow-sleeved cross-collar martial arts attire, dark black short robe, dark red waist sash, black leg-binding pants with light martial arts boots. Small amount of dark purple vein-like patterns on the side of the neck and chest. Combat ability: extreme high-speed Qinggong 'Purple Lightning Wandering Dragon Steps', vein-severing palm techniques, and heavy leg strikes 'Cloud-Severing Mountain-Cleaver'. Purple effects represent high-speed movement only, not energy attacks. Character B: 'Shen Yan', a black-clad bladesman, adult male, messy mid-length black hair, ink black and dark gray narrow-sleeved martial arts outfit, gray-black short cloak, a long saber in a deep black scabbard at his waist. Skilled in hidden blade techniques and judgment-based tile-treading footwork, but significantly slower than Bai Jin. He remains suppressed throughout the first half. 0.0—1.0s: Fight starts immediately. 20mm ultra-wide low-angle shot, camera almost touching the wet bluestone ground. Bai Jin occupies the massive foreground, one black boot appearing huge due to perspective, body lowered almost to the ground. Shen Yan stands in the courtyard in front of the main hall dozens of meters away. Bai Jin suddenly explodes off the ground; the stone slab beneath him cracks like a spiderweb, standing water splashes backward in a fan-shaped curtain, his hair and clothing pull straight back. The camera follows closely, with wooden pillars, stone steps, banners, and eaves blurring into high-speed stripes along the direction of motion. Bai Jin enters 'Purple Lightning Wandering Dragon Steps', forming a continuous high-speed trail with a white-purple highlight core, deep purple body, magenta edges, and minimal cyan-green dispersion. 1.0—2.5s: Bai Jin cuts into the courtyard at high speed but doesn't attack immediately. He first flashes past Shen Yan's left, leaving a black human afterimage, a white air trail, and water cut by the airflow; an instant later, he circles a giant wooden pillar from the right rear. Shen Yan quickly turns his head to find the target, hand gripping the hilt but not swinging wildly. The camera zooms into Shen Yan's eye; the pupil darts left, then snaps right. In the next frame, Bai Jin has already cut in from the side-rear. Bai Jin uses two joined knuckles to quickly brush aside Shen Yan's raised scabbard, followed immediately by a palm strike exploding from a very short distance, slamming heavily into the side of Shen Yan's face. The contact must be clear: the cheek compresses first, the head is knocked aside, hair and rain spin in the other direction, a few drops of blood fly out, and the shoulders and waist follow. The arm can have a very brief 2D smear stretch, but the actual arm does not lengthen. 2.5—4.0s: Shen Yan is sent flying dozens of meters laterally by the palm strike, body tumbling irregularly. His feet first brush against a stone lamp, shattering the top half, then he crashes through the decayed wooden railing and wall of the main hall. A massive amount of wood chips, tiles, and dust scatter. The camera cuts to an ultra-wide shot to show the displacement. Shen Yan's shoulders and back hit the wooden floor of the hall first, his body continuing to roll, the long saber still gripped in his hand, the scabbard scraping the floor creating continuous wood shavings. Finally, Shen Yan kneels on one knee, one hand supporting the ground, not yet fully stable. 4.0—5.8s: A purple-white high-speed light trail flashes outside the main hall's wooden wall. Bai Jin bursts through the wood chips into the hall, first flashing past Shen Yan's front, creating two purple human afterimages, then a third afterimage appearing behind Shen Yan. The real Bai Jin finally stops steadily to the side. As Shen Yan turns, Bai Jin has already stepped forward, his supporting foot crushing the floorboards, his knee lifting quickly, waist and hips fully rotating, his leg extending horizontally to execute 'Cloud-Severing Mountain-Cleaver'. His sole strikes Shen Yan's chest and abdomen heavily. At the moment of contact, Shen Yan's torso folds, the sword hilt is pressed into his body, and then his feet leave the ground entirely. 5.8—7.1s: Shen Yan is blown out from the other side of the hall by the side kick, the decayed wall exploding into a giant hole. He flies through a storm of wood pieces and broken tiles toward the next roof level. Shen Yan's feet hit the wet blue tiles first, which shatter consecutively as his body slides down the slanted roof at high speed. He immediately stabs the scabbard into the tile gaps to slow down, carving a long mark across the roof. Bai Jin doesn't pause, entering 'Purple Lightning Wandering Dragon Steps' again, shooting out from the hall's opening, a purple-white high-speed trail streaking across the flying eaves. 7.1—9.0s: High-density 3D pursuit. Bai Jin sprints along the eaves, circles behind an ancient pine, then runs briefly horizontally along the side wall of the bell tower, using the force to change direction. The purple light band forms a continuous S-shaped path between temple buildings, accompanied by black humanoid afterimages and white ink-style air brushes. Shen Yan leaps from a roof toward another corridor. Bai Jin cuts in early from the ridge above, first slamming his shoulder into Shen Yan's sword arm to deflect his posture, then immediately spinning his body to deliver a lateral elbow strike to the side of Shen Yan's face. The elbow strike frame features a strong 2D smear, with hair, cuffs, and forearm briefly stretched. Shen Yan is knocked off the roof again. 9.0—10.0s: Shen Yan falls from a height of two stories. Bai Jin chases down from above as an almost vertical purple high-speed trail. Bai Jin is not firing energy but descending at high speed himself. He passes Shen Yan mid-air, presses his palm against Shen Yan's back, and uses his own descending momentum to slam him downward. Shen Yan's trajectory is forced vertically down, crashing face-first into the next courtyard. The wet stone slabs shatter instantly, with white-gray dust, water splashes, and rubble exploding simultaneously, forming a tall smoke pillar in the center and a ring of spreading water curtains. 10.0—11.8s: Before the dust clears, Shen Yan forces himself out, running along a massive colonnade. Bai Jin follows in 'Purple Lightning Wandering Dragon Steps'. High-speed horizontal tracking shot; the two weave between dozens of giant dark brown wooden pillars, characters briefly obscured by the columns. The purple trail must realistically circle the pillars and never pass through them. Bai Jin kicks off a pillar to turn, wood chips exploding from the surface as he launches in the opposite direction. Shen Yan uses his scabbard to vault over a wooden railing, landing on an outer stone terrace. As Bai Jin chases out, the high-speed wind pressure sweeps up a whole row of banners. 11.8—13.2s: Shen Yan judges that the next strike cannot be fully dodged. He turns immediately, keeping his long saber sheathed, gripping both the hilt and scabbard with both hands, holding the entire sword horizontally across his chest, lowering his center of gravity in a front-back bow stance to form 'River-Blocking Guard'. In the distance, Bai Jin charges along the stone terrace, exiting the purple high-speed state in the final meters, pulling both palms to his waist before thrusting forward simultaneously in the final step with 'Mountain-Shaking Palm'. Both palms slam into Shen Yan's horizontal scabbard. 13.2—14.1s: The collision must be depicted in layers. In the first instant, Shen Yan successfully blocks the attack; the scabbard bends inward, his arms are pressed back to his chest by the immense force, and his feet slide back along the wet stone slabs, soles kicking up water splashes as the stones begin to crack. But Bai Jin's force continues to surge in; in the next instant, Shen Yan's defense completely collapses, and he is blown away along with his sword. The camera follows Shen Yan with a 30-45 degree camera roll, with eaves, sunset, cliffs, and mountains rotating and tilting together. 14.1—15.0s: Shen Yan flies toward an ancient stone bridge connecting two cliff walls. One foot touches the wet bridge surface first, followed by the second. His body continues to slide back at high speed, water on the bridge spraying to both sides. Shen Yan almost falls off the edge of the bridge before finally stabbing the scabbard into a stone crack; rubble bursts out, and his body finally forces a stop. The final shot is static: Shen Yan kneels on one knee in the center of the cliff stone bridge, head slightly lowered, a trace of blood at the corner of his mouth, shoulders heaving, one hand still firmly gripping the hilt. In the distance, on the roof of Duanyun Temple, Bai Jin stands quietly. The sunset stretches a vast distance between the two. The scene ends as Shen Yan just begins to look up, connecting directly to the next segment. Special Effects Fixed Rules: Purple only represents Bai Jin's high-speed Qinggong; white-purple center, deep purple body, magenta outer edge, minimal green dispersion. High speed can produce 3-5 humanoid afterimages, but there is only one real person. Attacks are primarily actual fist, palm, and leg techniques; effects only show speed and impact. All knockbacks must involve real contact first, then body deformation, then displacement. No random teleporting, no static standing, no turn-based combat, no purple turning into lasers, no characters passing through pillars or walls, no characters flying away prematurely, no instant recovery to normal after landing.

Fixed use of [Reference Image] as the sole protagonist and bear reference for 30 seconds. For Seedance 2.5, 16:9, 30 seconds, 60fps. Generate the entire video as a single continuous ultra-high-speed AAA fantasy cinematic. This video will not use the 'continuous fall from high altitude to underground' structure of the previous work; the setting, direction of movement, creatures, and visual development are entirely changed. The core is 'a 30-second sequence where the protagonist and bear run, fly, switch rides, and traverse a massive otherworldly city'. The protagonist and bear always move along a single line of motion; standard edit cuts, blackouts, fades, sudden location changes, portals, or warp holes are prohibited. The camera doesn't just follow beside the protagonist; it rapidly moves through frontal intercepts, rear tracking, over-the-head shots, ground-level skims, under the wings of giant creatures, between the protagonist and bear, and inside narrow architecture, almost never staying in the same position for 30 seconds. [Most Important: AAA transformation of Reference Image] Do not replace the protagonist in the [Reference Image] with a realistic human or an ordinary beautiful girl; rebuild her as a highest-quality AAA 3D character with the specific 'one-eyed monstrous character' traits of the reference image. Treat her as an adult female, but do not correct her to a human face. Completely maintain the giant single eye in the center of the face, white eyeball, purple-blue-red iris, black pupil, and the characteristic smile showing very large teeth. Do not increase the number of eyes to two. Do not make her face human. Extract the red to deep wine-colored round helmet, black rim, small black circular decorations on the helmet, black bodysuit/jacket, white skeletal motifs, red cord-like decorations, voluminous black lower outfit, and dark navy to black shoes from the reference image. Do not erase the watercolor-like unevenness of red, blue, purple, green, and black; keep them as hand-painted dyes, rubbed pigments, and fine color bleeds on the AAA material surface. Do not make the skin overly realistic human skin; process it as a high-quality stylized AAA character with a sense of real material. Give the surface fine roughness, paint unevenness, sewing, metal, rubber, cloth, leather, and fine scratches. Do not leave the black ink outlines of the reference image as pure 2D lines, but translate them 3D-style into black borders on the outfit, thick helmet edges, and shadow contour designs. Do not make it look like cheap game CG. Use Unreal Engine-grade AAA cinematic materials, feature-length animated film-level quality, GI, reflections, volumetric light, and lens expressions. [Bear Fixed] Maintain the pink-to-purple bear held by the protagonist in the [Reference Image] as a second character alongside the protagonist for the entire 30 seconds. It is not just a stuffed toy, but a small living bear that talks to the protagonist, runs on its own, flies on its own, and moves at the same speed as the protagonist. Maintain the round ears, black eyes, black nose, white-to-cream muzzle, body with watercolor-like mixes of pink, purple, blue, green, and pale yellow, seams, and soft round limbs from the reference image. In the AAA version, it's a mix of a luxury plush toy and a fantasy creature. Express short soft fibers, sewing threads, slightly rubbed fabric, soft compressive deformation, and the inertia of ears and limbs. The bear is not a copy or a pet of the protagonist, but a slightly cheeky sidekick. It has a different voice from the protagonist. Maintain the size difference between the protagonist and the bear; do not enlarge the bear. It should be a size that the protagonist can hold under one arm when needed. [Stage] The starting point is a massive 'floating megacity' at dusk. Not a typical office district, but a post-futuristic wasteland city with massive rust-red steel frames, glass elevated corridors, magnetic levitation trains running through the air, buildings built upside down, giant windmills, and only the frames of floating billboards remaining, with high-rise towers piercing the clouds and a giant ring-shaped urban structure rotating in the distance. Colors are sunset orange, deep ultramarine, reddish-purple, and cold cyan. Excessive neon is prohibited. Everything is highest-quality AAA 3D cinematic. In the second half, move continuously from the city to a sea of clouds, then an aerial waterway, then a giant glass forest, and finally a floating night market; do not switch abruptly between any of these. Transitions occur through physical continuity of architecture, weather, creatures, and terrain. [Basic Camera Design] Use 18-24mm wide-angle as the base for high-speed movement, naturally closing in to 28-40mm for dialogue and expressions. 60fps. The camera does not continuously track parallel like a drone. Change height, distance, angle, and speed every 0.4 to 1.2 seconds. Create a sequence of 3D movements like: backing away in front of the protagonist and bear -> passing grazing the protagonist's shoulder -> diving under the bear -> moving ahead of the giant creature -> circling to the front -> backing away at ground level -> passing through an architectural hole first. Meaningless 360-degree rolls are prohibited. However, actively use banks of 15-70 degrees to match the turns of giant creatures, gravity direction changes, or the curves of architecture. Add natural directional motion blur only to high-speed movement sections, while keeping the protagonist's single eye, smile, and the bear's expressions readable. [0.00-3.20s | Suddenly sprinting on the roof of an aerial train] Intense speed from the first frame. The protagonist and bear are sprinting at full speed on the roof of a black magnetic levitation train traveling at hundreds of km/h through the floating city at dusk. The protagonist is in front, the bear desperately runs 1-2m behind on short legs. The protagonist's red cords, black outfit, and helmet decorations sway in the wind pressure; the bear's ears and arms stream back violently. The starting camera is about 4m ahead of the train, 20mm lens skimming the roof while backing away to capture the two from the front. At 0.5s, the camera dives steeply to the protagonist's feet, with the metal seams of the train roof flowing violently under the lens. At 0.9s, it rises sharply from outside the protagonist's right leg, approaching the giant single eye and toothy smile at 24mm. While running, the protagonist looks back and shouts happily to the bear behind, 'You're slow!' The bear replies, 'Think about my leg length!' while out of breath. Perfect lip sync. At 1.5s, broken rails and a massive aerial gap appear ahead. Without slowing down, the protagonist says, 'Okay, let's jump!' The bear says, 'I didn't hear about this!' At 2.3s, the protagonist kicks off the train roof, and the bear jumps immediately after. The camera passes between the two at high speed, then continues into the air ahead to turn around and catch the protagonist and bear head-on as they approach from behind. [3.20-6.30s | Jumping onto a giant white crow] Dozens of black birds fly in the city sky, but a giant silver-white crow with a wingspan of over 15m descends rapidly from behind them. Do not summon it suddenly. Have it exist as a small white shadow in the distance from the 2-second mark; by 3.5s, its wings, beak, silver feathers, and bluish-black eyes become clear. The protagonist and bear do not fly freely but fall by gravity after the jump. The giant crow crosses diagonally below them. The protagonist twists her body 90 degrees in mid-air and makes a dynamic landing on the crow's back on one knee at 4.2s, grabbing a feather with one hand to absorb the impact. The bear misses the landing and falls toward the crow's wingtip, shouting a small 'Waaah!', but the protagonist immediately reaches out one hand to grab the bear's arm and pulls it up like a pendulum. The camera glides under the crow's wing, looking up from below at the protagonist pulling the bear up. At 5.0s, the bear rolls onto the back saying, 'Hold me from the start!', and the protagonist laughs loudly, 'That wouldn't be any fun!' The crow beats its wings once strongly and dives between the city buildings. The camera gets ahead of the head and backs away, putting the silver crow, protagonist, bear, and the city flowing at high speed behind them into a single frame. [6.30-9.30s | Ultra-high-speed breakthrough through high-rise city interior by white crow] The white crow doesn't cruise horizontally but meanders between high-rise buildings using a combination of descents and ascents. At 6.5s, it passes through two beams of a giant glass tower with its wings folded. The protagonist crouches low, pulling the bear to her chest. The camera flies into the narrow gap ahead of the crow, an 18mm ultra-wide shot with glass and metal closing in to about 10cm on each side. Immediately after, the crow bursts violently toward the lens, and the camera accelerates backward to avoid it. At 7.2s, the crow redeploys its wings and makes a sharp left turn. The camera slides over the top of the right wing and looks forward over the protagonist's shoulder. Ahead is a massive transparent 'aerial waterway' flowing through the center of the city. Water is held in the air like a giant ribbon, meandering between buildings. The white crow descends toward the waterway. The protagonist says, 'Eh, we're going in there?' The bear, with a pale expression, says, 'It's definitely the face of someone going in!' The crow plunges into the transparent water flow. [9.30-12.40s | Plunging into the aerial waterway -> Switching to a giant celestial carp] As the crow enters the water, do not flash the screen; physically continue the surface refraction, bubbles, water pressure, and water adhering to the feathers. Maintain AAA quality. The crow's feathers get wet and heavy, causing speed to drop sharply. The protagonist and bear are thrown forward by inertia. The protagonist lightly strokes the crow's neck once, saying 'Thanks!' while swimming forward in the water current. The bear doesn't swim but drifts toward the protagonist's side, spinning around due to its round body's buoyancy. 'I can't swim!' The protagonist says, 'You're floating, so it's fine!' At 10.4s, a massive semi-transparent celestial carp approaches at high speed along the current from downstream. 12m long, milky white and pale cyan, with transparent fins, thin golden scales, and a soft light inside. The carp doesn't appear magically but has been swimming in the distance of the waterway from the start. The protagonist flips her body in the water and reaches for the carp's back. At 11.2s, as the carp passes under her, she grabs the base of the dorsal fin and pulls herself in. The bear lands softly on top of the carp's head. The camera runs backward from grazing the carp's mouth, showing the giant eye, the bear on top of the head, and the protagonist moving to the back all at once. At 12.0s, the carp jumps out of the aerial waterway. Water droplets do not remain for seconds but fall down according to gravity. [12.40-15.60s | Flying celestial carp + Sea of clouds jump] The giant carp jumping out of the waterway does not fly freely like a bird but glides for a short time using the momentum from the waterway exit and the lift of its giant fins. The protagonist and bear head toward the sea of clouds at the edge of the city while riding its back. The camera dives directly under the carp, showing the protagonist and bear above through the transparent belly with refraction. At 13.0s, the camera passes behind the carp's tail and rises to the upper rear. A giant glass forest begins to appear ahead. Massive transparent trees hundreds of meters tall, branches branching like prisms, and wind blowing at high speed between them. The carp's flight speed gradually decreases. The protagonist points ahead, 'Next, over there!' Bear: 'Next what!?' Protagonist: 'I haven't thought about it!' Bear: 'I knew it!' At 14.3s, the carp approaches the upper part of the glass forest, but the gaps between branches are too narrow for the giant carp to enter. The protagonist strokes the carp's back once and, holding the bear under one arm, jumps to one of the branches. The carp naturally falls into another waterway below and disengages. The protagonist's shoes contact the transparent branch and begin to slide with a glass sound. [15.60-18.60s | High-speed sliding through glass forest + Bear starts self-flight] The protagonist slides at high speed on the massive transparent branches like skating, holding the bear in her left arm. The branches are not straight but curve downward sharply, and the protagonist accelerates with gravity. The camera moves at high speed upside down on the underside of the branch, capturing the soles of the protagonist's feet and the bear through the transparent glass, then circles to the side of the branch. At 16.3s, the branch forks. The protagonist leans right to make a high-speed turn. The bear says, 'Wait a sec, I feel like I can fly too!' The protagonist: 'Now!?' The bear jumps out of the protagonist's arms on its own. Instead of two small cloth wings popping open from its back, the seams on the sides of its body stretch out as if unraveling, becoming soft wings that were originally built-in. The wings maintain the color tones and materials of the reference image. The bear starts flying awkwardly next to the protagonist, 'I flew!' Immediately after, it almost hits a transparent branch, 'Whoa, dangerous!' The protagonist laughs and crouches low to jump to the next branch. The camera bursts into a narrow hole between branches ahead of the protagonist and follows the jumping protagonist and flying bear from the front without changing the 180-degree orientation. [18.60-21.80s | Jumping onto a giant horned rabbit's back and sprinting on the ground] Below the glass forest, amber-colored grasslands and herds of massive migrating beasts are visible. The transparent branches of the forest lower their altitude, finally continuing near the ground like a giant curved slide. The protagonist accelerates on the branch, with the bear flying beside her. Below and ahead, a giant horned rabbit about 8m long is running through the grassland at high speed. Long white-to-light-gray fur, silver-black antlers branching like a deer's, massive hind legs, and blue eyes. Do not show it suddenly; introduce it as a distant herd from the 17-second mark. At 19.2s, the protagonist jumps from the end of the transparent branch and aligns herself in the air with the horned rabbit's back. The camera looks straight back from between the horned rabbit's ears, capturing the approaching protagonist from above. At 19.7s, the protagonist lands on the back, absorbing the impact with both knees. The bear also flies in from the side and lands as if bumping into the protagonist's shoulder. The horned rabbit does not slow down. The protagonist grabs the fur with one hand and strokes the side of its neck once with the other. The horned rabbit reacts by tilting its ear slightly toward the protagonist. At 20.4s, Protagonist: 'Fast, fast!' Bear: 'You were just saying you wanted to go faster!' Protagonist: 'This is too fast!' The horned rabbit kicks off a giant rock with its hind legs and jumps over it; the camera leads grazing the ground to give the impact of the giant feet passing over the lens. [21.80-24.70s | Entering the floating night market, protagonist and bear running together] Past the grassland, a massive floating night market city appears continuously, detached from the ground. First, warm-colored lights in the distance, then roofs, cloth, suspension bridges, and three-dimensional alleys become discernible. The horned rabbit runs up a giant inclined bridge and enters the market's periphery. The market is an AAA-class otherworldly city. Wet stone, wood, brass, cloth awnings, floating lanterns, steaming stalls, and giant mechanical clocks. Do not generate text or readable signs. The protagonist jumps from the horned rabbit's back to a nearby roof at 22.2s. The bear follows with self-flight. The horned rabbit runs off into an alley. The protagonist sprints at full speed on the roofs, with the bear flying alongside at shoulder height. The camera starts from the protagonist's left side, accelerates ahead of her at 22.7s, and stays turned 180 degrees while backing away to capture her face and the bear. It then dives steeply under a narrow suspension bridge, looks up at the two through the gaps in the bridge, and immediately rises back to the rooftops. At 23.5s, the protagonist steps onto a giant cloth roof of a stall, and the bounce of the sinking cloth sends her leaping to the next roof. The bear asks from the side, 'Hey, where's our destination anyway?' The protagonist looks at the bear for a moment while running. 'Destination?' A 0.2-second pause. '...We don't have one?' The bear looks at the protagonist silently. [24.70-27.50s | Running up a giant clock tower and jumping into the sky together] A giant tilted clock tower ahead. Giant gears are exposed on the outer wall from the market roof. The protagonist jumps from the roof edge to the rotating gears of the clock tower and runs upward using the gear teeth as footholds. The bear flies beside her. The camera tilts nearly 90 degrees against the clock tower wall, following the protagonist from below while maintaining the world coordinate gravity. At 25.2s, the camera passes through a hole in a rotating gear first, and the protagonist jumps through the same hole. At 25.6s, the camera moves from behind the protagonist over her head to the front, approaching the giant single eye and smile at 28mm. Protagonist: 'Okay, last jump!' Bear: 'How many times today!?' At 26.0s, the protagonist jumps onto the hand at the very top of the clock tower and runs along the rotating giant minute hand. The bear flies beside her. At 26.6s, the moment the tip of the minute hand reaches its highest point, the protagonist uses the recoil to make a giant jump. The bear also beats its wings strongly to rise at the same time. The camera leads ahead into the sky above the tower, composing a shot of the two jumping toward the camera with the entire city in the background. [27.50-30.00s | Comical mid-air pose together] The protagonist and bear rise for a short time due to the inertia from jumping off the clock tower, then naturally transition to a fall. In the background are the night floating market, distant glass forest, giant ring city, and the sky changing from dusk to night. The protagonist half-rotates her body in the air to face the camera. The bear flies next to her, trying to mimic the protagonist's pose. The protagonist strikes a bold and comical pose reminiscent of the reference image, with her right hand spread wide, her left hand on her hip, and one leg bent. The bear also raises one arm but forgets to operate its wings and starts falling slightly downward. The protagonist doesn't notice and looks at the camera with a supremely proud face, saying 'Perfect! ...Wait, where's the bear?' Immediately after, the bear shouts from below the screen while falling, 'It's not perfect!' The protagonist's giant single eye moves downward, noticing for the first time that the bear is falling. At 29.2s, the protagonist laughs saying 'No way!' and dives down to grab the bear's leg with one hand. They return to the front of the camera hanging upside down, the bear with a grumpy face and the protagonist laughing loudly showing her teeth. At 29.6s, the protagonist gives a V-sign with one hand while dangling the bear with the other. Protagonist: '...Did it work out?' Bear: 'How!?' Perfectly synchronized with the final 'How!?', the protagonist laughs even harder. The camera rapidly approaches the two, filling the screen with the giant eye, the protagonist's smile, and the angry bear. The city behind flows rapidly upward, maintaining that the two are still falling. At 30.00s, the video ends on a clear final frame of the comical moment where the protagonist is holding the bear with one hand and both are still moving in the air. Black fades, blackouts, or static images are prohibited. [Dialogue/Voices] The protagonist is an adult female. A bit low but bright, mischievous, and not too high-pitched even when excited. The bear has a slightly higher-pitched voice characteristic of small characters but not a toddler voice. Do not confuse the voice qualities. Do not add lines not specified. Do not repeat the same line multiple times. Perfect synchronization with mouth movements. Do not stop the protagonist or bear for dialogue during high-speed action. Lines: 'You're slow!', 'Think about my leg length!', 'Okay, let's jump!', 'I didn't hear about this!', 'Hold me from the start!', 'That wouldn't be any fun!', 'Eh, we're going in there?', 'It's definitely the face of someone going in!', 'Thanks!', 'I can't swim!', 'You're floating, so it's fine!', 'Next, over there!', 'Next what!?', 'I haven't thought about it!', 'I knew it!', 'Wait a sec, I feel like I can fly too!', 'Now!?', 'I flew!', 'Whoa, dangerous!', 'Fast, fast!', 'You were just saying you wanted to go faster!', 'This is too fast!', 'Hey, where's our destination anyway?', 'Destination? ...We don't have one?', 'Okay, last jump!', 'How many times today!?', 'Perfect! ...Wait, where's the bear?', 'It's not perfect!', 'No way!', '...Did it work out?', 'How!?'. [Music/SFX] Music is high-speed quirky cinematic electro x orchestral breakbeat. Around 135-145 BPM. Low end is tight, with light drums, short brass, string staccato, comical woodwind accents, clean sub, digital percussion, and short synths. Heavy EDM drops are prohibited. Include metal vibrations and strong wind on the train roof, air pressure during jumps, giant wing flaps and wind cutting for the white crow, water pressure, bubbles, and low resonance in the aerial waterway, water film and giant tail vibrations for the celestial carp, high resonance sounds of transparent branches in the glass forest, giant footsteps, fur rubbing, and ground vibrations for the horned rabbit, and environmental sounds of cloth, wood, metal, and distant crowds in the floating market, along with metal gear meshing in the clock tower. Do not unnaturally stop the music during dialogue; lower it naturally by 1-2dB to bring voices forward. Remove the music for just 0.15s before the final 'How!?' and add a short comical percussion accent after the line. Leave the laughter naturally. [Absolute Priorities] Even when transforming the protagonist from the [Reference Image] into AAA, maintain the single eye, giant smile, helmet, black skeletal motif outfit, red cords, and colors. Do not change her into an ordinary woman with two eyes. Also maintain the bear's design from the [Reference Image] as the same individual sidekick character throughout. Do not replace the protagonist or bear with different characters during the 30 seconds. The entire video must be unified in highest-quality AAA cinematic style, without changing textures to live-action, 2D animation, or watercolor halfway through. The appeal this time is not texture changes, but extraordinary camera movement and location changes.

A cinematic, action-packed CGI sci-fi sequence set in a dusty desert outpost surrounded by large metal industrial scaffolding towers under a hazy, foggy sky. Two male martial arts fighters are locked in combat, one wielding glowing fiery orange energy shields and blast rings. Standing between them is a powerful East Asian female cyber-warrior in high-tech black and gold metallic body armor alongside a large, glowing blue-eyed mechanical spider mech. The camera cuts to an intense close-up of the female warrior's face, showcasing her long black hair, luminous white glowing rings around her pupils, and glowing orange circuitry veins running down her neck and collarbone. She charges forward through the sand, leaps high into the air, and transforms into a fiery orange energy shockwave ring as she flies across the sky. A small blue-lit robotic spider burrows into the desert sand, and she crashes down directly onto it with a massive fiery explosion impact. As the flames and dust clear, she rises into a dramatic superhero crouch in a sand crater, wind blowing through her hair. Hyper-realistic 8k resolution, photorealistic visual effects, high-contrast cinematic color grading, motion blur, volumetric fog, dynamic action movie aesthetic, shot on 35mm lens, 9:16 vertical ratio.

Create a 27-second dark fantasy cinematic battle sequence set in a destroyed futuristic city during an apocalyptic storm. Scene 1 — Establishing shot (0–4 sec): Begin with a wide aerial cinematic shot of a devastated futuristic city. Massive skyscrapers are damaged and burning, smoke covers the streets, dark storm clouds fill the sky, and flaming debris falls from above. The atmosphere should feel hopeless, epic, and dangerous. Use dramatic volumetric lighting, realistic fire, thick smoke, flying ash, and subtle camera shake. Scene 2 — Warrior reveal (4–8 sec): Cut to a powerful female warrior standing in the middle of the destruction. She wears highly detailed dark metallic fantasy armor with sharp sculpted elements and glowing blue energy flowing around her body. Her eyes glow intensely. Lightning strikes behind her as she raises her arms and channels enormous electrical energy. Use a slow cinematic push-in toward her face, dramatic backlighting, wind moving her hair and armor details. Scene 3 — Enemy attack (8–13 sec): Introduce a huge monstrous armored enemy emerging through the smoke. The warrior charges toward it while several shadowy creatures move through the ruined battlefield. Show fast, dynamic sword combat with powerful impacts, sparks, flying debris, and realistic weight. Use low-angle shots, tracking shots, and brief slow-motion moments during major attacks. Scene 4 — Lightning clash (13–19 sec): The warrior releases a massive blue-white lightning attack toward the enemy. The enemy blocks with its enormous weapon, creating a spectacular energy collision. Lightning spreads across the ground, illuminating the battlefield. The camera circles around the fighters while explosions, dust, smoke, and debris fill the background. Scene 5 — Final confrontation (19–24 sec): Move into intense close-up combat. The warrior dodges a devastating strike and counters with a glowing energy blade. The enemy becomes surrounded by fiery orange energy while the warrior remains surrounded by cold blue lightning, creating a strong visual contrast. Scene 6 — Final hero shot (24–27 sec): End with the warrior standing alone on cracked ground after the battle. She slowly raises her glowing weapon as golden-orange energy burns around her. The destroyed city and storm remain in the background. Finish with a dramatic low-angle hero shot and a slow camera pullback. Visual style: photorealistic dark fantasy, AAA video-game cinematic, high-detail armor, realistic human movement, cinematic destruction, volumetric fog, dramatic storm lighting, realistic fire and smoke, blue lightning effects, orange energy, dynamic camera movement, realistic physics, shallow depth of field, anamorphic cinematic look, high contrast, epic scale, ultra-detailed VFX. Motion: smooth character animation, realistic combat choreography, controlled camera shake, fast trac

Desert chase action dual pistol prompt High-quality anime video. 3D toon/cel-look chase action. The setting is unified as a bright desert, and the protagonist holds a total of two pistols, one in each hand. Maintain character identity, hoverboard chase on sand, a giant white snake dragon, direction of movement, camera significance, and a conclusion where the chase continues after firing. [Reference Image Roles and Priorities] image1 is for face and identity reference. Prioritize face shape, eyes, iris, eyelashes, cheeks, jaw, nose, mouth, bangs, and hair around the face. image2 is for full body and outfit reference. Refer to overall hair, twin tails, flower decorations, outfit structure, gloves, shoes, physique, full body silhouette, and unique color scheme. image1 and image2 represent the same single protagonist. In case of conflict, prioritize image1 for facial features and bangs, and image2 for overall hair, outfit, and silhouette. Do not copy indoor backgrounds, text, logos, or UI from reference images. [Protagonist Identity Fixing] A soft, small oval face, pale skin, large irises glowing from purple to lavender, delicate dark upper eyelashes, small nose and mouth, and focused eyes. Hair is long twin tails ranging from pale coral pink to milky pink. Maintain two purple flower ornaments, purple tassels, purple earrings, and small black ear studs in the same positions. Outfit is a black/lavender/purple street-wear idol/hacker style. A short lavender jacket, black top, chest straps, patches, belt and harness, black-purple layered pleated skirt, black fingerless gloves, and black-purple platform high-cut boots. [Art Style: 3D Toon/Cel-look] High-density 3D toon/cel-look. Thin dark navy to blue-gray outlines, two to three layers of cel shading, soft mid-shadows, and transparent environmental light. Multilayered highlights on hair and eyes. Distinct textures for sand, rocks, scales, metal, cloth, and leather. [Desert, Enemy, and Equipment Fixing] The stage includes golden dunes, red sandstone canyon walls, dry rifts, and crumbling ledges under a blue sky. The enemy is a single giant white snake dragon with segmented body, white scales, cyan light lines, and red eyes. The board is one hoverboard owned by the protagonist. Weapons are exactly two compact black-purple handguns held simultaneously. No weapon duplication or transformation. [Common Motion Agreement] Connect sand grounding, jumps, air movement, aiming, simultaneous firing, recoil, and subsequent sliding into one continuous trajectory. Use sand dust and the dragon's body as foreground wipes. Cut 1 | From extreme close-up of purple eyes to desert chase Cut 2 | Side chase on dunes Cut 3 | Transition to wall-running on sandstone canyon Cut 4 | Dual pistol shot under the dragon's belly Cut 5 | Return to sand dunes and sandstone jump Cut 6 | Overlooking the dry rift and rocky ledge Cut 7 | Close-range evasion of the dragon's mouth Cut 8 | Parallel chase with dragon's head Cut 9 | Face and dual pistols, chase continues with hit reaction

Hyper-realistic cinematic action sequence, 15 seconds, 16:9. A futuristic interceptor races through a canyon carved into the surface of Saturn's moon Titan. Towering orange ice cliffs rise thousands of meters overhead while dense atmospheric haze creates dramatic shafts of light. The route naturally narrows into twisting rock corridors, giant natural arches, frozen rivers, and immense caverns illuminated by glowing methane ice. The aircraft skims the canyon walls at incredible speed, passes beneath collapsing ice bridges, and exits through an enormous vertical opening into Titan's endless orange sky. The camera alternates between cockpit POV, ultra-low tracking, FPV chase, wide aerial reveals, and breathtaking flybys emphasizing the scale of the alien landscape. Style: IMAX realism, believable alien geology, cinematic atmosphere, crystal-clear action, uninterrupted forward momentum.

Seedance | 15-second Chinese Xianxia hardcore dual spears vs. sword short film. Generate a complete and continuous 15-second Chinese Xianxia hardcore combat cinematic short film. I. Overall Requirements: Cinematic realism, pure ancient-style Chinese Xianxia hardcore combat aesthetics, based on real ancient weapon distance logic as the skeleton, then magnified for Xianxia. Actions must be fierce, fast, heavy, and sudden. Use high-aggression impact editing, weapon-occluded cuts, fast pans, low-angle rushes, sudden stops, and clear acceleration and deceleration. Completely prohibit soft, dance-like choreographed moves. II. Core Action Principles: Real action pleasure comes from: Real weapon distance differences, clear and readable spatial geometry, explosive starts and stops, clear weight, inertia, and impact, body imbalance and regaining the axis, pressure formed by weapon containment lines, absolute control of the final strike stopping before the fatal distance. Explicitly prohibit: Soft actions, slow display of moves, meaningless posing, random sword energy, teleportation, vague light pollution, dance-like matching of moves, subtitles. III. Reference and Environment Setting: Character identity anchors: @Image 1 -> White-clothed Sword Immortal Senior Sister, @Image 2 -> Green-clothed Junior Sister. Environment DNA: All uploaded background and location reference images determine the same environment DNA. Before formal composition, silently reconstruct compatible: Real terrain, architectural language, spatial scale, texture, vegetation, water bodies, mountain mist, main light direction, air depth. And from the spatial relationships already implied in the reference images, find the most suitable real location for combat, such as: Narrow passages, bridgeheads, corridors, rock wall aisles, or equivalent combat axes. Never forcibly create a new location that conflicts with the reference images for the sake of action. IV. Environment Movement Rules: The background remains consistently vivid but absolutely neutral in narration. Continuous environment movements include: Wind, water, mountain mist, vegetation, clothing hems, reflections, distant figures. Requirements: All independent, natural, and continuous movement, maintain real parallax, do not actively create plot, do not actively help either side, do not actively change the combat outcome. Only when characters truly step on, collide with, or sweep past, is the local environment allowed to produce passive physical feedback, such as: Dust rising, water ripples, debris, leaf swinging. V. Character Settings: Character ID A | White-clothed Sword Immortal Senior Sister | @Image 1: Always maintain the same identity and appearance: 25-30 year old East Asian female, original face, original hairstyle, white embroidered silk Hanfu, silver waist seal, jade pendant, white cloth boots, a single silver straight sword. Performance and action requirements: Calm, precise, restrained, extremely fierce at the moment of explosion, all actions are efficient with no redundant moves. Character ID B | Green-clothed Junior Sister | @Image 2: Always maintain the same identity and appearance: Original identity, green linen Hanfu, wooden hairpin, black cloth shoes, a single dark steel sword. In this segment: Stands in the back with the same elderly master as a witness, showing instinctive emotional reactions midway, but does not intervene in the main combat. Supporting character: Enemy master, a light-armored enemy master, holding a short spear in each hand, actions must be extremely aggressive, possessing strong continuous suppression capability at short to medium range. Elderly Master: Always witnessing from the back, does not intervene in combat, only says one judgment at a critical moment. VI. Shot Structure: 0-5s | Wide Shot / Long Shot | Extreme static pressure -> Violent explosion. Camera Structure: 0.0-1.0s: 24mm static wide shot; after 1.0s: Hard cut to low-angle 28mm high-speed rushing shot. Content: The same white-clothed sword immortal senior sister confronts the same light-armored dual-spear enemy general. The same green-clothed junior sister and the same elderly master stand in the back as witnesses. 0.0-1.0s: Use 24mm static wide shot. Requirements: Both sides completely still, use absolute silence to create pressure, combat space, distance relationships, retreat paths, and pressure axes must be clearly established. 1.0s Explosion: The enemy suddenly explodes without warning. Immediately hard cut to low-angle 28mm high-speed rushing shot: Two spear tips stab directly toward the camera, camera pans quickly, the same white-clothed sword immortal only sidesteps at the last moment to dodge the spear point. The silver sword and the first short spear collide violently, the metallic bang instantly cuts to a side angle, the second short spear immediately hits her scabbard hard, real force pushes her back two full steps sideways, then stops instantly. Leave only half a beat: Dust continues to move, silk clothing continues to swing, hair continues to be moved by the wind. Segment focus: The beginning must be extremely quiet, the explosion must be sudden, the first round of conflict must make the audience immediately feel the fierce pressure of the dual spears. 5-10s | Medium Shot / Cowboy Shot | High-pressure continuous attack -> Disarmed near-death situation. Camera and editing logic: Maintain the same white-clothed sword immortal, the same dual-spear enemy, and the exact same geographic space. This segment completes 4 aggressive cuts, triggered only by: Weapon impacts, sudden changes in movement direction, body imbalance, characters entering the frame, weapons blocking the screen. Never use average mechanical cuts. Content: The first spear shaft sweeps across the screen, forming a natural occlusion cut. Cut to 35mm close-up handheld follow-shot, the enemy completes three segments of high-pressure offense: straight thrust, spear tail counter-hit, low sweep. The same white-clothed sword immortal's reaction: Parry the first strike, extreme body evasion to dodge the second, but the third low sweep actually grazes her ankle, she clearly loses balance for the first time. Immediately hard cut to 70mm close-up: The same silver sword is violently knocked out of her hand by the enemy's other short spear. Never use slow motion to admire the flying sword. Immediately cut back to 24mm low-angle: The enemy's two short spears have crossed to seal her chest, forming a true dead end. At this point: The same junior sister instinctively takes a step forward, the same elderly master only reaches out to grab her sleeve, whispering only one word: 'Watch.' Segment focus: The rhythm must get fiercer, a visual crisis where the sword immortal seems to be suppressed to death truly appears, the junior sister and master give minimal reaction, not stealing the focus. 10-15s | Close-up / Extreme Close-up | Violent close-in entry -> Regaining the sword momentum. Content: The dead end continues, no breathing room. The same white-clothed sword immortal does not retreat, but suddenly rushes close against the two short spears. The sequence of actions must be clear: Use the empty scabbard to wedge between the two safe spear shafts, forcing the dual spears into a close-range dead corner where distance advantage cannot be fully utilized, then use a shoulder to violently knock the enemy's body axis off balance, spinning to the outside along his forearm, using the enemy's already committed forward inertia to carry him past the camera. The enemy's spear shaft instantly fills the screen, forming a weapon-occluded hard cut. The next moment: The same silver sword that flew out earlier falls at high speed from the top of the screen. The same white-clothed sword immortal does not look back from start to finish, her right hand accurately grabs the sword hilt behind her back. Immediately cut to 85mm impact close-up: Wrist explosive correction, silver sword edge suddenly stops at a distance of one finger from the enemy's throat. All previous violent sounds suddenly disappear simultaneously. Closing shot: Extreme close-up of the enemy's shrinking pupils, then cut to the same white-clothed sword immortal's completely calm eyes. The same elderly master in the distance only says: 'Long weapons fear closeness, and the same goes for dual spears.' Only then does the same junior sister truly exhale the breath she was holding. Segment focus: The ending action must be fierce, short, and precise. It's not a flashy display but a victory decided by a close-in momentum-seizing strike. The final 'sudden stop' must form a strong memory point. VII. Editing and Camera Requirements: The entire segment is still 0-5s, 5-10s, 10-15s, three clear narrative segments, but approximately 10-12 clear cuts must be completed internally. Cut rules: All cut points can only be triggered by: Weapon impacts, weapon sweep occlusions, sudden changes in movement direction, body imbalance, characters entering the frame, weapons filling the screen. Explicitly prohibit: Average mechanical cuts, groundless random cuts, soft smooth transitions, excessive slow motion display. VIII. Audio Requirements: Original synchronized Mandarin dialogue, enhanced metallic bangs, enhanced short spear wind-breaking sounds, enhanced sole-stopping friction sounds, enhanced fabric swinging sounds, enhanced real breathing. The first half's action sounds must be violent and clear, the final sword-stop moment should form a clear sound-evacuation feeling. IX. Motion Texture: 24fps cinematic motion texture, natural motion blur, actions have clear weight, inertia, impact, and sudden stops. Strictly prohibit: Teleportation, floating, softness, dragging, slow move displays, illogical move demonstrations. X. Total Output Requirements: Total duration: 15 seconds, Aspect ratio: 16:9 landscape, Cinematic realism, three clear narrative segments, approximately 10-12 high-aggression clear cuts completed internally, original synchronized Mandarin dialogue, consistent character identity, clothing, weapons, and geographic relationships throughout, no subtitles generated. XI. Negative: blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, glitching cuts, disappearing props, soft dance-like combat, slow ornamental moves, random sword aura, random energy effects, teleporting movement, unreadable weapon trajectories, floating body mechanics, fake impacts, weak inertia, artificial slow motion glamor shots, modern elements

A 15-second high-quality 2D fantasy anime. Using the young wizard in the attached image as the protagonist, strictly maintain his messy black hair, round glasses, blue-gray eyes, dark blue and light blue magic robe, leather belt, magic book, staff, and short stature in every cut. Delicate line work, watercolor-style coloring, fantastical blue and gold light, theatrical anime quality. [0-4 seconds] An ancient observatory protruding into a sea of clouds before dawn. In a strong wind, the boy opens a large magic book with both hands. The camera approaches quickly from a low angle. The pages turn violently, and the written constellations glow bluish-white. Hair and robe flutter in the wind. [4-10 seconds] Countless star particles erupt from the magic book, spiraling around the boy. The camera makes a large half-circle around the boy. The star particles transform in the air into a giant translucent star dragon, spreading its wings. The boy looks up at the sky in surprise through his glasses. [10-15 seconds] The giant star dragon flies from above the boy toward the sea of clouds, splitting the clouds and turning the night sky into a sunrise. A close-up of the boy's face. The departing dragon is reflected in his eyes, and finally, the boy smiles slightly. A grand and moving lingering feeling. Smooth animation, natural body movements, flawless fingers, no text or subtitles.

I. Generation Goal Generate a complete, continuous 15-second Chinese Xianxia martial arts movie short film. Overall Style Requirements: - Cinematic realistic quality - Pure traditional Chinese Xianxia aesthetic - Mature and high-level martial arts action design - Arri Alexa cinematic texture - Clear and stable facial micro-details - Natural volumetric lighting - Delicate film grain - 24fps cinematic motion quality - Natural motion blur II. Core Action Philosophy Process real martial arts body principles for cinema and Xianxia: - Flexible movement - Stable footing - Diagonal dodging of weapon tips - Strict distance control - Feigned retreats to lure the enemy - Seizing attack lines - Rapid offense-defense reversals Visual Appeal Points: - Clear and readable spatial geometry - Dangerous timing differences (time gaps) - Distance gaps between long and short weapons - Realistic silk fabric movement - Metal clashing impact - A memorable finishing move Prohibited for 'coolness': - Excessive meaningless sword aura/beams - Overblown light pollution/effects - Teleportation - Random special effects - Illogical energy attacks III. Reference and Environment Settings Character Identity Anchors: - @Image 1 → Character ID A | Sword Immortal Senior Sister - @Image 2 → Character ID B | Junior Sister Environment DNA: All backgrounds and location reference images uploaded in this round jointly determine a unified environment DNA. Before composition, analyze and merge compatible elements: Real terrain, architectural language, spatial scale, material aging, vegetation, water bodies, weather, mountain mist, cloud movement, primary light direction, reflection relationships, atmospheric depth, foreground/midground/background layers, character movement paths, and logical combat routes, then recombine them into a unique, unified, plausible, physically reasonable, and spatially continuous new space. IV. Environmental Movement Principles Background must remain vivid but narratively neutral. Continuous natural movements include: Wind, water, mountain mist, vegetation, fabric, distant figures, dust, reflections. Requirements: - Movements follow real physical laws - Maintain natural parallax - The environment does not create danger or assist any side - Only when characters physically interact with the environment (step, sweep, collide) are normal physical feedbacks allowed (e.g., dust rising, water ripples). V. Character Settings Character ID A | Senior Sister | @Image 1 Consistently a 25–30 year old East Asian female: Oval face, fair natural skin, dark almond eyes, black long hair partially pinned with white jade, tall and slender. Wears a white embroidered silk Hanfu, silver belt, and jade pendant. Carries a single silver straight sword. Performance: Very steady, calm, restrained, precise reactions without unnecessary flair. Character ID B | Junior Sister | @Image 2 Consistently a 20–25 year old East Asian female: Round, agile face, braided black hair, petite. Wears a green linen Hanfu and dark belt. Observing the battle from a distance with an elderly master. Supporting Characters: - Enemy Spearman: Wearing dark charcoal gray martial arts attire, holding a spear with dark red tassels. Professional and explosive movements with a long-weapon advantage. - Elderly Master: Observing from a distance, non-intervening, providing a single judgment at the end. VI. Shot Structure 0–5s | Wide or Long Shot | Establishing Battle Space and Distance Advantage Camera: ~24mm, slow lateral tracking. Content: Establish positions and weapon distance advantage. Senior Sister faces the Spearman while the Junior Sister and Master observe. The Spearman thrusts suddenly; the Senior Sister performs a narrow diagonal step, parrying the spearhead with the side of her sword and returning to stillness instantly. Focus on judgment and distance control. 5–10s | Medium or Cowboy Shot | Three-stage Attack and 'Disarmed' Reversal Camera: ~40mm, stable tracking. Content: Spearman performs a three-part attack (thrust, sweep, tail-strike). Sister retreats half a step, leaving a gap. Enemy commits to a lunge; she dodges and guides the spear away with a circular sword motion, closing the distance. The enemy spins the spear, knocking her sword into the air. Visual reversal: 'The sword immortal is disarmed.' Junior sister reacts with shock; Master is steady. 10–15s | Close-up or ECU | Close-quarter Blind Spot Conclusion Camera: ~75mm. Content: Enemy pursues for the kill. Sister steps into the spear's blind spot, controlling the spear shaft with her forearm and scabbard. She spins past the enemy as her falling silver sword descends behind her. She catches the sword behind her back without looking and stops the blade an inch from the enemy's throat. Ending: Close-up on enemy's pupil, focus shift to Sister's calm eyes and hair in the wind. Master: 'She captured the sword, but not her momentum.' VII. Sound and Motion Requirements - Sound: Mandarin dialogue, precise metal clashing, wind, footsteps, fabric rustle, breathing. Music must not drown out sound effects. - Motion: Clear weapon trajectories, realistic body weight/inertia, natural motion blur, visual continuity, stable weapon assets. No teleporting, random aura, or subtitles. VIII. Total Output Requirements - Duration: 15s - Aspect Ratio: 16:9 - Three continuous clear shots - Optimized for Seedance 2.0 Fast - Consistent identity and spatial logic IX. Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, glitching cuts, disappearing props, random sword aura, excessive energy effects, overexposed light pollution, teleporting movement, unreadable weapon trajectories, fake combat, weak body mechanics, exaggerated anime action, modern elements

Seedance | 15s Wuxia Relationship Short Film | True Authorization I. Core Goal: Generate a continuous 15-second Chinese Xianxia film short. Overall Style Requirements: Cinematic realistic texture, pure ancient Chinese Xianxia aesthetics, mature martial arts power dynamics, restrained emotional tension, sophisticated character blocking, Arri Alexa cinematic quality, clear and stable facial micro-details, natural volumetric lighting, fine film grain. The core twist centers on: 'True Authorization.' Everyone initially assumes the more famous Senior Sister sword immortal is the one making decisions. As the plot progresses, the Senior Sister gradually transfers the frame center, negotiating rights, and final decision-making power to the Junior Sister. The true emotional anchor is: Under the public eye, the Senior Sister restrains the urge to answer for her. True support is allowing someone to have their own voice. II. Reference and World Setting Character Identity Anchors @Image 1 -> Character ID A | Senior Sister Sword Immortal @Image 2 -> Character ID B | Junior Sister Environment DNA All background and location reference images uploaded determine a set of environment DNA. Before formal composition, silently deduce and integrate compatible: real terrain, architectural language, spatial scale, materials, vegetation, water bodies, weather, cloud layers, mountain mist movement, main light direction, reflection relations, overall color grading, atmospheric depth, and realistic movement paths into a unique, unified, physically logical, stable, and continuous new space. III. Environmental Operation Principles The background must remain vivid but absolutely neutral in narrative. Continuous environmental movements: water flow, drifting mountain mist, vegetation responding to natural wind, slowly changing clouds, distant character activity, reflection changes, and spatial ambient sound. Requirements: Naturally continuous throughout, maintaining realistic physical logic, no active conflict creation, no resolving conflict for characters, no triggering of dramatic turns. IV. Character Settings Character ID A | Senior Sister Sword Immortal | @Image 1 Consistently the same 25-30 year old East Asian female: oval fair face, dark almond eyes, long black hair half-pinned with a white jade hairpin, tall and slender, wearing the same white embroidered silk Hanfu from white cloth boots to semi-transparent wide sleeves, silver waistband, jade pendant, a single silver longsword. Performance: Restrained, calm, stable, not scene-stealing, truly handing over decision power. Character ID B | Junior Sister | @Image 2 Consistently the same 20-25 year old East Asian female: rounded lively face, black hair in braids, petite, wearing the same turquoise linen Hanfu, dark belt, wooden hairpin, black cloth shoes, a single dark steel sword. Performance: First half has inertia, subconsciously waits for confirmation; second half enters true decision-maker state; growth via breath, gaze, posture, and tone. Supporting Characters: Enemy swordsman (standing opposite, carries bias, initially only acknowledges senior, later forced to turn to junior); Elderly Master and two disciples (silent witnesses, master nods at the end). V. Shot Structure 0-5s | Wide/Long Shot | Bypassed by the world. In the redesigned space: Enemy stands opposite. Master and disciples are silent witnesses. Senior and Junior Sister stand together. Enemy looks only at Senior: 'I only talk to you, let her step down.' Senior calmly replies: 'You've got the wrong person.' Key focus: Establish default power structure; enemy's gaze on sister; junior bypassed; sister's first response starts rewriting power direction. 5-10s | Medium/Cowboy Shot | Yielding center and voice. Maintain same characters, costumes, swords, and geography. Senior Sister steps back half a step, leaving frame center to Junior Sister. She says: 'The leader today is her.' Enemy disbelieves: 'Her?' Sister: 'Even I listen to her.' Enemy's smile fades. Key focus: Half-step back is actual transfer of power; sister acknowledges junior's leadership. 10-15s | Close-up/Extreme Close-up | Leaving decision to her. Enemy turns to Junior: 'Conditions?' Junior steadies breath, answers: 'Retreat three miles, leave weapons, take the wounded.' Enemy looks at Senior for confirmation. Senior: 'Don't look at me.' Junior looks back instinctively. Senior softly: 'You are leading.' Pause. Junior turns back, straightens up, says: 'So be it.' Extreme Close-up: Senior Sister stays half step behind, slight smile of relief. Background: Master nods. Key focus: Junior completes decision loop; sister's restraint is key; master is witness. VI. Performance Focus Senior Sister: Extremely stable emotion, natural authority, brilliance in restraint, movements change relationships. Junior Sister: Starts with passive inertia, then takes center, then independent voice; readable growth. Enemy: Initially bypasses junior, then wavers, finally faces the negotiator. VII. Technical Requirements 15s, 16:9, three continuous shots, native Mandarin dialogue, precise lip-sync, clear pauses, consistency in identity/costume/props/space, restrained performance, natural physics for silk/hair, realistic parallax, neutral vivid background, no subtitles. Camera: Stable, restrained, heavy cinematic feel, serving relationships, no showing off. VIII. Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, glitching cuts, disappearing props, exaggerated melodrama, overacting, cartoon performance, environment triggering story, environment solving conflict, random magical event, fake emotional climax, static background, frozen distant figures, fake parallax, modern elements

Seedance | 15s Deadpan Comedy Short | Sect Administrative Procedures I. Project Goal Generate a complete, continuous 15-second Chinese Xianxia movie short. Overall Style Requirements: - Cinematic realistic texture - Pure ancient Chinese Xianxia aesthetics - The solemnity of an epic martial arts duel colliding with the absurd comedy of a "deadpan sect administrative process" - Restrained silent-film style visual reactions - Dry, absurd logic - "Three-beat progression" structure - Precise comedic pauses - Arri Alexa cinematic look - Clear and stable facial micro-details - Fine film grain - Natural volumetric light The plot must be immediately understandable to the audience upon the first viewing: A powerful challenger arrives full of murderous intent, expecting a legendary duel; however, two masters (Characters A and B) treat "accepting a challenge" as a tedious daily administrative task. They use increasingly logical yet absurd procedures to slowly wear down the enemy's heroic momentum. II. Reference & World Setting Character Identity Anchors: - @Image 1 → Character ID A | Sword Immortal Elder Sister - @Image 2 → Character ID B | Junior Sister Environment DNA: All backgrounds and location reference images uploaded this round jointly determine the same set of Environment DNA. Before formal composition, silently integrate and deduce compatible: real terrain, architectural language, spatial scale, textures, vegetation, water bodies, weather, mountain mist, main light direction, reflections, atmospheric depth, and realistic walking paths. Reorganize these into a unique, unified, physically logical, and stable continuous new space for this round. Environmental Principles: The background is always natural and vivid but absolutely neutral in narration. Continuous natural environmental movements include wind, water, vegetation, mist, banners, distant ordinary disciples, reflections, and spatial ambient sound. These movements only provide a sense of the "world's real existence" and must not trigger laughs, change the plot, or make decisions for the characters. III. Character Settings Character ID A | Sword Immortal Elder Sister | @Image 1: - Same 25–30 year old East Asian female throughout - Tall and slender - Oval face, dark almond eyes, fair natural skin - Long black hair half-tied with a white jade hairpin - Wears a white embroidered silk Hanfu with a silver waistband and white cloth boots - Carries a unique silver longsword - Performance Core: Extremely restrained, deadpan, always acting as if handling a normal sect matter. Comedy comes from her serious execution of absurd procedures. Character ID B | Junior Sister | @Image 2: - Same 20–25 year old East Asian female throughout - Petite, round and agile face, black braided hair - Wears a turquoise linen Hanfu with a dark belt and black cloth shoes - Carries a unique dark steel sword - Extra Prop: Ordinary bamboo slips with no readable text - Performance Core: Acts extremely naturally, treating the challenge as routine registration. Serious tone, professional attitude. Comedy comes from her complete lack of awareness that she is being ridiculous. Enemy Swordsman: - Same adult East Asian male throughout - Imposing, murderous intent, strides into the scene, carries the same longsword - Initially believes he is about to start a legendary duel - Performance Core: Not a clown, but a very serious challenger whose momentum is gradually flattened by administrative procedures. His emotional change is a key readable clue. Elderly Master: - Same elderly master throughout, stable position - The authoritative supplement in the procedural logic - Calm tone, responsible for the final "finishing blow" punchline. IV. 15-Second Three-Shot Storyboard 0–5s | Wide or Long Shot | Challenge Begins, Procedure Interruption: In the open area based on the reference images: The enemy swordsman strides in, draws his sword, and shouts with epic intensity: "Today, either you fall, or I fall!" Character A is just about to reach for her sword hilt. At this moment, Character B naturally walks between them with bamboo slips and asks deadpan: "Name, sect, weapon. Please report them first." The enemy freezes. Key point: The enemy is in full combat mode while the Junior Sister acts like she's at a registration desk. The contrast must be clear immediately. Comedy relies on logical mismatch, not overacting. 5–10s | Medium or Cowboy Shot | Procedure Escalates, Enemy Deflates: Maintain character, clothing, sword, and spatial consistency. Character B continues asking seriously while counting on her fingers: "Any old injuries?" "Who pays if things get broken?" The enemy's momentum is broken, and he shouts in frustration: "I'm here for a duel!" Character A replies expressionlessly: "I know, that's why I'm asking." The camera shifts focus to the Elderly Master a few steps away, who says without looking up: "The last one didn't register and cost me two doors." The camera returns to the enemy, who looks at the master, the bamboo slips, and his own drawn sword; his murderous aura clearly begins to deflate. Key point: The administrative logic becomes more "reasonable" but more absurd, grinding down the enemy's heroic narrative. 10–15s | Close-up or Extreme Close-up | Witness Rule, Final Punchline: The enemy takes a deep breath, suppressing his anger, and asks: "Can we fight after I fill this out?" Character B replies seriously: "We still need a witness." Both sisters turn to look at the Master. The Master says calmly: "Today is a day off." A full beat of silence. The enemy slowly lowers his sword and asks: "Then should I come back tomorrow?" Character A replies with zero hesitation: "Tomorrow she is off." Character B nods solemnly. The enemy stares at them and finally asks: "Do you just not want to fight at all?" Extreme close-up: A tiny, tacit look between the sisters, then they turn back and say in unison, completely serious: "You noticed?" In the background, the Master hides a slight smile behind a teacup. The enemy stands there, lost all will to fight, and slowly sheathes his sword. He lost the whole duel without a single strike. Precise cut to black. V. Performance & Comedic Rhythm Requirements Comedy Mechanism: A clear three-beat progression (High-pressure start -> Procedure erodes momentum -> Confirmation they don't want to fight). Performance: Restrained, realistic reactions, no funny faces or slapstick. Comedy comes from deadpan delivery, dry logic, natural pauses, and emotional gaps. Pauses: Must be clearly preserved after: "Name, sect, weapon", "I'm here for a duel", "Cost me two doors", "Still need a witness", "Today is a day off", "Do you just not want to fight?", and "You noticed?". VI. Camera & Sound Requirements - Duration: 15s strictly - 16:9 Landscape - Three continuous clear shots - Native synced Mandarin dialogue with precise lip-sync - Consistent character identities, props, and geography - Realistic silk fabric and hair physics - Natural parallax between foreground and background - Realistic spatial ambient sound, no generated subtitles - Camera Principle: Restrained, stable, realistic cinematic inertia, serving only the narrative and comedy, no flashy movements. VII. Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, glitching cuts, disappearing props, exaggerated slapstick, cartoon comedy, overacting, forced meme expression, background triggering punchline, environment affecting story, wind reacting to dialogue, random magical event, fake parallax, frozen background, static scenery, modern elements

[PROSE & NARRATIVE ACTION LOCK]\nA 23yo Chinese-Indonesian woman, 165cm, French-bob with blue streaks, sits at an SCBD desk typing on a keyboard (0.0s-1.2s). Urgent pop-ups flood screen; she exhales, sipping iced latte as her watch flashes red alert (1.2s-3.0s). At 3.0s, [HARD CUT: SLAM SHOT], she slams the cup, kicks chair back, flicks off glasses (3.0s-4.8s). At 4.8s, [HARD CUT: DYNAMIC SPIN], she flings her blazer up, popping gum with a smirk as cyan-magenta bio-electricity crackles across fingers (4.8s-7.0s). At 7.0s, [HARD CUT: GROUND IMPACT], sneakers stomp the floor with an ionic wave; the airborne blazer morphs into a bio-neon jacket she catches and zips (7.0s-9.4s). At 9.4s, [HARD CUT: VISOR DEPLOY & ROOFTOP SHIFT], cyan visor locks on and copper bracers snap shut as she leaps through balcony glass into Jakarta sunset sky (9.4s-11.0s). At 11.0s, [HARD CUT: MONUMENTAL WIDE], she flips mid-air, fires a magenta energy whip, swings over Sudirman, and lands 3-point on a helipad at 12.8s kicking up glowing dust (11.0s-13.4s). At 13.4s, [HARD CUT: TIGHT 75mm CHOKER CLOSE-UP], she blows a pink gum bubble that pops, speaking strictly from 13.4s to 14.2s in fluent Indonesian Gen Z cadence: \"Deadline jam lima? Gue yang tentuin.\" From 14.2s to 15.0s, [SETTLE STANCE HOLD], standing upright, hands on hips, living chest breathing, 0% frozen frame.\n\n[ACTING, MICRO-EXPRESSIONS & BIOLOGICAL REALISM]\nAnya exhibits somatic realism: eyelid fatigue shifting to pupil dilation (AU5), laryngeal swallow, asymmetric Duchenne smirk (AU12+AU6), masseter jaw pulse upon landing, specular iris reflections of neon arcs, natural skin micro-pores with satin finish, 0% tears, 0% plastic smoothing.\n\n[CAMERA RIG, OPTICS & MOTION BLUR]\nPanavision DXL2 Large Format, Primo 70mm and 40mm Primes, 90-degree fast shutter on impact to 180-degree natural blur, snap-push zooms, tight 75mm prime on dialogue, Light Iron Color 3, cyan-magenta chiaroscuro against dusk, operator sway.

Generate a 15-second horizontal 16:9 original martial arts action anime video from the provided first frame. CRITICAL ENTITY LOCK: There must be exactly 2 main characters in the entire video: KANAN (male acrobat in yellow wraps with energy staff) and ZEPHYR (female scavenger in off-white techwear with gold visor). Do not add secondary scavengers, droids, or background onlookers. Maintain strict visual consistency for Kanan's cyan staff rings, yellow wraps, goggles, and Zephyr's gold visor, off-white suit, and amber palm gauntlets throughout all cuts. Character identity: KANAN: Male dune acrobat, yellow wraps, dark vest, protective goggles, cyan energy staff, fluid martial arts movement. ZEPHYR: Female sun scavenger, off-white techwear suit, gold heat-reflective visor, amber palm energy gauntlets, fast strike martial arts style. Video style: High-budget Japanese anime feature film quality, post-apocalyptic martial arts aesthetic, high-contrast sun-drenched lighting, bright lens flares, fluid hand-to-hand animation, continuous real-time velocity. Camera and pacing: Relentless real-time martial arts action without any slow-motion pauses: 0.0s - 5.0s: High-speed tracking shot as fighters sprint across sand dunes and trade acrobatic staff deflections amidst blowing sand. 5.0s - 10.0s: Low-angle arc camera following fast martial arts maneuvers, sand sweeps, and palm strike deflections across rusted satellite ruins. 10.0s - 15.0s: High-velocity energy shockwave impact at full speed, transitioning into a clean sliding recovery and a balanced standoff freeze. Action timing: 0.0s - 2.5s: KANAN and ZEPHYR charge downhill across sand dunes. KANAN vaults off a buried satellite rim, spinning his cyan energy staff into a sand vortex. 2.5s - 5.5s: ZEPHYR leaps through the sand wave, using her amber gauntlets to deflect three fast staff strikes. She kicks hot sand upward to create a defensive distraction. 5.5s - 8.5s: KANAN sweeps his staff along the ground to displace the sand under ZEPHYR's feet. ZEPHYR vaults over the sweep and launches a flying palm strike. 8.5s - 11.5s: FULL-SPEED KINETIC IMPACT: ZEPHYR's palm gauntlet collides directly with KANAN's staff guard at full velocity. A bright shockwave of cyan and amber energy blasts sand 360 degrees outward. 11.5s - 15.0s: KANAN redirects the kinetic blast into a staff counter-pulse. ZEPHYR slides backward across the dune on her boots, coming to an instant halt in a ready stance. Final cinematic freeze frame. Motion quality: Fluid 2D animation, rea

The camera is already flying through a sky filled with living mountains slowly migrating across the atmosphere of an alien world. Entire ecosystems exist on their backs while roots of stone and crystal trail behind them through the clouds. Two flight craft weave between these moving giants, diving through valleys and tunnels carved into the mountains themselves. The camera tracks tightly as colossal peaks drift past the lens. A transparent station attached to one of the mountains overlooks the migration. The craft climb above the herd to reveal thousands of living mountains stretching across the horizon like a moving continent.

A frustrated young male creator sits alone in a dark creative studio, staring at a completely blank computer monitor. He rests his hand on his forehead, looking stuck and out of ideas. The room is moody and cinematic, with soft monitor glow and deep shadows. The blank screen suddenly begins transforming into a vivid cinematic world. The camera smoothly pushes toward the monitor and transitions seamlessly inside it, revealing an enchanting fantasy forest filled with enormous ancient trees, glowing purple-pink foliage, exotic plants, soft mist, and a crystal-clear stream reflecting warm rays of sunlight. Magical particles float gently through the air as the camera slowly travels forward through the forest. The scene then dramatically transitions into a sleek futuristic black supercar speeding through a winding mountain road at dusk. The car accelerates aggressively around the curves, tires producing subtle smoke and sparks, glowing red taillights reflecting across the wet asphalt. Massive mountains surround the road with a bright full moon in the background. Ultra-cinematic commercial look, photorealistic details, dynamic camera movement, smooth transitions, volumetric lighting, atmospheric fog, realistic reflections, shallow depth of field, dramatic contrast, premium VFX, realistic motion blur, 4K quality. End with a powerful tracking shot behind the supercar as it disappears into the mountain road. Duration: 15 seconds. Aspect ratio: 16:9. No text, no logos, no watermark.

Use [Reference Image] as the sole protagonist character reference. For Seedance 2.5, 16:9, 30 seconds, 60fps. Generate the entire video as a single continuous ultra-high-speed cinematic one-cut. Conventional editing cuts, scene transitions via blackouts, fades, or sudden location changes are prohibited. Every change during the 30 seconds must have a physical reason for moving from the previous scene to the next. Connect the character's gaze, body rotation, running, jumping, hand movements, ghost summoning, ghost flight, camera orbits, entry into the ghost's mouth, free falling, web shooting, wall attachment, tension, swinging, and mid-air rotations into a single motion trajectory. The tempo is extremely fast, with visual accents occurring every 0.3 to 1.0 seconds, but never as a high-speed edit of multiple unrelated shots. The key is that the entire 30 seconds are physically connected despite being ultra-fast. The protagonist, ghosts, or camera must always be in motion from start to finish. [Most Important Concept] The vivid anime girl from [Reference Image] exists directly within a AAA-grade photorealistic live-action near-future city. Only the protagonist is not photorealistic. Maintain the original 2D anime/high-end digital illustration texture of [Reference Image] for 30 seconds, while depicting only the city, architecture, sky, glass, metal, roads, light, shadows, reflections, fog, steam, distant views, air, web lines, and environmental physics as thorough photorealistic live-action. Create visual beauty through the heterogeneous coexistence where a character looks like an anime figure but receives real live-action city light, runs in real space, and flies from real buildings via webs. Keep the protagonist's outline sharp, but do not make her look like a sticker pasted on the background. Reflect directional light from the environment, floor reflections, color bleeding from buildings, and shadows from obstacles onto the anime shading. Establish contact shadows on live-action floors, color and silhouette reflections on glass, and natural occlusion by foreground objects. [Protagonist Consistency] Completely maintain the adult female character from [Reference Image]. Keep the bright blue hair, bangs, side hair, tied-back hair, large red ribbon, red and white hat, white pom-poms, heart-shaped sunglasses with purple-to-pink lenses, small earrings, white fur, black mini dress, red nails, and rings. Do not change the face, eyes, hairstyle, hair color, body proportions, outfit, or accessories midway. Maintain the sharp, high-quality 2D cel-shaded anime + digital illustration texture with clear linework, blue/red/white colors, and smooth cel shading. However, do not make her a flat 2D image; maintain 360-degree physical consistency as the same person even in profile, from behind, low angles, bird's-eye views, or mid-air rotations. Apply natural lag to hair, ribbons, white fur, and clothing according to speed, wind pressure, and acceleration. [Ghost Consistency] Use the white ghost from [Reference Image] as the basic design for summoned entities. Maintain the soft, semi-transparent white-to-light-blue body, round cute silhouette, red-to-pink eyes and mouth, and light pink cheeks, making them charming companions rather than scary spirits. Summoned individuals are of the same type but not exact copies; give them small individualities in size, expression, tail curve, and flight style. Mass spawning is prohibited. Increase by one at a time in the early stage up to a maximum of 5. They fly three-dimensionally around the protagonist at different heights, distances, and speeds. The ghosts maintain the same anime-world texture as the protagonist, but the photorealistic city behind them should be naturally visible through their semi-transparent parts. [Setting | NEO FUTURE CITY] A giant near-future city entering the blue hour at dusk. Simple building rooftops are prohibited. The starting point is a massive floating aerial public plaza in the mid-section of a skyscraper cluster. Located hundreds of meters above ground, several skyscrapers are connected by glass skybridges, magnetic transit rails, massive structural beams, and aerial walkways. AAA-grade sci-fi movie photorealistic texture. Features giant glass facades, titanium, wet black stone floors, concrete, translucent panels, movable louvers, subtle glowing traffic guidance lines, and unmanned transit systems moving in the distance. Do not use excessive cyberpunk neon. Establish it as plausible near-future architecture. Above are giant multi-level roads and magnetic rails; in the distance, aerial transit craft. Natural evening light shines through thin clouds, casting blue and gold reflections on glass buildings. Skyscrapers are placed at 20-50m intervals for later web-swinging, clearly showing outer walls, beams, and antenna structures where webs can attach. [Camera Design] AAA-grade action movie + high-end sports commercial quality. Full-frame digital cinema camera. 18-21mm for high-speed movement, 24-28mm for character close-ups, ultra-wide macro intensity when entering the ghost's mouth. 60fps. Maintain a shutter speed that makes high-speed motion readable, adding natural motion blur only in necessary directions. Prohibit meaningless 360-degree rotations, random rolls, or AI-style sudden zooms. Camera rolls, banks, and orbits must have a clear center of motion like the protagonist, a ghost, or web tension. Do not keep speed constant; create a camera rhythm that repeats sudden acceleration, ultra-close-ups, orbits, momentary speed differences, and re-acceleration. [0.00-3.50s | Girl Running + First Ghost Summon] Intense motion from the first frame. The protagonist from [Reference Image] is running at full speed toward the camera in the giant near-future aerial plaza. The external camera retreats at high speed from a low angle about 1.5m in front of her. She laughs happily; the white fur, red ribbon, and blue hair sway violently backward due to speed. The anime protagonist's shadow flows fast over the wet live-action floor, and the colors of the blue hair and red ribbon reflect momentarily on the surrounding glass. Around 0.70s, she continues running while swinging her right hand sideways. A 2D paint stroke, like a single stroke of red, blue, and white paint, is drawn in live-action space. The moment it forms an arc, the first ghost pops out with a 'pop!' sound. It's not a magic circle, but a flashy yet brief painting effect. The ghost flies alongside her shoulder; she looks sideways and laughs, saying 'Can you keep up?' in Japanese with perfect lip-sync. The ghost bounces happily, but she never stops. [3.50-7.50s | Successive Summons While Running] The protagonist runs at high speed, weaving through structural beams and transparent panels of the plaza. The camera orbits rapidly from the front to the left side as she half-rotates her body while running. Cutting the air diagonally with her left hand summons the 2nd ghost from a pale blue brushstroke with a 'pop!'. A flick of the wrist produces a short red paint splash for the 3rd. Circling her arm overhead cuts the space with a white ribbon-like stroke, summoning the 4th from its tip. Finally, pushing both hands forward once creates a slightly larger 5th ghost where thick blue, red, and white brushstrokes cross. Do not generate them all at once; they must visually increase 1->2->3->4->5, and previously summoned ones do not disappear. The five fly at different heights and radii (above, side, waist, front-above, behind), forming a three-dimensional spiral around her. One passes very close to the lens, showing the character and city refracted through its semi-transparent body, but this is not an edit cut. [7.50-10.20s | Giant Ghost Moves Forward] She accelerates further toward the edge of the aerial plaza. Ahead is a massive urban drop of hundreds of meters, but she doesn't slow down. The five ghosts also speed up, and the largest one flies ahead of her, flying backward while rapidly enlarging to 3-4 meters in diameter. Maintain the cuteness and basic design. The giant ghost looks at her and shouts 'Here we go~!' in a bright, cute voice, lip-syncing perfectly. As the line ends, it opens its mouth wide. The inside is not a black hole, but an exotic space where red, blue, white, and pink 2D paints swirl at immense speed. She laughs, saying 'Eh, in there?!' in surprise but keeps running. A powerful suction comes from the ghost's mouth, pulling her hair, ribbon, and fur forward, with the smaller ghosts also drifting toward it. The camera moves behind her, getting physically sucked into the giant ghost's mouth along with her. [10.20-12.20s | Single Large PAINT IMPACT] She jumps off the edge of the plaza, and the momentum and suction pull her into the giant ghost's mouth. As she passes the edge, generate the flashiest effect in the 30 seconds: a massive 2D paint stroke of red, blue, white, and pink, along with liquid ink, cel-anime speed lines, and rough brush splatters that circle the entire photorealistic screen at high speed, forming a 'PAINT IMPACT' as if a giant brush repainted the film itself. Not an explosion or laser. The camera passes through the paint layer physically. The mouth interior is not a long tunnel but is passed in under 1 second, revealing the high sky over the same near-future city. The paint splits left and right, ejecting her and the camera into the urban sky. This is the only major effect transition; do not repeat similar transitions. [12.20-14.50s | Ejection into Urban Sky] She is thrown at high speed into the sky hundreds of meters above the photorealistic near-future city. Below are multi-level roads, giant glass buildings, and transit rails. She does not fly freely; after the upward inertia is lost, she enters a clear gravitational fall. The camera dives from her side to get beneath her, looking up from a low angle with the city as a background. The five ghosts rejoin her one by one. Only the ghosts can fly freely. One aligns next to her; while falling, she laughs and shouts 'Whoa, so high!'. The ghost laughs too. She quickly switches expressions, fixing her gaze on a skyscraper ahead to the right. [14.50-18.00s | 1st Web Shot -> Attachment -> Tension -> High-speed Swing] A giant glass and metal skyscraper is ahead to the right. While falling, she swings her right arm forward from the shoulder and snaps her wrist sharply. A single, thin, strong white-to-light-blue web line is clearly shot from her hand. The web is depicted as a translucent high-strength fiber existing in live-action space. Show the entire process: extending from her hand, crossing the air, and reaching the building wall. It attaches with a 'click!' and a tiny bit of live-action dust. Immediately after, the line pulls taut. Tension transfers from her right arm to her shoulder, chest, waist, and legs. Her downward velocity converts to forward velocity, entering a massive pendulum orbit centered on the attachment point. The camera follows, banking steeply to the right. The five ghosts chase at high speed—one in front, two on the sides, the rest behind. Ghosts do not use webs. [18.00-21.00s | Ultra-high-speed Swing Near Building + Dialogue] She reaches the lowest point of the swing at maximum speed, passing within 1-2m of the live-action building wall. Reflections of the anime girl and ghosts race across the glass. She tilts her body sideways, bending her legs to narrowly avoid building protrusions. The camera moves into the narrow gap from behind-below, using a 21mm wide-angle to emphasize speed and distance. A ghost catches up: 'Wanna go faster?'. Looking ahead, she happily says 'Of course!'. Perfect lip-sync. Do not stop the swing, hair/ribbon/fur movement, or background flow during the dialogue. Just as the camera zooms into her face from the side, she opens her right hand to release the first web. Tension vanishes, and inertia shoots her diagonally upward at high speed. [21.00-24.00s | 2nd Web + Further Speed Increase] She maintains her upward momentum while quickly half-rotating her body. The camera performs a purposeful 120-degree orbit around her to capture her face from the front-above. Another skyscraper appears front-left; she extends her left arm and shoots a 2nd web line. Show the tip flying through the air and attaching to the upper wall of the left building. 'Click!'. The line tenses, pulling her left arm, and the tension shifts her direction sharply to the left. She describes a giant arc, passing under a near-future aerial transit rail. The camera rolls left with the tension, causing the horizon to tilt for a physical reason. The ghosts cross paths around her while chasing. A larger ghost pulls up alongside: 'Can you still go?'. She laughs cutely in the wind, shouting 'I'm going to pick up even more speed!'. This is a key line for the latter half. She holds the left-hand web and continues the swing at max speed while speaking. [24.00-27.00s | Web Release -> Big Jump over City] She enters the upward part of the 2nd swing, rising toward a building rooftop. At the moment when speed converts most to height, she releases the web, getting ejected into the air higher than the roof. The five ghosts rise together. While ascending, she extends one leg forward, twists her body 180 degrees, and spreads her arms. Her blue hair, red ribbon, and fur follow with a large lag due to strong wind. The camera rises sharply from beneath her, overtakes her, and moves to the front, facing her while retreating. Behind her is the massive photorealistic near-future city hundreds of meters below. Her ascent slows, transitioning naturally into free fall. The five ghosts begin to gather around her, adjusting their speed. [27.00-30.00s | Cute Mid-air Pose + Painting Effect] For the finale, do not use shutter or lens effects. Falling through the urban sky, she faces the camera. The five ghosts are positioned three-dimensionally (sides, top, feet, behind). She smiles mischievously, brings her right hand to her face for a small V-sign, slightly bends her left leg, and assumes a cute, confident mid-air pose characteristic of [Reference Image]. However, she is not static; falling, hair, ribbon, and ghost movements continue. One ghost near her shoulder mimics her pose. Another smiles at the camera. She releases the V-sign and draws a large stroke in the air with her right index finger. A vivid blue->red->white->pink 2D paint line emerges from her fingertip, perfectly synchronized with her arm's trajectory to form a giant arc. The line circles around the character and ghosts, forming an abstract painted silhouette reminiscent of a one-stroke heart and star, without using letters or logos. The five ghosts don't disappear; they converge around her, each leaving a small colored tail. The camera zooms in as she winks. The live-action city, vivid 2D anime girl, white ghosts, and giant brushstrokes reach maximum density. At 29.70s, she thrusts the V-sign slightly toward the camera, and the ghosts strike cute poses simultaneously. The paint line circles the perimeter to decorate the frame. At 29.90s, a giant red/blue/white/pink stroke crosses very close to the lens without blacking out. The final frame at 30.00s is the vivid moment of the anime girl and five ghosts striking poses in the center of the hand-drawn effect high above the photorealistic city. Fades are prohibited. [Sound/Music] High-tempo bright electro-pop x futuristic breakbeat from the start. Heavy EDM is prohibited. Use a tight kick, clean bass, crisp percussion, short digital synths, and playful plucks to balance cuteness and high-speed action. Sync short 'pop' and 'swish' sounds to each summoning; music layers increase as more ghosts appear. Ghost voices are cute and light with an airy feel, but avoid excessive echo. The protagonist has a bright, free-spirited, slightly mischievous young adult female voice. No toddler voices. Dialogue is only at specified points: 'Can you keep up?', 'Here we go~!', 'Eh, in there?!', 'Whoa, so high!', 'Wanna go faster?', 'Of course!', 'Can you still go?', 'I'm going to pick up even more speed!'. Do not repeat lines or speak at other times. Do not confuse the protagonist's voice with the ghosts'. Perfect lip-sync for all. The suction into the ghost's mouth features a low suction sound + high-speed paint swoosh. A strong musical accent for the PAINT IMPACT at 10s. Release wind pressure sound when ejected into the sky. Web shooting 'thwip', wall attachment 'click', short low frequency on tension, and pitch-shifting wind noise during swings. Sync a dry brush swipe sound to the final painting effect. [Absolute Priorities] Maintain the protagonist from [Reference Image] as the same person throughout. She remains an anime character for 30 seconds and must not be photorealized, 2.5D, or 3D. The background is a strictly AAA-grade photorealistic live-action city. Intentionally keep the heterogeneous texture difference between the protagonist and the background. However, ensure light, shadows, reflections, contact, occlusion, wind, and speed are physically consistent within the same space. No camera props or selfies. No generated photos. Summon ghosts one by one via the protagonist's hand movements, increasing to 5 without deleting others. Summoning is a brief 2D paint expression. Only one major effect change: entering the giant ghost's mouth. The mouth opens right after the ghost says 'Here we go~!', and the camera/protagonist are sucked in. Continuous ejection from the mouth to the city sky. She falls via gravity, not free flight. Web lines must shoot from her wrist/hand; show the tip flying, attaching, the line tensing, and the tension converting the fall into a swing. Establish 'shoot->attach->tension->swing' at least twice. Ghosts fly alongside and remain until the end. Natural dialogue between them during high-speed movement. Finish with a cute pose of the protagonist and 5 ghosts synchronized with a giant 2D paint line she draws. [Prohibited] Photorealism of the protagonist, 2.5D, 3D anime, game CG, plastic 3D, face changes, hair color/style changes, outfit changes, loss of accessories (ribbon, sunglasses, fur), character duplication, becoming another person, age changes, additional human characters, scary/zombie/spirit ghosts, blood, gore, fusion of protagonist and ghost, extra limbs, 6 fingers, fused fingers, body twisting, joint failure, random teleportation, standard edit cuts, midway blackouts,

Create a cute cinematic 14-second 2D anime-style video featuring two completely new child characters in a sunny park: a girl with long dark-brown hair, expressive brown eyes, a pink dress, white socks and pink shoes, and a boy with messy black hair, dark eyes, a blue hoodie, black shorts and white sneakers. Keep their new faces, hairstyles, outfits and proportions consistent. They happily fly a bright orange-red kite through the colorful park until a strong breeze sends it into a tall tree. The boy uses a fallen branch to free the kite while the girl helps, and they catch it together with happy smiles before running through the park and flying it again in warm golden sunlight. Use beautiful Japanese anime-inspired 2D animation, clean line art, expressive eyes, soft cel shading, vibrant colors, detailed backgrounds, smooth movement, cinematic lighting and natural wind effects. No face changes, character redesign, extra characters, duplicates, distorted faces or hands, flickering, photorealism, 3D style, text, logo or watermark. smooth cinematic anime animation.

Vertical 9:16 video, photorealistic high-end nail salon aesthetic mixed with soft anime-style wide shots. A seamless step-by-step close-up sequence of a professional manicure process on elegant long almond-shaped nails: filing, dusting off white powder with a fluffy brush, wide shot of a cute anime girl sitting in a bright luxury nail salon vanity, applying glossy gel base coat, curing under a UV lamp, buffing iridescent pearl white chrome powder with pearl accents, applying shiny top coat, and placing cuticle oil drops. Ends with the client admiring her glossy metallic pearlescent silver nails under warm glowing lights. Macro-focused shots, ultra-realistic textures, pastel tones, smooth transitions, soft cozy ambiance, cinematic lighting, 4k resolution.

Japanese full-color anime style. Japanese TV anime style. Cel-shaded animation. Anime-style screen capture focused on flat coloring. Noble's mansion. The long-haired blonde character from Image 1 is sitting in a chair. In the chair opposite him sits the red-haired character from Image 2, who is the stepbrother of the character in Image 1. 0-4 seconds: A diagonal high-angle shot of the blonde long-haired character from Image 1. "Brother, what brings you here today?" 4-8 seconds: A diagonal low-angle shot looking up at the red-haired character from Image 2. "Withdraw from Father's business, Reiss." (4 seconds) 8-13 seconds: A side-view shot of the upper body of the blonde long-haired character from Image 1. "Did turning the business into a profit displease you that much?" Eyes closed quietly. Panning left. 13-17 seconds: A close-up of the red-haired character's clenched fist from Image 2. "You're after the family headship, aren't you?" Trembling with anger. 17-20 seconds: A straight-on shot of the red-haired character's upper body from Image 2. Pointing at the screen and shouting, "Even though you're just a concubine's child!" 20-21 seconds: Water is poured over the head of the red-haired character from Image 2. 21-25 seconds: A diagonal low-angle shot looking up at the black-haired maid from Image 3. She has a condescending gaze. The black-haired maid is holding a water pitcher with looking down eyes. "...Pardon me. My hand slipped." Cold gaze. 25-28 seconds: A scene where the red-haired character from Image 2 screams "You!" and raises his arm. Side-view shot, showing the red-haired character from Image 2 from the shoulder up. 27-28 seconds: A close-up of the red-haired character's shoulder from Image 2. The hand of the blonde long-haired character from Image 1 firmly grabs the shoulder of the red-haired character from Image 2. The face of the blonde character is not shown, only the hand grabbing the shoulder. 28-30 seconds: Half of the red-haired character's face from Image 2 is shown facing the camera in the foreground. In the background, the face of the blonde long-haired character from Image 1 appears. "...Please leave." He speaks with an intimidating tone. The red-haired character turns pale with fear at the intensity.

Cinematic anime short film clip, 15 seconds. Muddy motocross track, dark overcast cloudy sky, wet churned up dirt everywhere, deep ruts in the track, no crowd just raw open terrain. CHARACTER: Use uploaded character sheet as strict visual reference.

The small things that go wrong on a crowded beach, shot like a European summer film — rich, warm, unhurried — so the first frame feels like a still from a beloved movie. TECHNICAL NOTES: Shot on 35mm film, fast spherical primes (Cooke-style, 40/50/85mm), T2.8 for a soft painterly falloff, a warm glow in the highlights. Delivered clean full-frame 16:9, no borders. Camera is elegant, intimate handheld — composed but breathing, close to the skin, unhurried. Light is hard Mediterranean noon softened by dappled shade and reflected sand-bounce, warm and enveloping. Grade: rich, saturated, romantic — golden sun-kissed skin, deep foliage and pine greens, terracotta and ochre, a saturated cobalt sea, creamy highlights that stay present rather than blowing out; gentle filmic contrast, a nostalgic warm bias. Cast: natural, effortlessly sun-tanned Mediterranean beachgoers. Costume: chic and timeless — high-waisted swim briefs, elegant one-pieces and simple bikinis in cream, rust, olive and gold; tortoiseshell sunglasses, fine gold chains, a wristwatch. SHOT 1 (00:00–00:02) — Hook - A close, sun-warmed face turns lazily toward the lens in dappled light, a cobalt sea burning bright and soft behind; the whole frame glows golden. - Slow intimate handheld drift-in, shallow, painterly. SHOT 2 (00:02–00:04) — Sand - Children run past a woman lying on her front; sand scatters warmly across her oiled, golden back and an open sandwich. - Low, close handheld, soft focus riding her skin. SHOT 3 (00:04–00:06) — Umbrella - A gust tugs a canvas umbrella loose; tanned figures rise unhurried to steady it, laughing. - Loose elegant handheld, saturated crowd soft behind. SHOT 4 (00:06–00:08) — Seagull - A gull snatches a tomato from a sandwich in a curly-haired man's hand; he startles, then grins. - Quick warm handheld reframe up to the bird against blue. SHOT 5 (00:08–00:10) — No Room - A family folds into a thin gap between sun-browned bodies, close and companionable. - Static, intimate, richly saturated. SHOT 6 (00:10–00:12) — Stray Ball - A ball rolls into a couple's shade and tips a glass bottle; the spill beads on warm sand. - Soft handheld settle on the spill. SHOT 7 (00:12–00:14) — To the Water - A tanned young man in high-waisted briefs wades into the cobalt shallows, sunlight scattering off the surface. - Handheld alongside, sea deep and saturated. SHOT 8 (00:14–00:15) — Wind - A woman gathers a towel that catches the breeze, fabric warm and luminous against the sea. - Close handheld, golden and soft.

Main Subject: \n\nYoung Korean woman, 23, natural everyday appearance, shoulder-length black hair loosely tied with a simple fabric scrunchie, wispy bangs, realistic skin texture, minimal makeup, calm and slightly playful personality. Wearing a faded red zip-up hoodie over a white T-shirt, loose dark-blue jeans, white sneakers, and a small navy canvas shoulder bag. Maintain consistent identity, clothing, hairstyle, and appearance throughout.\n\nLocation: Small neighborhood laundromat on a rainy afternoon. Old front-loading washing machines, plastic laundry baskets, fluorescent ceiling lights, fogged windows covered with rain droplets, vending machine, folding table, faded Korean instruction stickers, and a quiet residential street visible outside.\n\nVisual Style: Ultra-realistic documentary realism. Ordinary, candid behavior with authentic early-2000s Korean neighborhood details.\n\nCamera Style: Early-2000s Sony MiniDV camcorder. Heavy handheld shake, imperfect framing, autofocus hunting between spinning washing machines and her face, fluorescent flicker, exposure pumping near the bright window, faded colors, soft contrast, slight motion blur, DV compression artifacts, mild digital noise, no stabilization.\n\n00:00–00:03\nShe enters carrying a blue laundry basket, shaking a few raindrops from her hoodie. The camera briefly focuses on the wet floor before finding her.\n\n00:03–00:06\nShe loads clothes into an old washing machine and pours detergent from a small plastic bottle. The machine door closes with a heavy click.\n\n00:06–00:09\nShe sits on a plastic chair watching the clothes spin through the round glass door. She rests her chin on one hand, looking slightly bored.\n\n00:09–00:12\nRain becomes heavier outside. She notices the camera and makes a small amused expression before pointing toward the spinning laundry.\n\n00:12–00:15\nThe washing machine suddenly finishes its cycle with a loud beep. She stands and opens the door as the camera moves closer, loses focus, and abruptly cuts to black.\n\nAudio: Natural ambience only—washing machines humming, water spinning, fluorescent buzzing, rain against windows, footsteps, plastic basket sounds, machine beeps, distant traffic. No music. No narration.\n\nGoal: A forgotten MiniDV home video from 2004, capturing a completely ordinary rainy afternoon at a small Seoul laundromat–mundane, intimate, imperfect, nostalgic, and convincingly real.

Main Subject: Young Korean woman, 20, natural everyday appearance, medium-length dark brown hair in a loose half-up style with wispy bangs, realistic skin texture, minimal makeup, slightly sleepy but cheerful personality. Wearing a soft mustard-yellow cardigan over a white cotton blouse, dark navy pleated midi skirt, white ankle socks, brown loafers, and a small canvas shoulder bag. Maintain consistent identity, clothing, hairstyle, and appearance throughout. Location: Quiet traditional neighborhood in Gyeongju on a cool autumn morning. Low tiled-roof houses, stone walls, narrow lanes, small gardens, bicycles, old utility poles, fallen leaves, and a tiny family-owned bakery with a fogged glass window and handwritten Korean signs. No tourists or busy traffic. Visual Style: Ultra-realistic documentary realism. Genuine candid behavior, subtle expressions, natural movement, authentic Korean small-town atmosphere. Camera Style: Early-2000s consumer MiniDV camcorder. Heavy handheld shake, imperfect framing, autofocus hunting between bakery windows and her face, exposure pumping between indoor and outdoor light, faded colors, soft contrast, slight motion blur, DV compression artifacts, mild sensor noise, no stabilization. 00:00–00:03 She walks down the quiet lane carrying her shoulder bag, stepping through fallen leaves. She notices the small bakery and slows down. 00:03–00:06 Inside, she looks through the glass display and chooses a freshly baked red-bean bun. Warm bakery light reflects on the glass while autofocus struggles to find her face. 00:06–00:09 She steps outside, takes a small bite, and smiles while warm steam escapes from the fresh bread. A bicycle passes behind her. 00:09–00:12 She continues toward a nearby stone wall, checking the time on her small wristwatch as morning sunlight reaches the street. 00:12–00:15 She notices the camera, laughs softly, gives a quick little wave, and walks away toward the old neighborhood. The operator accidentally points at the pavement before the recording cuts to black. Audio: Natural ambience only–morning birds, footsteps on leaves, bicycle bell, distant scooter, bakery door chime, quiet conversation, paper bag rustling, soft wind. No music. No narration. Goal: A forgotten MiniDV recording from 2004, capturing an ordinary autumn morning in historic Gyeongju–quiet, nostalgic, imperfect, intimate, and completely believable.

Model appearance (no reference photo, text description only): An androgynous, 90s-inspired editorial model look — sharp, striking features, high cheekbones, thin natural unshaped eyebrows, smooth, flawless, beautiful skin, unfiltered lips, a direct, slightly defiant gaze. Lean, angular build. Not glamorous or symmetrically "Instagram-perfect" — an interesting, characterful face, but with impeccable, well-cared-for skin. Dark or ashy hair, styled loosely for backstage. archive fashion week backstage atmosphere and energy; editing rhythm and rapid cut pacing; vintage visual aesthetic and color grade; 2000 cam lighting quality and mood, emotional tone (candid, unposed, lively); and the makeup look (metallic/silver glittery smoky eye, blended, applied quickly backstage-style). Camera: shot as if on an amateur early-2000s camera/analog camcorder — visible grain, characteristic slightly faded color, natural handheld shake, sharp manual zoom, home-video compression artifacts. Sound: no music. Live backstage sound — background chatter, fragments of conversation, clicks of other people's cameras, rustle of brushes and sponge on skin, laughter, her own voice answering interview questions — natural, indistinct speech over the ambient noise. Storyboard — 8 dynamic backstage moments, short hard cuts, each shot a new detail or moment: 0–2s: Extreme close-up — a damp sponge presses foundation into the skin, macro texture of blending, product settling evenly onto smooth skin, handheld camera shaking, sharp zoom. 2–4s: Quick cut. Close-up — a brush lines the lower waterline with dark pencil/shadow, the eye slightly squinting, macro detail, a natural light watery reflex from touching the waterline. 4–6s: Quick cut. A brush/sponge applies silver eyeshadow to the lid, macro texture of glitter, sharp manual zoom. 6–8s: Quick cut. She answers interview questions — camera pulls back, natural conversational expression, gaze sometimes at the lens, sometimes away. 8–10s: Quick cut. Genuine laughter — head tilted back slightly or hand covering her mouth. 10–12s: Quick cut. She sits while a makeup artist's hands adjust her makeup/hair in frame — relaxed, half-ignoring the process. 12–13.5s: Quick cut. A coquettish glance over her shoulder, playfully sticking her tongue out at the camera. 13.5–15s: Quick cut. A wide departing shot — she turns and walks into the busy backstage space, silhouettes of people, clothing racks. Facial naturalness: alive expression throughout — talking, laughing, reacting, blinking naturally

Create a 15-second, 16:9 photorealistic cinematic film titled “A Last Train Home.” Keep the same young South Asian woman, face, hair, outfit and proportions throughout. 0–3s: Alone at a deserted rainy station, she waits as train headlights approach. Wet reflections, warm lights, blue moonlight and mist. Mood: loneliness. 3–6s: The train arrives through heavy rain. She turns toward it with quiet hope. 6–9s: Inside the warm train, she finds an old photograph, holds it, and gives an emotional smile. Rainy window behind her. 9–12s: The train moves through the rainy city, with blue night tones, warm lights and reflections. She holds the photograph. 12–15s: Sunrise fills the train as she looks at green fields and distant hills, holding the photograph close with a peaceful smile. Use natural expressions, realistic rain, reflections, cinematic lighting, shallow depth of field, filmic color grading, smooth camera movement and realistic motion. End with the train moving toward sunrise and fade to black.

The Casino Roof Battle Visual language: A 1973 giallo-action scene on the roof of an abandoned casino at night, neon signs flickering below. The woman wears a blinding gold lamé costume with sculptural hips, immense lashes, dark green lids, heavy rouge and deep glossed lips. The man wears a white satin gambler-villain outfit with a black cape and decorative chain details. Saturated red and gold, strong shadows, crisp side-on medium shots. English dialogue only, nasal old-film dubbing. Music: jazz bass, snare and suspense strings. Spatial continuity: Woman remains screen left facing right. Man remains screen right facing left. They stay on the same axis and never swap sides. 0:00–0:03 — Establishing two-shot Woman’s expression: delighted hostility. Man’s expression: mocking confidence. 0:03–0:06 — Fight shot 1 The woman rushes him and grabs his lapels, slamming him backward. Her expression is joyous and ferocious. His expression becomes startled and alarmed. 0:06–0:09 — Fight shot 2 He tries to throw her off, but she drives a shoulder into him again and spins him partly sideways without crossing positions. Her expression is determined and almost cheerful. His expression is frustrated and shocked. 0:09–0:12 — Victory shot He collapses against the low wall on screen right, winded. She stands on screen left with squared shoulders. Her expression becomes serenely smug. His expression is miserable and breathless. 0:12–0:15 — Final dialogue Woman: “Your bluff was physically unconvincing.” Man: “It usually helps if no one touches me.”

Generate a 21-second, 9:16 vertical, ultra-realistic cinematic video. [Scene] A bright, warm modern beige living room with white cloth sofa, plants, and natural light. [Character] A young East Asian girl (5-8 years old), cute, standing in the center. Natural expressions: surprise, shy smile, happiness. [Core Mechanism] A pair of giant real human hands enters the frame to change the girl's clothes as if she were a doll. Natural and smooth hand movements. [Timeline] 0-2s: Opening with girl in pajamas. 2-4s: Giant hand removes the top. 4-6s: Hand removes the pants; puts on a beige cardigan. 6-8s: Hand puts on olive green shorts. 8-11s: Hand puts on rainbow socks. 11-13s: Hand puts on beige snow boots. 13-15s: Hand puts on a knit hat. 15-17s: Hand gives her a brown handbag. 17-21s: Final reveal of the outfit. Hand points at her like "done." [Camera] Fixed front full shot with slight handheld breathing feel. [Constraints] Character consistency (face/hair), realistic human hand proportions, smooth clothing transitions, no artifacts or text.

Photorealistic absurd surreal comedy scene in a small indoor hotel pool. Low-budget real-world environment: tiled pool, metal ladders, fluorescent lighting, depth markers on the wall, white plastic chairs, a vending machine in the background, and a few casually dressed people in swimwear watching. A giant beautiful young woman is submerged in the pool up to her neck and upper chest, as if her body is enormous and fills most of the pool. She has wet slicked-back dark hair, expressive eyes, realistic wet skin, and a slightly amused then surprised expression. Her face and shoulders are huge compared to the rest of the environment. A normal-sized adult man in swim trunks carefully walks down the pool ladder as if he is just entering the water. He holds a white cup in one hand. As he steps down, it becomes clear that because of the bizarre scale mismatch, he is stepping directly onto the giant woman’s face. He awkwardly steps on her nose and cheek while trying to keep balance. The woman looks confused and cross-eyed toward him. The man slips, grabs at her nose for balance, and then slides toward her open mouth. By the end of the shot, he accidentally tumbles into her mouth in a ridiculous surreal way. She looks shocked. The people in the background react with surprise and laughter. Camera style: Recorded like a viral smartphone clip. Slight handheld movement, but mostly stable from the poolside. Realistic phone-video look, mild compression, indoor echo, natural pool reflections, believable lighting, not cinematic. The scene should feel like a bizarre real video someone captured by accident. Tone: Absurd, funny, surreal, and visually shocking — not gory, not violent, not horror. The humor comes from the impossible scale and the dead-serious realism. Action timing: 0–3s: Show the indoor pool and the giant woman occupying most of it. The man starts climbing down the ladder. 3–7s: He continues stepping down and accidentally plants a foot on her face/nose. 7–11s: He loses balance, grabs at her nose, and slips toward her mouth. 11–15s: Her mouth opens in surprise, he tumbles into it, and the bystanders react. Important details: Keep the scale relationship very clear: giant woman, tiny man, normal-size background people. Keep the environment realistic and grounded. Preserve believable water interaction and reflections. No gore, no body horror, no extra limbs, no broken anatomy, no glitchy transitions.

Create a seamless 9-second surreal animation blending Tokyo with giant desserts: a massive silver fork plunges into a miniature Tokyo street and lifts a strawberry shortcake slice with whipped cream, strawberries, and red glaze; cut to street level where a giant chocolate slab sits in the pavement as thick caramel pours over it and pedestrians pass by; show a bustling skyscraper avenue decorated with strawberries and whipped-cream rooftops; finally, zoom out to reveal the entire city district sitting on a giant decorated cake on a white plate as a huge fork carves out and lifts a slice of the city-cake, seamlessly returning to the aerial Tokyo view.

With a single slow-motion portable handycam that floats and organically weaves through a completely frozen scene, everyone and moving objects stop mid-motion like a live diorama. The handheld camera slowly spins and floats around the ball, providing one uninterrupted motion.

Create a short, visually satisfying mixed-reality stop-motion video where [DRAWN OBJECT/SKETCH] seamlessly transforms into [REAL INGREDIENT/OBJECT]. A person uses a smartphone and colored pencil on a [SURFACE], sketching each item before it magically becomes real through smooth, seamless transformations. Show [INGREDIENTS] being added to [CONTAINER], then [COOKING/BLENDING/PREPARATION ACTION], followed by serving into [DISH]. Finish with the artist sketching [TOPPINGS/FINAL DETAILS], which instantly materialize into real food, creating a vibrant, aesthetically arranged [FINAL DISH]. Use realistic textures, natural hand movements, satisfying stop-motion timing, clean composition, soft lighting, and seamless transitions between pencil drawings and real objects.

Format: 16:9 | Look: Photorealistic, 8K, 35mm film, cool-toned night grade, shallow depth of field, lens grain, cinematic hard cuts, dynamic camera per shot CHARACTER DESCRIPTION The Woman (pre-transformation): Mid-20s, Western/European features, strikingly attractive and beautiful. Fair, luminous skin with a soft cool undertone that catches the neon well. Long, wavy chestnut-brown hair loose around her shoulders. High cheekbones, full lips, expressive light hazel-green eyes. Slim, toned, elegant build, roughly 5'8". Wardrobe: a fitted black leather jacket over a simple fitted top, dark high-waisted jeans, ankle boots — modern, stylish, urban-night aesthetic. Minimal jewelry (thin gold necklace) that will later echo the beast's gold markings. The Beast (post-transformation): Tall, lean, unmistakably feminine — never bulky or male-coded, and not wolf-like. Slender long-limbed frame, long legs ending in black-clawed paws. Thick ruff of cream-white fur across the chest. Coat mottled black, gold and cream. Crown of tall branching gold horns. Fine gold facial markings tracing her cheekbones and brow (echoing the woman's features). Amber-gold slit-pupil eyes. Long membranous black-and-gold wings. Trailing golden tendrils drifting from the horns and spine. The transformation is progressive and layered — skin to hide, fur in a traveling wave, muscle lengthening, nails to claws, horns branching, wings unfurling — surface-level morphing only, no wounds, no tearing, no blood. SETTING A generic Shibuya-style crossing at night — no real landmarks, brands, or logos; invented signage text only. Neon in magenta, cyan and violet, wet asphalt mirroring the light, drifting steam, light rain haze. A huge blood-red moon dominates the sky above the skyline throughout. PART 1 — 0:00–0:15 — "THE OMEN AND THE COLLAPSE" Shot 1 (0–4s) — The Red Moon Opens tight on the enormous blood-red moon above the skyline. Long continuous extreme crane-down past glowing tower screens (invented neon kanji-style signage) toward the crowded crossing below, then pushes through the crowd to settle on the woman walking among them, one hand pressed to her temple, head down, unsteady, hair damp with mist. Shot 2 (4–7s) — First Tremor Hard cut to a low tracking shot at hip height following her boots across wet asphalt — reflections of neon rippling with each step — as her stride breaks rhythm and she catches herself against an invisible wave of dizziness. Shot 3 (7–11s) — The Collapse Cut to handheld medium pushing with her through the crowd. She stumbles, breath shallow, face pale under the neon wash. Her knees buckle and she drops to the pavement. Shot 4 (11–15s) — The Empty Ring Wide crane shot as the crowd scatters back in a widening ring, phones raised, murmurs rippling outward, until she is alone, kneeling, at the dead center of the empty crossing — rain haze drifting through the neon light around her. PART 2 — 0:15–0:30 — "THE TRANSFORMATION" Shot 5 (15–26s) — One Unbroken Take Extreme low angle at asphalt level, 24mm wide lens, only her right hand in frame, pressed flat on the wet asphalt, oversized by wide-lens perspective. Camera zooms hard on the back of her hand: pale skin darkens into hide, cream-white and mottled gold-black fur emerges strand by strand, fingers lengthen, nails darken into black claws that gouge the asphalt. Camera tracks and pushes along the forearm following the wave of fur, rises to the shoulder, swings around to find her face in tight close-up — still human, jaw clenched, terror in her hazel-green eyes — then pushes into an extreme close-up of her eyes as the irises flood amber-gold, pupils narrowing to slits, the last trace of the woman vanishing. Camera pulls back into a fast orbiting handheld shake as the white ruff floods her chest, gold markings spread across her face, golden horns branch from her temples, wings unfurl behind her. The orbit whips to the front as the fully formed beast rises to full height. Shot 6 (26–30s) — The Roar Low, wide-angle hold as she plants her clawed paws and throws her head back — a full-body roar, wings snapping open to full span, golden tendrils whipping outward, neon light rippling across her mottled coat. PART 3 — 0:30–0:45 — "PANIC, PURSUIT AND THE HOWL" Shot 7 (30–33s) — Panic Fast handheld at street level as the beast sweeps her clawed limbs and lunges forward, wings flaring wide. The crowd breaks and scatters, people sprinting past the lens, camera whip-panning through the chaos as neon streaks smear across the wet asphalt. Shot 8 (33–36s) — The Chase Beat Low tracking shot alongside her as she bounds on all fours between stalled cars and scattering crowds, claws sparking faintly off metal and asphalt, wings half-folded for speed, camera struggling to keep pace — motion blur and lens flare from passing neon signs. Shot 9 (36–39s) — The Climb Fast upward-tracking shot as she launches from the crossing and scales a building face in fluid bounds, claws punching into the facade, mottled coat and white ruff catching the neon, wings half-folded. Camera rises with her, orbiting outward to reveal the city sprawling below. Shot 10 (39–42s) — The Perch Cut to a slow dolly-in on the rooftop ledge as she lands in a low crouch, wings settling against her back, chest heaving, gold facial markings glowing faintly under the moon, ears (horn-crown) tracking the city sounds below. Shot 11 (42–45s) — The Howl Low-angle wide on the rooftop tower — she stands tall, lit by neon from below, the enormous red moon directly behind her, her crown of golden horns and outstretched wings rim-lit against it. She rears back and howls, fur and golden tendrils rippling in the wind. Camera cranes slowly up and back into a vast wide, city lights sprawling out, rain drifting across the frame as the shot holds on the silhouette-and-glow composition for the final beat. EXTRA ACTION BEATS (optional inserts to extend further / alternate cuts) Insert after Shot 4: a tight cutaway of a dropped phone on wet asphalt, screen cracked, still recording, neon reflected in the cracked glass. Insert after Shot 6: a low static shot of scattered crowd members frozen mid-flight, silhouetted against a neon billboard, as the beast's shadow sweeps over them. Insert after Shot 9: a brief aerial drone-style shot circling the tower as steam vents hiss below her perch, city traffic streaking far beneath. Alternate ending extension: slow-motion feather/fur wisp caught in the wind drifting past camera as the howl fades, rain intensifying, screen fading to the red moon.

Create a cute cinematic 14-second 3D cartoon using the two reference characters, keeping their faces, hairstyles, clothes and appearance consistent. A young woman in a pink outfit walks through a sunny green park with a cute blonde little boy in a dinosaur T-shirt. The boy notices a colorful butterfly, points excitedly, and they happily follow it through the flowers. The boy picks a small flower and gives it to the woman, and she smiles and gently hugs him. Use smooth expressive animation, soft cinematic lighting, vibrant colors, detailed park scenery, natural camera movement and a warm family-friendly atmosphere. No character changes, face distortion, extra characters, outfit changes, flickering, deformed hands, text or watermark. 16:9 vertical, high-quality cinematic 3D animation.

polished 3D animated comedy. Inside a busy cartoon pizzeria kitchen, one expressive chef confidently spins a large pizza dough disc and tosses it high into the air. He waits underneath with both hands ready, but the dough never comes back down. He slowly looks up, confused. Suddenly the dough swoops back into frame like a giant floppy bird, flapping wildly through the kitchen. It smacks the chef flat across the face, completely covering his head for one beat. He yanks it off in panic, but the dough whips around again, snatches his tall chef hat clean off his head, then flies toward the open kitchen window with the hat stuck proudly on top of it. The chef lunges after it and misses, crashing chest-first into the counter as the dough escapes outside. End on the chef staring through the window in stunned disbelief while the dough disappears into the distance wearing his hat. Bright high-quality 3D animation, exaggerated squash-and-stretch, fast readable physical comedy, expressive body language, clean cause-and-effect, energetic camera movement, one chef, one kitchen, no extra characters, no text, no logos, no famous celebrity faces, no recognizable actors, no movie-star resemblance, no public-figure likenesses.

Create a cinematic 3D animation of a young couple revisiting their life journey through a glowing scrapbook—proposal at a Ferris wheel, wedding, new home, Italy road trip, and growing family with two children and a golden retriever. Return to the present as they close the album with the text “самые важные главы жизни еще впереди,” ending with a loving forehead kiss and warm smile. Use emotional storytelling, soft lighting, and seamless transitions.

3D animated comedy. In a cozy living room at night, one fluffy house cat is deeply asleep on the couch in a completely innocent pose while a digital clock on a side table reads 02:59. The moment the clock flips to 03:00, the cat’s eyes pop open and it instantly transforms into a tiny maniac. It blasts off the couch, sprints across the room at absurd speed, bounces off the sofa, skids across the table, races up the wall, runs upside down across the ceiling, then rockets back down, knocking a pillow, a sock, and a blanket flying in every direction. The whole burst should feel fast, wild, and ridiculous, like pure 3 a.m. cat madness. Then the noise wakes the owner, who stumbles in half asleep, messy and confused, trying to understand the chaos. Final payoff: the owner turns toward the couch and sees the cat already back in the exact same spot, curled up fast asleep like a perfect angel, while the room around it is a total mess. The cat must look completely innocent, as if the owner imagined everything. Bright polished 3D animation, exaggerated facial expressions, strong physical comedy, fast readable timing, playful cinematic camera, clear cause-and-effect, one cat, one living room, one short burst of chaos, funny and crystal clear. No text, no logos, no famous celebrity faces, no recognizable actors, no public-figure likenesses.

30-SECOND CINEMATIC SHORT Create a 30-second cinematic sequence with the visual discipline of a major feature film. The world is entirely made from paper, cardboard, ink, and delicate handcrafted materials. 00:00–00:05 — THE FIRST FOLD Begin with an extreme macro shot of a blank sheet of textured paper. A single fold slowly forms across its surface. Lens: 100mm macro Camera: Completely static Soft daylight moves across the paper as the fold continues. Cut precisely as the paper reaches its final shape. 00:05–00:11 — THE CITY EMERGES Transition into a 50mm shot. The folded paper begins forming miniature streets and architectural structures. Buildings rise naturally from the surface through carefully constructed paper folds. The camera slowly tracks sideways. Tiny windows catch the light. Nothing feels computer-generated; every surface should have believable paper fibers, folds, shadows, and imperfections. 00:11–00:17 — THE REVEAL Move into a 35mm shot. The camera pulls backward. The small arrangement is revealed as an enormous handcrafted paper city stretching across a large table. Roads connect different districts. Bridges cross miniature rivers. Paper trees move gently from an unseen breeze. 00:17–00:23 — MORNING Transition into a slow overhead crane shot. Warm sunlight gradually spreads across the paper city. Shadows from the buildings become longer and more defined. Tiny paper windows begin reflecting the sunlight. The camera continues rising, revealing the geometric relationship between the streets and buildings. 00:23–00:27 — THE DETAIL Cut to an 85mm close-up. Focus on a tiny paper clock mounted on one building. The clock moves forward by one minute. Rack focus from the clock to the miniature skyline behind it. 00:27–00:30 — FINAL IMAGE Return to an extreme wide shot. The entire paper city sits beneath a large studio window. The morning light completely fills the miniature world. Camera slowly pulls backward until the city becomes a small object within the larger room. Fade to black. CINEMATOGRAPHY Large-format feature-film aesthetic. 100mm macro for texture. 85mm for detail. 50mm for natural perspective. 35mm for environmental shots. Controlled dolly and crane movements. Realistic depth of field. Natural focus transitions. Soft motion blur. Subtle lens imperfections. No artificial camera shake. MATERIAL DIRECTION Every surface must visibly behave like physical paper. Visible paper fibers. Natural folds. Tiny imperfections. Soft cardboard edges. Realistic contact shadows. Believable paper thickness. Natural material deformation. LIGHTING Soft morning daylight. Large window as the primary source. Natural bounce light. Gentle shadows. Subtle warm highlights. No artificial neon effects. No excessive visual effects. CONTINUITY The same paper city throughout the entire sequence. Buildings maintain identical shapes and positions. Roads remain connected. Paper materials remain consistent.

Create a 30-second cinematic animated short featuring two original fantasy characters: a young female Taoist warrior and her gigantic reluctant giant partner. Visual style: high-end Western animated feature film, stylized realism, expressive characters, cinematic lighting, detailed environments, polished 3D animation, natural physics, expressive facial animation, comedic timing, dynamic cinematic camera movement. Keep the character designs completely consistent throughout the entire video. Scene 1, 0-5s: The New Target Begin with a quiet atmospheric mountain landscape. A gentle mountain breeze passes through the frame. The camera slowly tilts downward, revealing the top of a large woven bamboo conical hat [from the reference image] The girl's long dark hair moves naturally in the wind beneath the hat. The camera continues slowly downward until her face [from the reference image] is revealed. She looks completely calm and slightly annoyed. She picks up the phone and says in a casual, matter-of-fact voice: "He's the new target? Got it." She rolls her eyes. Subtle wind, distant birds, soft mountain ambience, cinematic silence before the dialogue. Scene 2, 5-11s: The Reluctant Partner Cut to a dramatic wide shot. The female Taoist stands on the edge of a huge mountain cliff [from the reference image]. Vast mountains, mist and clouds stretch into the distance. Behind her, the gigantic blue-gray stone-skinned giant [from the reference image] is crouching on the ground, taking a rest. His enormous body towers over the landscape. She casually turns toward him and says: "Time to work." The giant slowly raises his head with an exhausted expression. He looks deeply annoyed. "I haven't rested enough yet!" His voice is deep and powerful, but his expression should feel more like an exhausted coworker complaining about overtime than an angry monster. The Taoist remains completely unfazed. Scene 3, 11-23s: The Car Argument Cut to the Taoist walking confidently toward a rugged vintage open-top convertible [from the reference image] parked nearby. She gets into the driver's seat, starts the engine and casually says: "Come on, come on. Let's go." The giant stands behind the vehicle, looking at the tiny car with disbelief. He complains: "Why do YOU get to drive while I have to run again?" The Taoist looks back at him without sympathy. "Then build a car big enough for you." She immediately turns forward and drives away. The giant stands frozen for a beat. The vehicle disappears down the mountain road, leaving a small cloud of exhaust and dust drifting directly into his face. He slowly wipes the dust from his face with an irritated expression. Comedic timing, exaggerated facial animation, but grounded physical movement.

intense epic 3d artwork, a mech cyber panda warrior character clenching her fist, pink, white, and black colors, neon glowing eyes, kinetic chromatic burst motion blur background intense epic 3d artwork, a mech cyber panda warrior character clenching her fist, pink, white, and black colors, neon glowing eyes, kinetic chromatic burst motion blur background, epic, cinematic, multi-shot, no glitches, amazing vfx

Use the first reference image (9 panel storyboard contact sheet whose cells are labelled TAKE 01 to TAKE 09) as ordered key poses TAKE 01-TAKE 09 for ONE unbroken LOCKED shot: the camera never moves, the action passes through TAKE 01 to TAKE 09 inside a fixed frame, without cuts or holds; completely ignore its line-sketch style, panel frames, gutters and any text. LOOK & WORLD: Modern 2D cel-shaded anime with painted backgrounds and clean confident linework. On a city rooftop at dusk, two fighters in simple training clothes trade staff work inside a chalk circle. Warm low dusk light, long shadows on concrete, cool blue city haze behind. One fixed 24mm lens, deep focus, cinematic. The lens never changes. CONTINUITY (binding, not flavour): Locked frontal wide at waist height, unchanged from TAKE 01 to TAKE 09. Stage anchors hold their positions throughout: rooftop water tank screen-left, chalk circle center, city skyline screen-right, crate at right. The chalk circle is unbroken through TAKE 04; scuff marks start at TAKE 05 and persist. One staff drops at TAKE 06 and stays on the ground for the rest of the take. The screen-left fighter is upright through TAKE 07 and grounded from TAKE 08 on. Stage anchors hold their screen positions throughout, the camera height never changes, and every state pinned to a TAKE number is true from that TAKE onward and false before it. Never invent a shot that the sheet does not have. ACTION & CAMERA: the camera is bolted down. No cuts, no zoom, no pan, no tilt, no tracking, no shake, no reframing. Only the action moves, entering and leaving a fixed frame. (TAKE 01) locked, no movement, Facing off: both fighters stand at opposite edges of the chalk circle, staffs low. (TAKE 02) locked, no movement, First approach: the screen-left fighter steps in, staff rising to guard. (TAKE 03) locked, no movement, First clash: the staffs meet high in the center of the frame. (TAKE 04) locked, no movement, Bind: bodies close, staffs crossed, both leaning into each other. (TAKE 05) locked, no movement, Break: the screen-right fighter pivots away, first scuffs on the circle. (TAKE 06) locked, no movement, Disarm: the screen-right fighter staff spins out to frame-right and lands on the ground. (TAKE 07) locked, no movement, Empty hands: the screen-right fighter raises open palms, the other still guarding. (TAKE 08) locked, no movement, Takedown: the screen-left fighter is swept and lands flat inside the circle. (TAKE 09) locked, no movement, Hand up

Cinematic 2.5D animation: fully painterly rendering, characters and environments look like gouache concept-art paintings in motion, visible brush texture on skin, cloth and buildings, flat posterized color blocks with hard-edged light shapes, matte finish,

Task: Dreamina Seedance 2.5 — Omni ReferenceDuration: 30 secondsAspect ratio: 16:9Editorial target: 24fps, F000-F719 [Generation Goal]A high-speed combat battle involving only three characters: @Image1, @Image2, and @Image3. Develop offense and defense over 10 cuts. No weapons; use a combination of rushes, dodges, palm strikes, elbow strikes, kicks, throws, and aerial attacks. Do not fix the expressions of the three characters. Change them to alertness, surprise, irritation, concentration, provocation, exhilaration, and relief in response to the offense and defense. Express specifically through eyes, eyebrows, mouth, gaze, breathing, and head tilt. Return characters to a recognizable form only at moments of stopping, posing, contact, or landing. During high-speed movement, always break down the finished human body and replace it with a small number of liquid brushstrokes, rough watercolor planes, and short afterimages based on each character's color scheme. Do not simply add speed lines or motion blur to a finished figure. The flowing brushstrokes themselves are the character. [Material Roles]@Image1 = Character A. One and the same person. Uses yellow bob hair, cat ears, green eyes, light blue and white clothes, yellow trim, triangular tie decoration, pink backpack, pink boots, light blue tail with a yellow triangular tip as identifying information. @Image2 = Character B. One and the same person. Uses a yellow-green round bob, two leaf-like ears, golden eyes, white puff sleeves, pink bowtie, dark green shorts with suspenders, green boots, a long yellow-green tuft continuing from the back of the head with a round tip as identifying information. @Image3 = Character C. One and the same person. Uses vivid blue long hair and ponytail, two white horns, pointed ears, golden eyes, star-shaped earrings, dark red Japanese-style clothing, black skirt, black belt and back ribbon, black knee-highs, and red geta as identifying information. Do not treat three-view diagrams as clones. Do not swap the faces, hair, costumes, color schemes, ears, horns, tails, back hair, or decorations of the three. However, the expressions in the reference images are not fixed and should change according to the battle situation. Do not use white backgrounds or explanatory text. Do not generate weapons. Do not use the backpack, tail, back hair, horns, or geta as detached weapons. [Environment]A giant festival square based on a white paper surface. A giant vermilion pillar and watercolor stalls on the left, a sequence of vermilion torii gates in the center back, a red lantern shelf on the right, and light blue cubes and firework-like brushstrokes in the sky. Draw the background with translucent watercolors, broken ink lines, pencil composition lines, and wide white margins. Do not enclose the floor, pillars, torii, stalls, or lanterns with closed outlines. The background lines also tremble slightly, changing in density and position, but the amount of variation is weaker than that of the characters. Maintain the background layout in all cuts. Do not include combatants, crowds, or additional characters other than the three. [Unstable Line Rule]Draw the entire video with 'continuously generated lines and watercolors.' Do not clean up each frame from the previous one; redraw each frame as a new rough sketch. Do not create a single outline surrounding the person. Do not enclose the head, shoulders, arms, torso, or legs with continuous lines; always leave 30-60% of the body's outer circumference as white gaps in the paper. Recognize the character's form intermittently through short, broken pencil lines, irregular watercolor boundaries, dry brush marks, eraser marks, and doubled redraw lines. Do not maintain the same line in the same position for more than 3 frames. Lines change in position, thickness, density, length, pressure, and curvature every frame. They break in the middle, split into two, and parts disappear. Do not use a single uniform line width for the whole body. Do not keep line density constant. Irregularly alternate frames with extremely few lines, frames with dense redraw lines, and frames where most of the outer circumference disappears. Do not create a finished outer circumference for the face either. Do not draw both eyes, nose, mouth, jaw, and cheeks clearly at the same time. Leave only the minimum lines to read the expression, such as the outer side of one eye disappearing, the mouth cutting off midway, eyebrows becoming double, or cheek lines melting into watercolor. Do not fit the watercolor accurately inside the lines. Create bleeding, protrusion, white unpainted areas, uneven shading, and misregistration. [Line and Color Delay]Body movement takes the lead, pencil lines follow about 1-2 frames late, and watercolor color planes follow about another 1-2 frames late. Late lines and color planes do not completely overlap the body and remain as short afterimages torn in the direction of movement. Afterimages are not copies of the finished person but brushstrokes and color planes that indicate only the direction of the arms, legs, hair, tail, and costume. Even when stopped, do not return the lines to a clean-up state. The person's face, clothing, and posture become legible, but the gaps in the outer circumference, double lines, watercolor bleeding, and misregistration are maintained. [Mandatory Motion Deformation]Draw only about 60% of the human body for weak actions. Draw only about 30-40% for medium-speed actions. Draw only about 10-20% for high-speed actions. At maximum speed, represent the character with only 3-5 brushstrokes and 2-3 rough color planes. Character A high-speed action = yellow, light blue, white, and pink brushstroke group. Character B high-speed action = yellow-green, dark green, white, and a small amount of pink brushstroke group. Character C high-speed action = blue, dark red, black, and a small amount of golden brushstroke group. During high-speed action, do not connect the head, torso, arms, and legs with a correct human structure. Turn arms into long liquid brushstrokes, kicking legs into extremely elongated color planes, and rotating torsos into crushed arcs, triangles, or twisted watercolor surfaces. Temporarily disconnect limbs during jumps. Compress the torso in the direction of travel during sudden stops. Shift the hips and shoulders in different directions during turns. [Expression Direction]Character A: Starts with alertness, changes to a proud smile upon successful evasion, surprise when an attack is diverted, playful provocation during a melee, strong concentration in the late stages, and finally a satisfied, breathless smile. Character B: Starts with cautious observation, sharp concentration when interrupting, momentary surprise when an attack is dodged, a competitive smile during aerial attacks, a desperate expression in the final stages, and finally a soft, relieved expression. Character C: Starts with calm confidence, smiles slightly after dodging the first attack, furrows eyebrows during a pincer attack, sharp concentration while counterattacking, an exhilarated smile during the three-way melee, and finally a provocative one-sided smile. Do not finish the face during high-speed action. Indicate emotions with a single stroke of the eye, the tilt of an eyebrow, a short line of the mouth, or the direction of head brushstrokes. Make expressions legible only at moments when speed drops. [Cut 1 | 0.0-2.5s | F000-F059 | Triangular Confrontation]Initial state: Fixed wide. Character A is in the left foreground, Character B is in the right foreground, and Character C is in the center back. They form a triangle with each other in sight. Primary event: Show the appearance and positions of the three from F000-F023. A narrows eyes and is wary of C, B looks back and forth between A and C and breathes cautiously. C raises their chin slightly and shows composure with a one-sided smirk. From F024-F041, A lowers their hips, B opens half a step to the side, and C slides their geta to lower their center of gravity. From F042-F059, the three step out simultaneously. End state: The travel directions of the three converge in the center. The expressions and body circumferences are legible, but the lines are not closed and are redrawn every frame. [Cut 2 | 2.5-5.0s | F060-F119 | A vs C First Strike]Initial state: Low angle near the ground. A on the left, C in the center, B in the right back. Primary event: From F060-F075, A breaks down into yellow, light blue, and pink brushstrokes and rushes low from left to right. From F076-F087, A connects a leg sweep to a spinning kick. C breaks down into blue, dark red, and black brushstrokes, rotates, and evades the kick by a hair's breadth. Only just before contact from F088-F095 do A's eyes open wide and C smiles slightly. Do not draw the whole face, show only through fragments of eyes, eyebrows, and mouth. From F096-F119, A passes by C and slides to the right, leaving fragments of a surprised face. End state: A is in the right foreground, C is in the center, B is in the right back. C's composure increases, and A refocuses their expression. [Cut 3 | 5.0-8.0s | F120-F191 | B's Interruption]Initial state: Medium wide horizontal position. C in the center, A on the right, B in the right back. Primary event: From F120-F137, B lowers their eyebrows, breaks down into yellow-green and dark green brushstrokes, and enters from right to left. From F138-F153, B performs a spinning leg sweep passing under both of their feet. From F154-F167, A jumps vertically, and their legs and tail stretch into watercolor bands. The moment the evasion is successful, the corners of A's eyes droop and a smirk forms provocatively. C sinks low, folding their body to dodge B's kick. From F168-F179, B's eyes open wide for a moment, showing surprise at being dodged. From F180-F191, B furrows their eyebrows and concentrates on the next attack while sliding left. End state: B lands on the left, C in the center, and A in the center back. [Cut 4 | 8.0-11.0s | F192-F263 | C's Continuous Counterattack]Initial state: Low diagonal composition. B on the left, C in the center, A in the back. Primary event: From F192-F207, C's smile disappears and eyebrows drop sharply. C breaks down into 3-5 blue, dark red, and black brushstrokes and accelerates rapidly from left to right. From F208-F223, C performs a palm strike on B, an elbow strike on A, and a spinning back kick in one continuous motion. During the attack, C does not return to a finished body. Only a single golden stroke for the eye and a concentrated eyebrow line appear within the brushstroke group. From F224-F239, B evades backwards as a green arc, and A escapes sideways by splitting into short yellow and pink lines. From F240-F251, the arms and legs of the three cross for just a moment. A opens their mouth briefly in surprise, and B grits their teeth. From F252-F263, C passes to the right side and shows an exhilarated one-sided smile. End state: C is on the right, B on the left, A in the center back. All three are standing. [Cut 5 | 11.0-14.5s | F264-F347 | A and B Pincer Attack]Initial state: High diagonal overhead view. C on the right, A in the center back, B on the left. Primary event: From F264-F283, A and B exchange a short look. A shows a confident smile, B shows a focused face with lowered eyebrows. The two run towards C simultaneously; A performs a low flying kick and B performs a spinning kick. From F284-F303, C furrows their eyebrows and changes their body into thin blue, red, and black lines to slip through the gap between the two kicks. From F304-F323, A and B open their eyes wide just before colliding with each other, tuck their legs, and change direction in mid-air. From F324-F347, the three move apart in a circle. C gives a small smile for breaking the pincer attack, A furrows their eyebrows regretfully, and B exhales and gauges the next distance. End state: B on the left, C in the center, A on the right. [Cut 6 | 14.5-18.0s | F348-F431 | Sequence of Grabs and Throws]Initial state: Circling tracking shot. B on the left, C in the center, A on the right. Primary event: From F348-F367, A smiles provocatively and tries to jump onto C's shoulder. C changes to a focused, almost expressionless face, catches A's wrist to change their direction of motion, and throws them low. Make only the contact point briefly legible. From F368-F387, A becomes a group of yellow and pink brushstrokes and falls while rotating. Surprised eyes and an open mouth appear for a moment, then immediately break down into lines. From F388-F407, B throws a heel kick at C while passing in mid-air. B narrows their eyes and shows a competitive smile. From F408-F419, C catches it with their forearm and flows B's leg line outward. C's eyebrows rise slightly, showing surprise at the unexpected strength. From F420-F431, A regains posture just above the floor, clenches their mouth regretfully, and performs a three-point landing. End state: A is in the right foreground, B is in the left foreground, and C is in the center back. [Cut 7 | 18.0-21.5s | F432-F515 | 3D Offense and Defense Using the Background]Initial state: Wide shot showing the left vermilion pillar, center torii, and right lantern shelf simultaneously. Primary event: From F432-F451, B runs to the left pillar. Their expression changes to an exhilarated smile, and they step on the pillar once to bounce off diagonally. From F452-F471, A jumps to the right stall roof, kicks the roof, and reverses to the center. A shakes off the regret and narrows their eyes joyfully. From F472-F491, C anticipates the trajectories of the two and stays out of both attack lines with a low rotation while keeping a sharp gaze. During the high-speed section, the three become brushstroke groups with their own independent color schemes. From F492-F503, only the lanterns and stall cloths shake from the wind pressure. From F504-F515, the three land in the center and open their eyes wide at each other's unexpected movements. Immediately after, all three laugh. End state: The three form a narrow triangle in the central square. Exhilaration for the battle appears in the three. [Cut 8 | 21.5-25.0s | F516-F599 | Six-Beat Close-Quarters Melee]Initial state: Waist-high close-up wide. The whole bodies and feet of the three are visible. Primary event: From F516-F527, A starts from a joyful smile and enters a right roundhouse kick. From F528-F539, C lowers their eyebrows and flows it outward with their forearm. From F540-F551, B clenches their mouth and delivers a low leg sweep. From F552-F563, C opens their eyes in surprise, tucks their knees, and jumps. From F564-F575, A changes to focused eyes and drops a heel from mid-air. From F576-F587, B grits their teeth, pushes A's leg away, and changes the attack direction toward C. From F588-F599, C shows an exhilarated smile and slips behind the two. Replace characters with brushstroke groups for each high-speed action. Recombine only for 1-2 frames for attack blocks and expression changes. End state: A is on the left, B on the right, C in the center back. All three are out of breath, but their fighting spirit has grown stronger. [Cut 9 | 25.0-28.0s | F600-F671 | Simultaneous Three-Way Rush]Initial state: Frontal wide. A on the left, B on the right, C in the center back. Primary event: From F600-F611, the three sink down simultaneously. A removes their smile and looks at the landing point. B furrows their eyebrows strongly and holds their breath. C raises only one side of their mouth, provoking the two. From F612-F635, A (yellow, light blue, pink), B (yellow-green, dark green, white), and C (blue, dark red, black) become minimal brushstroke groups and rush toward the center. Do not show finished faces, hands, torsos, or costumes. From F636-F647, the three brushstroke groups cross in the center. Eyes, eyebrows, and mouths appear briefly as independent short lines, showing the concentration of the three. From F648-F659, opaque watercolors, ink splatters, and dry brush strokes that scrape the paper scatter radially. From F660-F671, the three pass through in the directions they crossed. End state: A, B, and C cross the center and decelerate while facing away from each other. [Cut 10 | 28.0-30.0s | F672-F719 | Recombination and Conclusion]Initial state: Fixed wide. The three are sliding in different directions from the center. Primary event: From F672-F687, brushstrokes shorten and characters return in the order of posture, hair, costume, and watercolor. The body's outer circumference does not close. From F688-F703, A stops on the right, B on the left, and C in the center foreground. A smiles satisfiedly while out of breath. B closes their eyes once, exhales, and smiles softly. C looks at the two over their shoulder and shows a provocative one-sided smile. From F704-F711, the three turn around simultaneously. From F712-F719, they change their expressions back to focused faces and take low combat stances. End state: No conclusion reached. All three can be identified as the same characters from the reference images by their color scheme, hair, and costume. However, line gaps, double lines, watercolor bleeding, and misregistration remain after stopping. [Visual Treatment]Hand-drawn 2D animation on bright white paper. Use ink, pencil, watercolor, dry brush, white unpainted areas, eraser marks, and redraw lines. Do not create a line that goes all the way around the person's outer circumference. Do not enclose the face, hair, clothes, arms, or legs as closed shapes. Lines within the screen change in thickness, density, length, position, pressure, curvature, and gap rate every frame. Do not correct high-speed action body collapse, line gaps, facial omissions, or color plane delays as generation defects. Do not return to smooth vector lines, uniform line width, closed outlines, symmetrical faces, fixed expressions, accurate coloring, 3D-like human interpolation, or processing that only adds effect lines to a finished character. [Audio]High-tempo percussion-based BGM. Character A has light boot sounds, B has soft landing sounds, C has hard geta sounds. Clothing wind noise, short hitting sounds, floor scraping sounds, sounds of kicking pillars and stalls, the sound of lanterns swaying, and breathing. No dialogue. [Maintain Consistency]Only three characters. Do not generate three-view diagrams as clones. Do not swap the color schemes, faces, hair, costumes, ears, horns, tails, back hair, backpacks, or shoes of the three. Do not fix expressions. Change eyes, eyebrows, mouth, gaze, and breathing in response to each attack, dodge, failure, success, surprise, and exhilaration. Do not stabilize lines. Do not make line width constant. Do not maintain the same line in the same position. Do not close the body's outer circumference. Do not fit watercolor accurately inside the lines. Always break down characters during high-speed action. Do not let them run, jump, or attack while maintaining a finished human body. Move the brushstroke groups themselves as the characters. Make them identifiable as the same person at the moment of stopping and contact, but do not return to a clean-up outline. Do not generate weapons. Do not generate additional people, clones, or human-shaped afterimages. Maintain background layout, character movement directions, and triangular positional relationships. Frame numbers are editorial goals; prioritize event order, expression changes, and the end state of each cut. pure video, no subtitles, no dialogue, no watermark.

High-quality anime footage. Movie-theater-class high-density 2D hand-drawn action anime. High-speed melee combat. Set in a dark near-future giant corridor, the protagonist continuously breaks through multiple enemies without stopping. Most important are the overwhelming speed, grounding, chain of reaction forces, clear contact, extremely changing camera distances, short 'face rewards' (clear face shots), and instantaneous graphic impacts in red, black, and white in the latter half. Do not create segments of standing still to re-pose. Use the reaction of each attack as the start of the next movement or attack. [Protagonist Fixed] Maintain the protagonist from the source reference image as the same person in all cuts. Do not change face shape, age feel, eye size and spacing, iris color, jaw, bangs, hairstyle, hair color, outfit, body type, or unique color scheme. Even during high-speed movement, distant views, or backlit explosions, do not allow the character to become a different person, look older, or become simplified. The moment the shot returns to a close-up, accurately revert to the same face, eyes, and hair as the source reference image. The protagonist's weapon is only one slender single-edged sword. Maintain the blade, guard, and hilt as a single continuous structure. Do not multiply, branch, or exchange weapons. Hands always grip only the hilt, and do not overlap the blade and hands in front of the face. [Enemies Fixed] Four humanoid enemy soldiers wearing black near-future armor. All four maintain a similar armor design and do not mix faces, hair, outfits, or weapons with the protagonist. Enemies maintain humanoid bodies, black armor, and limbs even during combat. After being hit, they do not transform into crystals, abstract objects, or clumps of light. [Art Style Fixed] Entirely 2D hand-drawn anime. Thin, delicate line art, dark but not overly black outlines, two-to-three-tier anime shading, transparent mid-shades, multilayered highlights on hair and eyes, high-density background art, and exaggerated key animation for theater-quality action. Use short smear frames with boldly stretched shapes during high-speed action as needed, but return the face to clarity during face reward frames. Do not use thick black outlines, simplified TV-style animation, low-density backgrounds, flat single-layer cel shading, smooth 3DCG, live-action style, semi-realism, dull colors, or mixed art styles. [Setting/Color/Light] The setting is a dark, massive near-future corridor. Maintain the same location throughout. Use black, dark navy, and dark blue-green as basic background colors, with thin red linear lighting on walls, floors, and ceilings. Red light lines point toward the vanishing point at the end of the corridor, emphasizing speed direction and depth. Fix color roles: Cyan-Blue = Protagonist's acceleration and electrical afterglow; Lime-Yellow Green = Sword attack trails; Orange-Yellow = Sparks and explosions after contact; Red-Black-White = Graphic impact used only for the final shock in the latter half. Do not exchange color roles between cuts. Keep skin tones and protagonist-specific colors stable, avoiding random white balance changes. [Common Movement Rules] Do not process running as a parallel slide of the body. Sequence: Grounding foot fixed to floor -> Pelvis passes over supporting leg -> Opposite knee moves forward -> Toes of supporting leg kick the floor -> Short aerial phase -> Opposite foot grounds forward -> Knee compresses -> Next kick. The pelvis and ribcage themselves must advance across floor joints. Even if the camera follows, do not eliminate the protagonist's movement amount in world coordinates. Landing must always pass briefly through: Foot grounding -> Knee compression -> Hips sinking -> Upper body following. Sword contact reaction force should not be absorbed only by the wrist; flow it through: Sword -> Wrist -> Elbow -> Shoulder -> Ribcage -> Hips -> Supporting leg, passing that reaction and hip rotation into the next movement without stopping. Hair, clothing hems, and decorations follow the body with a slight delay. [Contact Rules] All important attacks follow this order: Clear gap between sword and enemy before contact -> Sword contacts specified position -> Very short hit-stop only at the moment of contact -> Sparks, armor displacement, enemy center-of-gravity change, and explosion only after contact. Do not blow enemies away before contact. Do not generate explosions before contact. Do not create static screens showing only effects; continue character movement while sparks or explosions are visible. [Full 10 Cuts Breakdown] [Cut 1 | Floor Ignition] Starts with an ultra-low-angle extreme close-up of the protagonist's feet. The front foot grounds strongly. The sole catches the floor, and the knee compresses for a split second. The moment the toes kick off, thin lime-green friction sparks fly backward. Cyan electrical afterglow runs briefly from ankle to leg. Camera runs parallel to the floor with the protagonist, showing grounding, kicking, and the pelvis advancing. The protagonist doesn't stop, converting the kick into high-speed acceleration. Ends with upper body leaning forward entering the next stride. [Cut 2 | Corridor Acceleration] Inherits the forward lean and speed. Protagonist advances into the corridor using alternating high-speed strides. Floor joints and red wall lights blur backward. Camera follows from a low front position, retreating at the same speed. Strong wide-angle perspective shows face, chest, knees, legs, and the corridor vanishing point. Orange explosion occurs far behind, but protagonist doesn't look back. Explosion pressure pushes hair and hems forward momentarily. Just before the next enemy appears, protagonist grounds the right foot and drops hips low for a slash. [Cut 3 | Face Reward to First Slash] Inherits right-foot support and forward speed. Camera catches up at high speed, zooming in momentarily from chest to face. Face and eyes are clear, but movement doesn't stop. Protagonist glares at the enemy, smirking slightly and exhaling a short breath. Rotates hips and swings the sword in a single path from bottom-diagonal to top-diagonal. Clear gap before contact. Short hit-stop when blade hits enemy torso armor. Thick lime slash trail and orange sparks cross the screen diagonally after contact. Enemy's upper body buckles; protagonist converts reaction into leftward movement. Lime blade light crosses the lens, providing a transition to the next camera axis. [Cut 4 | Explosion Mask and Axis Change] Starts from the blade light mask. Camera is now at a diagonal side medium shot. As the first enemy collapses, they hit background equipment causing a short orange explosion. Smoke partially covers the screen but doesn't hide the protagonist for long. Using the hip rotation, protagonist steps laterally to change course toward the second enemy. Camera cuts across the smoke as a foreground element to discover the new distance. No stopping or re-posing. [Cut 5 | Continuous Melee] Protagonist grounds left foot forward and steps into the second enemy. Parries enemy weapon once. Gap before contact. Tiny metal sparks at contact. Enemy weapon is deflected, protagonist's shoulders and hips rotate in reverse. Protagonist uses that recoil to transition into a horizontal slash from the same rotation. Horizontal slash makes single contact with torso. After hit-stop, enemy buckles sideways. Side-tracking camera shows enemy weapon, protagonist's full sword, supporting leg, and contact point in one frame. Protagonist maintains forward direction after slash, passing the enemy. [Cut 6 | Space Opening] Inherits momentum from the previous cut. A third enemy attacks from the front. Protagonist grounds one foot and converts compression into a vertical and forward leap. Camera switches to a wide long shot, making character size significantly smaller. Shows protagonist, 3 enemies, red lighting corridor, and vanishing point to maximize scale. Protagonist tilts body axis once in mid-air to clear the attack line. No excessive mid-air spins. Descends while keeping the landing point in view. [Cut 7 | Landing Reaction to Re-acceleration] Protagonist lands. Sequence: Grounding -> Compression -> Hips sinking -> Upper body follow. Absorbs reaction without stopping, sinking hips directly into the next forward kick. Cyan afterglow runs back from landing point. Camera slides low to the protagonist's side. Before fully standing, protagonist accelerates toward the third enemy. Camera briefly parallels the face to show eyes and profile. No long face rewards. [Cut 8 | Third Contact and Maximum Pressure] Protagonist maintains acceleration. Minimally dodges enemy attack and swings sword horizontally. Gap before contact. Hit-stop at contact. Lime trail, orange sparks, and armor displacement occur simultaneously. Enemy's center of gravity shifts significantly. Protagonist absorbs reaction into hips to twist toward the next direction. Camera zooms extremely close to the protagonist's eye. The moment the eye occupies the main screen area, transition to graphic insert. [Cut 9 | Graphic Impact Intercut] Do not delete reality for long. Extreme close-up of eye -> flashes to solid red -> extremely short 2D graphic frame composed of black background, white slash plane, and red speed lines -> sword trajectory abstracted as a single white directional line -> immediate return to reality. This is an editorial accent for impact, not a transformation of space. No red ribbons or decorative beams. Upon return, the sword trajectory from Cut 8 continues into the attack on the fourth enemy. [Cut 10 | Maximum Impact / No-Stop Termination] Protagonist approaches the fourth enemy at continuous speed. No stop before the final attack. Grounds left foot to compress hips, pushing off to swing sword in a large diagonal slash. Clear gap before contact. Strongest hit-stop at contact. After contact, large lime slash trail, orange sparks, and cyan afterglow expand from a single point. Insert red-black-white graphic frames at the pressure point. Enemy is blown off balance. Protagonist flows the reaction into hips and supporting leg, converting it into the next forward acceleration. Camera passes high-speed next to protagonist during impact. Final frame: protagonist is not static. No sword-raising. No victory pose. No looking at the camera. No quiet pull-back shot. Cut while protagonist is kicking off for the next step, hair and clothes flowing back, mid-movement. This vector naturally loops back to Cut 1. [Sound] High speed but not constant max volume. Sync short 'huff/ha' breaths with strong steps. Hard floor sounds for grounding. Short metallic 'whoosh' for sword movement. Short metallic clashing for weapon contact. Metallic impact and low thud for hitting armor. Explosions sync with contact results. In graphic inserts, compress ambient sound momentarily to bring sharp slash sound to the front. No long screaming, dialogue, or excessive roaring. [Important Guards] Never return to neutral stance after attacking. Always convert reaction, dodge, compression, or rotation into the next move's propulsion. Do not just move the camera to simulate speed; advance the character relative to the floor. Change at least 2 camera parameters (position, height, size, axis) between adjacent cuts, but do not reset movement direction for camera changes. Create distance amplitude: Extreme close-up feet -> Front wide -> Face close-up -> Side medium -> Low angle contact -> Ultra-wide long -> Profile close-up -> Eye extreme close-up -> Red/Black/White Graphic -> Real-time high-speed melee. Do not express speed through effect volume alone. Ensure at least one key point (supporting leg, trajectory, contact, reaction) is readable. Do not overlap effects over the face during face rewards. No text, subtitles, logos, or watermarks. No extra protagonists, clones, or weapon multiplication. No sudden enemy additions. No enemy crystallization. No long slow-motion. No long glares or weapon-clashing. No static victory poses or sword-raising. No final wide pull-away shot.

[REFERENCE CONTROL] Use the uploaded storyboard image as the primary visual reference for story structure, character design, costume design, environment, emotional progression, and shot order.

use the attached storyboard <<<image_1>>> & A warm, wholesome 15-second animated short in cozy storybook watercolor style with soft ink linework, painterly textures, and gentle film grain.

Create a 30-second cinematic biographical video. The entire process must be a completely seamless, coherent, and uninterrupted long shot. There should be no cuts, fades, dissolves, resets, or montage effects. The camera can orbit, track, dive, rise, rotate, and pass through foreground obstacles, or temporarily obscure the character, but the final result must always appear as an uninterrupted moving shot. **Core Concept** A man's entire life, from youth to old age, condensed into an uninterrupted long shot. This man is Su Shi, the greatest literatus of the Northern Song Dynasty. He remains the same person, a continuous presence throughout the film, but his age, face, physique, hairstyle, clothing, expression, and temperament must change fluidly as he progresses through life stages. Every change must occur within the frame, tied closely to his movement with physical causality. His evolution must be seamlessly connected: from a youth in Meishan to a scholar taking exams, then a talented young official entering the bureaucracy, followed by an outspoken official opposing reforms, then a prisoner in jail, a free-spirited poet in the fields of Huangzhou, an Academician returning to court, a local governor of Hangzhou, an old man exiled to Lingnan, and finally a wise man dying with a smile on his way back north. His face, body, beard, official robes or plain clothes, hats, and handheld objects must transition naturally with life stages. These changes must happen during motion and never as abrupt jumps. **Tone** Poetic, magnificent, soulful, open-minded, and filled with literary charm. The environment can be spectacular and dramatic, but the overall atmosphere always carries Su Shi's characteristic openness and warmth. Suitable for all ages, with no blood or violence. **Audio Design (Seedance 2.5 Voice Format Instructions)** This video uses Seedance 2.5 native audio generation, where visuals, voice, and sound effects are completed synchronously in one generation. The voice format rules are as follows: - **Voiceover Narration** (Classical male voice recitation): Enclosed in double quotes "...", preceded by "Narrator recites in a steady classical middle-aged male voice:" - **Environmental Sound Effects**: Marked with angle brackets <...>, e.g., <Wind rustling through bamboo forest> - **Background Music**: Marked with parentheses (...), e.g., (Distant guzheng melody, slow and deep) - **Mixing Priority**: Voiceover narration is clear and prominent, background music is lower than the voice, and environmental sound effects serve as the ambient layer. Due to the limited voice capacity within 30 seconds (about 40-50 Chinese characters), please select 6 of the most representative poems or famous lines as voiceovers, while the rest of the scenes use sound effects and music to carry emotion. **Narrative Timeline** The 6 lines not only fit the current scenes but also form a complete spiritual arc of life: 1. "Where can life be compared? It should be like a flying swan treading on snow and mud." → **Departure**: Life is like a flying swan, unknown destination. 2. "Picking all cold branches but unwilling to roost, lonely and cold on the sandbank." → **Fall**: Nowhere to stay, extreme loneliness. 3. "Bamboo cane and straw shoes are lighter than a horse. Who's afraid? A life of misty rain in a coir raincoat." → **Awakening**: Letting go of everything, no longer afraid. 4. "Comparing West Lake to Xi Shi, light makeup or heavy is always suitable." → **Settling**: Reconciling with the world, loving the present. 5. "Eating three hundred lychees a day, I wouldn't mind being a Lingnan man forever." → **Open-mindedness**: Living flavorfully no matter where. 6. "Where this heart finds peace is my hometown." → **Return**: Having traveled the world, peace of heart is home. Hearing them consecutively summarizes Su Shi's life spirit: Unknown destination → Nowhere to stay → No longer afraid → Loving the present → Living well anywhere → Peace is home. --- ## Opening: Meishan Youth (c. 1050, Su Shi aged 13-14) Opens with early morning in Meishan, Sichuan. A youth stands on a bamboo-shaded slope, overlooking the misty Minjiang Plain. Distant layered green mountains, nearby the eaves of his father's study. Morning light spills through bamboo leaves onto his face. He holds a scroll, eyes bright and curious. The youth has handsome features, wearing simple Sichuan student robes, hair tied in buns. The visual should immediately feel poetic and cinematic. He hears a distant horn sound (symbolizing the call of the imperial exams), looks up, eyes flashing. <Wind rustling bamboo forest, faint distant horn> The camera moves as he turns and starts running down the bamboo path. --- ## First Transition: Meishan Youth → Scholar Heading for Exams (1056, age 20) He runs through the bamboo forest, shadows sweeping over him. As the shadows pass, his face matures slightly, his stature grows, and the buns change into a young scholar's hairstyle. Simple student robes turn into a traveler's long robe. The scroll becomes a bundle. The bamboo forest dissolves behind him, the mountain path becomes a government road to Bianjing. City outlines appear. His pace is steady, with the high spirits of a youth leaving Sichuan for the first time. The figure of his brother Su Zhe appears beside him (blurred as a background element). Narrator recites in a steady classical middle-aged male voice: "Where can life be compared? It should be like a flying swan treading on snow and mud." <Horse hooves, wind on the government road> --- ## Second Transition: Scholar → Young Talent on the Golden List (1057, age 21) An imperial banner fluttering in the wind sweeps across the frame. As it passes his body, his traveler's robe turns into the blue official robe of a new jinshi. The bundle vanishes, replaced by a scroll of imperial decree. His gait becomes confident and high-spirited. The government road becomes the stone-paved main street of the Bianjing imperial city. Towering palace buildings, red walls, yellow tiles, and bustling crowds appear on both sides. He walks through the crowd with a triumphant smile. Sunlight hits his young face, full of spirit. <Crowd noise, distant celebratory music> --- ## Third Transition: Young Talent → Outspoken Official (1069-1071, age approx. 33) A vermilion palace pillar sweeps across the foreground, temporarily obscuring him. When he reappears from behind the pillar, his face is more mature, with a neatly trimmed short beard. The color and patterns of the official robe become deeper, and he wears a black official hat. His expression turns from triumphant to serious and grave. He is walking briskly through a court corridor, clutching a memorial. The environment becomes oppressive, the pillar's shadow elongating. Notices of Wang Anshi's reforms and arguing officials are faintly visible in the distance. His pace is urgent and steady, with the determination of one speaking truth to power. <Urgent footsteps echoing in the empty corridor> --- ## Fourth Transition: Outspoken Official → Prisoner in the Crow Terrace Poetry Affair (1079, age 42) The light in the corridor suddenly dims. The shadow of the pillar becomes the iron bars of a prison. His official robe loses color in the shadows, turning into a prisoner's coarse cloth. The black hat vanishes, hair becomes disheveled. The memorial in his hand turns into heavy chains. His face is haggard, short beard messy, eyes sunken. He stumbles through a dim cell, but his gaze still holds unyielding light. There are water stains and carvings on the walls. A weak beam of light shoots from a high window onto his face. Narrator recites in a steady classical middle-aged male voice: "Picking all cold branches but unwilling to roost, lonely and cold on the sandbank." <Heavy chains dragging on the ground, water dripping echoes> The camera rises with that beam of light. --- ## Fifth Transition: Prisoner → Open-minded Poet in Huangzhou (1080-1084, age approx. 45) That light expands and warms, becoming the moonlight at Chibi in Huangzhou. The prison walls shatter and dissolve, becoming the Chibi cliffs by the Yangtze River. The prison floor under his feet turns into riverbank rocks. The prison clothes turn into loose literary plain robes in the moonlight, a coarse belt tied at the waist. Chains vanish, a jug of wine appears in his hand. His hair is tied back again, but not in an official bun—instead a casual literati style with a Dongpo hat. His face relaxes, though thin, his eyes have a transcendent light. He stands under the Chibi cliffs, the river wind blowing his robes. A small boat is in the distance on the river. He looks up at the moon, lips moving slightly as if reciting "The Yangtze River flows east." The entire person is completely liberated from the oppression of imprisonment, exuding a transcendent air. Narrator recites in a steady classical middle-aged male voice: "Bamboo cane and straw shoes are lighter than a horse. Who's afraid? A life of misty rain in a coir raincoat." <River wind howling, waves hitting rocks> He turns and walks along the riverbank with a leisurely pace.

I. Core Positioning Cinematic realistic texture, pure ancient Chinese xianxia aesthetics. Styled as an elegant martial arts duel with deadpan reactions, spatial visual comedy, and dry logic. Features Arri Alexa film look, micro-facial details, fine grain, and natural volumetric lighting. Comedy structure: Misleading -> Escalation -> Reversal. The humor stems from the Senior Sister promising 'not to intervene' and physically keeping her hands off while providing non-stop verbal guidance. II. Reference Control Character Anchors: @Image 1 for Senior Sister (Role A). @Image 2 for Junior Sister (Role B). Environment: Shared DNA across three shots with consistent geography, terrain, and weather. III-IV. Characters and Environment Senior Sister (Role A): 25-30 years old, tall, fair skin, white silk hanfu, silver sword. Calm, serious, and logically dry. Junior Sister (Role B): 20-25 years old, petite, celadon hanfu, dark steel sword. Accustomed to her sister's style. Enemy Swordsman: A 'normal person' who becomes increasingly frustrated. Master: An authoritative figure watching from afar. V. Core Narrative Logic 1st Beat (Misleading): Enemy demands a one-on-one. Sister retires and promises not to intervene. 2nd Beat (Escalation): During the duel, Sister keeps her hands behind her back but gives precise instructions ('Shoulders down', 'Left side'). 3rd Beat (Reversal): Enemy snaps, 'This is not intervention?' Sister replies, 'Hands didn't move.' Master confirms, 'By the rules, hands stayed still.' Junior Sister adds, 'She mainly uses her mouth.' VI. Storyboard Script 0-5s | Wide Shot: Enemy shouts 'One at a time!' Senior Sister sheaths her sword and backs away: 'Fine, I won't intervene.' Junior Sister steps forward. 5-10s | Medium Shot: Duel starts. Senior Sister stands aside with hands behind her back, coaching: 'Settle your shoulders... Steady feet... Left side.' Junior Sister yells: 'Sister, I got this!' Enemy stops and shouts at the Sister: 'Is this not intervention?' Sister looks at her hands: 'Hands didn't move.' 10-15s | Close-up: Enemy looks to the Master for justice. Master nods solemnly: 'By the rules, indeed no hands were used.' Enemy is speechless. Junior Sister adds: 'She mainly moves her mouth.' Senior Sister starts to correct her, but Junior Sister points a finger: 'Don't guide this line.' Sister freezes with a tiny eyebrow twitch. Enemy drops his sword in defeat. VII-VIII. Performance & Action Senior Sister must remain serious; the humor comes from her complete lack of irony. Junior Sister is利落 and familiar with the routine. Enemy is the audience surrogate for disbelief. Master provides the absurdly serious ruling. Actions must be clean and readable, especially the hands-behind-back pose and the three sword exchanges.

I. Project Positioning Movie-grade realistic texture, pure ancient style Chinese Xianxia aesthetics. The overall style focuses on restrained and mature wuxia emotional scenes, emphasizing: Measured subtext in character relationships Elegant and clear martial arts blocking Arri Alexa cinematic texture Clear and stable facial micro-details Fine film grain Natural volumetric light The core reversal of this segment centers on "trust": Initially, everyone interprets the Sword Immortal Senior Sister's inaction as coldness. Finally, it is discovered that her true protection is not preventing the Junior Sister from getting hurt, but preserving her right to choose, to bear, and to prove herself. Narrative Principles: The emotional center always belongs only to the two women in @Image1 and @Image2 Supporting characters are only responsible for creating external pressure and social evaluation Supporting characters cannot replace the two protagonists as the emotional center The environment must be vivid but absolutely neutral in the narrative II. Reference Control 1. Character Identity Anchors @Image1 strictly serves as the identity anchor for Character ID A, Sword Immortal Senior Sister @Image2 strictly serves as the identity anchor for Character ID B, Junior Sister 2. Environment Reference All background and location reference images uploaded in this round together determine a single set of environment DNA. Before formal composition, silently deduce compatible elements: Real terrain Architectural language Spatial scale Material era Vegetation ecology Water system direction Weather Clouds Mountain mist Main light direction Reflective relations Overall color Aerial depth Real traffic flow lines Then replan into a unique, unified new space for this round. Requirements: Cannot mechanically copy any single reference image Maintain identity within the same world Three shots share a consistent geographical logic Clear layers of foreground, middle ground, and background The world continues to run naturally, but plot progression can only come from characters III. Environment Principles The background is always vivid but maintains absolute neutrality in the narrative. If the reference images contain the following elements, they should run naturally: Water flows continuously Mountain mist drifts naturally according to terrain Vegetation responds to natural wind Clouds move slowly Reflections change continuously Distant spatial ambient sounds truly exist Hard requirements: The environment cannot create plot events The environment cannot solve character problems The environment cannot change character goals The environment cannot create setups for jokes The environment only provides a sense of existence in the real world IV. Technical Orientation Organized overall according to Seedance 2.0 capabilities, focusing on: Continuity of character identity through image reference Native synchronization of audio and lip-sync Clearly readable martial arts movements Stable positioning of supporting characters Stable geographical space Director-level control over performance and camera 15-second high-quality multi-shot audio/video output capability V. Character Settings Character ID A | Sword Immortal Senior Sister The same @Image1 Sword Immortal Senior Sister. 25–30 year old East Asian female, oval face, fair skin, dark almond eyes, long black hair half-up, fixed with a white jade hairpin, tall and slender. Fixed styling: White embroidered silk Hanfu Translucent wide sleeves Silver waist seal Jade pendant White cloth boots A single silver longsword Character temperament: Calm Mature Measured Rarely explains herself The true way of protection is not acting on behalf, but respecting the other party as a swordsman completing their choice independently Character ID B | Junior Sister The same @Image2 Junior Sister. 20–25 year old East Asian female, round and agile face, black braided hair, small stature. Fixed styling: Cyan-green linen Hanfu Dark belt Wooden hairpin Black cloth shoes A single dark steel sword Character temperament: Intelligent Resilient Not fragile Will briefly misunderstand between external pressure and the senior sister's silence But when truly trusted, can immediately stand firm Supporting Character Setup Elderly Master One person Distant secondary witness Represents the sect elder's perspective Has concerns about the Junior Sister's injuries and risks Doesn't steal the scene, only provides social pressure Sect Disciples Two people Distant secondary witnesses Exist only as secondary figures in the environment Do not participate in the main emotional line Enemy Swordsman One person Fights with the Junior Sister Only responsible for creating external pressure and the evaluation of the "cold-hearted senior sister" Does not become the true emotional center VI. Core Narrative Structure The core of this segment is not "who is strong or weak", but "what true trust looks like". The three-beat structure is as follows: First Beat The outside world misinterprets the senior sister's coldness. The master asks why she doesn't stop it; the small enemy calls her heartless. Second Beat The Junior Sister challenges on her own and completes the outcome. The senior sister never acts on her behalf, only uses minimal action to prevent others from intervening. Third Beat The senior sister admits she is "afraid," but what she fears more is the Junior Sister waiting for others to decide every step for her from then on. A single sentence redefines the true meaning of "not acting." VII. Storyboard Script 0–5s | Full Shot or Long Shot In the sect sword-testing area replanned according to this round's reference images, the same cyan-clothed Junior Sister is facing an enemy swordsman. An elderly master and two sect disciples exist only as distant secondary witnesses. Character ID A is always the same @Image1 Sword Immortal Senior Sister. Character ID B is always the same @Image2 Junior Sister. The master lowers his voice and asks the same white-clad Sword Immortal: "Her injury isn't healed yet, you really won't stop her?" The Sword Immortal Senior Sister answers calmly: "It's her own choice." The enemy swordsman sneers: "What a heartless senior sister." Camera requirements: Clearly establish the sword-testing space and character positioning The Senior Sister never steals the focus; she is an observer and judge The Junior Sister is at the front of the combat line The master and disciples are distant but distinguishable The environment continues its natural movement but never interferes with the plot 5–10s | Medium Shot or Cowboy Shot Maintain the same cyan-clothed Junior Sister, the same enemy swordsman, the same white-clad Sword Immortal, and an identical geographical space. The enemy completes only one clear downward sword strike. The Junior Sister raises her sword to block, her body pressed back a step, but she doesn't wait for anyone to rescue her; she instead actively readjusts her foot positioning and uses a clean counter-attack to parry the enemy's blade away. The master subconsciously takes half a step forward. The same Sword Immortal Senior Sister only calmly raises one hand to stop the master, yet never draws her own longsword from start to finish. The Junior Sister sees this in her peripheral vision, stabilizes her breathing again, and then parries the enemy's weapon through a precise wrist-rotating sword movement. Camera requirements: The fight must be simple, clear, real, and readable The focus is not on flashy techniques but on the Junior Sister completing this trial herself The senior sister's movement to stop the master should be light but very definite Her not acting does not mean she is not present After the Junior Sister sees this, her breath and confidence stabilize again 10–15s | Close-up or Extreme Close-up The defeated enemy remains in the background in a shallow depth of field, blurred. The same Junior Sister's breathing is slightly rapid; she turns to look at the same Sword Immortal Senior Sister and asks: "Are you really not afraid of me losing?" The same white-clad woman answers without hesitation: "Afraid." Maintain an emotional weighted half-beat silence. Then she steps closer and personally readjusts the Junior Sister's slightly tilted sword-grip posture, calmly adding a sentence: "But I'm more afraid of you waiting for others to decide every step for you from now on." The Junior Sister's originally tense gaze gradually softens, and she asks again: "So do I count as having won now?" The Senior Sister looks down at the longsword lying at the Junior Sister's feet and answers expressionlessly: "Pick up your sword first." Extreme Close-up: The Junior Sister lets out a soft laugh with suppressed emotion Behind her, in the shallow depth of field, the master gives a barely perceptible nod of approval Camera requirements: The word "Afraid" should be short, true, and defenseless Readjusting the sword-grip posture is one of the most critical physical actions of this segment This is not comfort, but evidence of trust landing in action "Pick up your sword first" is responsible for pulling the emotion back from weight to deadpan humor The ending must be warm but never saccharine

Seedance Prompt | Being Seen I. Project Positioning Cinematic realistic quality, pure ancient Chinese Xianxia aesthetics. The core is a restrained, mature emotional relationship drama. Using: High-level martial arts character blocking, Arri Alexa cinematic texture, clear and stable facial micro-details, fine film grain, and natural volumetric light. The core reversal revolves around "being seen": Everyone remembers the one who swings the final sword, but the person who truly determines the outcome might be the one who found the first breakthrough point but went unnoticed for a long time. Narrative Principles: The emotional center belongs strictly to the two women in @image1 and @image2. Supporting characters provide external evaluation, social pressure, and witness the reversal, but cannot replace the leads as the emotional center. The environment remains vivid but neutral in narration. II. Reference Controls 1. Character Identity Anchors: @image1 strictly serves as the identity anchor for Character ID A (Sword Immortal Senior Sister). @image2 strictly serves as the identity anchor for Character ID B (Junior Sister). 2. Environment Reference: All newly uploaded background and location reference images together determine the environmental DNA. Requirements: Do not mechanically copy any single reference image, maintain a consistent world identity, and three shots must share consistent geographical logic. III. Environmental Principles The background is always vivid but neutral. Elements like water flow, mist, and vegetation must move naturally without triggering plot developments. IV. Technical Direction Organized according to Seedance 2.0 Fast capabilities, emphasizing: character consistency from image references, native lip-sync, stable gaze and character positions, and directorial control over performance, lighting, and camera movement. V. Character Settings Character ID A | Sword Immortal Senior Sister (@image1): 25-30 years old, East Asian, oval face, fair skin, black hair with white jade hairpin. Cold, stable, direct, and values truth. Character ID B | Junior Sister (@image2): 20-25 years old, East Asian, round lively face, braided hair, petite. Smart, restrained, used to putting herself in a secondary position. VI. Core Narrative Structure Three beats: 1. External credit given to Senior Sister. 2. Senior Sister immediately corrects it, pointing to Junior Sister's breakthrough. 3. Junior Sister says "No one saw it then," Senior Sister replies "I saw it," followed by a repositioning action to let Junior Sister stand in front. VII. Storyboard Script (0-15s) 0-5s | Wide/Long Shot: Battle aftermath. Senior Sister and Junior Sister stand apart. Master says "The credit for breaking the formation goes to you." Senior Sister replies "Wrong." 5-10s | Medium/Cowboy Shot: Enemy argues "The final sword was clearly yours." Senior Sister ignores him, looks at Junior Sister: "She found the first eye of the formation." Junior Sister is surprised. 10-15s | Close-up/ECU: Junior Sister: "But... no one saw it then." Senior Sister: "I saw it." Followed by: "The first step is often harder than the last sword." Senior Sister steps back, letting Junior Sister stand in front. VIII. Performance Requirements Senior Sister: Calm, stable. Junior Sister: From restrained to touched, standing tall. Master: Restrained approval. IX. Actions and Blocking...

Seedance Prompt | Weakness and Harbor I. Project Positioning Cinematic realistic quality, pure ancient Chinese Xianxia aesthetics. The overall focus is on restrained emotional relationship scenes, with a slight touch of cool temperature. Style Requirements: High-level three-layer character blocking (foreground, middle, background) Arri Alexa cinematic quality Clear and stable facial micro-details Delicate film grain Natural volumetric light Restrained, quiet, and clear progression of character relationships Core Twist: Redefining the traditional meaning of "weakness." A truly important person does not make a master weaker; instead, she makes her realize for the first time why she must return alive. Narrative Principles: The main emotional thread belongs only to the two women (Image 1 and Image 2). Enemies are only responsible for external judgment and pressure. Masters and disciples exist only as distant witnesses. Supporting characters must not steal the plot. The environment is always vivid but completely neutral in narration. II. Reference Control 1. Character Identity Anchors: Image 1 strictly serves as Character ID A (Sword Immortal Senior Sister). Image 2 strictly serves as Character ID B (Junior Sister). 2. Environment Reference: All newly uploaded background and location reference images are treated as the same environmental DNA. Before formal composition, silently deduce compatible: real terrain, architectural language, spatial scale, materials, vegetation, water bodies, weather, clouds, mountain mist, main light direction, reflection relationships, overall color, atmospheric depth, and reasonable movement lines. Recombine these into a unique, unified new space for this round. Requirements: Do not mechanically copy a single reference image; maintain identity in the same world; three shots share the same geographical logic; clear layers of foreground, middle, and background; the environment exists vividly, but the plot is driven only by characters. III. Character Settings Character ID A | Sword Immortal Senior Sister Same Image 1 Senior Sister. 25–30 year old East Asian female, oval face, fair natural skin tone, dark almond eyes, long black hair partially tied up with a white jade hairpin, tall and slender. Fixed styling: White embroidered silk Hanfu, translucent layered wide sleeves, silver waistband, jade pendant, white cloth boots, a single silver long sword. Character temperament: Calm, direct, concise, emotions not exposed; the truly important weight is placed in the last sentence. Character ID B | Junior Sister Same Image 2 Junior Sister. 20–25 year old East Asian female, rounded and lively face, black hair braided, petite figure. Fixed styling: Cyan-green linen Hanfu, dark belt, wooden hairpin, black cloth shoes, a single dark steel sword. Character temperament: Smart, sensitive, restrained; does not over-show weakness; when touched, will use a gentle comeback to protect herself. Supporting Characters: Only as world witnesses, must not steal the scene. Enemy Swordsman: One, defeated, kneeling several steps in front of the two. Elderly Master: One, small volume in the background. Sect Disciples: Two, small volume in the background. Requirements: Supporting characters always submit to the main narrative, do not steal the spotlight, do not speak, do not create new plots, only provide scene pressure and social witness. IV. Environmental Principles The environment must remain vivid but neutral in narration. If elements exist in reference images, they should continue naturally: water flows, mountain mist drifts, vegetation responds to wind, clouds move slowly, reflections change with camera angles. Ambient sounds exist naturally. Hard Requirements: The environment must not trigger the plot, change character goals, create plot devices, or "over-act" for emotional beats. All transitions must come from character dialogue, actions, choices, and gaze relationships. V. Three-layer Blocking Principles Foreground: Facial details, hands, silk fabric, hair strands, long sword. Ensure stable facial details and readable micro-expressions. Middle Ground: Spatial relationship between the two leads and the kneeling enemy. Focus on the change in side-by-side positioning. This layer handles core relationship blocking. Background: Elderly master and two sect disciples. Exist only as silent witnesses, blurrier and smaller. Only provide world presence and social relationship pressure. VI. Narrative Core The true proposition is: "Are you my weakness?" The answer is completed in three beats: Beat 1: The enemy makes a traditional judgment: having someone you care about means having a weakness. Beat 2: The Senior Sister does not deny "caring," but redefines "weakness." She doesn't stand in front of the Junior Sister; she stands beside her. Beat 3: "Weakness" is rewritten as "Harbor (Home)." It's not a burden, but the reason to return alive. VII. Storyboard Script 0–5s | Wide or Long Shot: In the main confrontation area naturally selected from the spatial relationships of the reference images, two women stand side-by-side from head to toe, with a minimal distance between them. A few steps in front of them, only one defeated enemy swordsman is kneeling. An elderly master and two disciples exist as small-volume, silent background figures. The enemy looks up at the white-clad sword immortal and sneers: "With her, you have a weakness." The Senior Sister glares slightly at the Junior Sister and simply replies: "Yes." The Junior Sister's expression visibly tightens. Shot Requirements: Foreground, middle, and background must be clear. The two leads are the absolute visual center. The enemy is in the mid-foreground. Background master/disciples must exist but not distract. The first beat must hold; the "Yes" should be short, steady, and weighty. Junior Sister's reaction is shown only through eyes and slight facial tension. 5–10s | Medium or Cowboy Shot: Maintain the same white silk Hanfu woman, green linen Junior Sister, enemy, and identical geographic space. The Senior Sister does not take a traditional protective stance. She takes a tiny step to the side until they are truly side-by-side on the same line. She looks at the enemy and says calmly: "Before, I only thought about how to win." A full half-beat pause, then: "Now, I also have to think about how to come back together." The mockery on the enemy's face fades. The Junior Sister doesn't turn her head, only slowly looks at the sister beside her. Shot Requirements: The second beat is achieved through "blocking change." The core is not protection, but standing together. The step must be small but meaningful. "How to come back together" is the second core sentence. Junior Sister's emotion is shown through her eyes turning slowly toward her. Background continues natural movement with zero causal impact. 10–15s | Close-up or Extreme Close-up: The Junior Sister finally asks softly: "So, am I really your weakness?" The Senior Sister truly looks at her this time and replies: "No." A weighty half-beat pause, then: "You are the harbor (home)." The Junior Sister immediately looks away, trying to hide her reaction, muttering: "Who has a harbor with you?" The Senior Sister replies without changing expression: "Then stand further away." The Junior Sister doesn't hesitate; instead, she immediately moves half a step closer to her, saying: "The wind is strong, I can't hear you." Extreme Close-up: A tiny smile finally appears on the Senior Sister's restrained face. The Junior Sister also tries to press down the corners of her mouth. They look forward together again. The kneeling enemy in the back remains blurred (shallow depth of field), and the background figures are even blurrier. Shot Requirements: The third beat must land on "Harbor." "You are the harbor" is the most important line. Junior Sister's reaction should be internal, like trying to act fine after being struck. "Wind is strong, can't hear" is a light comedic touch. Moving half a step closer must be clear and natural. The ending is not a heated reconciliation, but a quiet confirmation of their stance. VIII. Performance Requirements Senior Sister: Stable external emotions, important info said in shortest words, power comes from restraint, "Yes" sounds like an undeniable fact, "Harbor" must have weight, final smile must be tiny. Junior Sister: Tightens up at "weakness," moved by "coming back together," truly touched by "harbor." Protects herself with a gentle comeback, then moves closer. Enemy: Only external judgment, no scene-stealing. IX. Action and Positioning: Side-by-side with minimal distance, enemy kneeling in front, Senior Sister glares at Junior, Senior Sister steps to side, Junior Sister looks with eyes, Junior Sister moves half-step closer. Principles: Minimal action, maximum meaning, no exaggeration. X. Cinematography: 16:9, three continuous clear shots, Arri Alexa look, stable/observational, natural volumetric light, film grain, facial details, shallow depth of field for extras. XI. Sound: Native Mandarin dialogue, precise lip-sync, ambient environmental sound, restrained music, significant pauses between core lines. XII. Technical Specs: 15s total, 16:9, Seedance 2.0 Fast, character/identity consistency, physical realism, no subtitles. XIII. Negative: blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, background shift, glitching cuts, disappearing props

I. Project Positioning Cinematic realistic texture, pure ancient style Chinese Xianxia aesthetics. The overall temperament centers on restrained emotional relationship scenes, with a hint of very light deadpan temperature. Style requirements: Elegant classical character blocking Arri Alexa cinematic look Clear and stable facial micro-details Natural volumetric light Tangible silk and weathered textures Restrained, quiet, and weighted emotional progression Core Reversal: A statement that initially sounds like "pushing someone away" gradually reveals another layer of meaning through the character's real actions. It is not estrangement, but allowing the other person to truly leave their own shadow and grow into themselves. II. Narrative Core The true theme of this scene is not "sending off," but letting go. On the surface: The Junior Sister wants to go down the mountain The Senior Sister says "I won't stop you" It sounds like coldness, like letting go, like lack of care In reality: The Senior Sister isn't getting rid of her The Senior Sister is acknowledging that she should go on her own path What she provides is not restraint, but protection after stepping back The emotional change to be achieved in the segment: Beat 1: A harsh word causes slight pain Beat 2: A small but real caring action overturns the surface understanding Beat 3: A sentence that truly points out the relationship, making "letting go" become "fulfillment" III. Reference Control 1. Character Identity Anchors @Image 1 strictly serves as the identity anchor for Character ID A, Sword Immortal Senior Sister @Image 2 strictly serves as the identity anchor for Character ID B, Junior Sister 2. Environment Reference All background and location reference images uploaded in this round collectively serve as a single set of environment DNA. Before formal composition, silently execute a round of environment planning, synthetically deducing compatible: Real terrain Architectural language Spatial scale Material era Vegetation ecology Water system direction Weather Clouds Mountain mist movement Main light direction Reflective surfaces Overall color Aerial depth Feasible character walking routes Then recombine into a unique, unified new space for this round that has never appeared before. Requirements: Do not mechanically copy single reference images Maintain identity within the same world Three shots share the same geographical logic Vivid environment but absolutely neutral narrative IV. Character Settings Character ID A | Sword Immortal Senior Sister The same @Image 1 Sword Immortal Senior Sister. 25–30 year old East Asian female, oval face, fair natural skin tone, dark almond eyes, black long hair half-up, fixed with a white jade hairpin, tall and slender. Fixed styling: White embroidered silk Hanfu Translucent layered wide sleeves Silver waist seal Jade pendant White cloth boots A single silver longsword Character temperament: Restrained Quiet Short speech Emotions hidden in actions The truly important parts are not through confession, but through handling details Character ID B | Junior Sister The same @Image 2 Junior Sister. 20–25 year old East Asian female, round and agile face, black braided hair, small stature. Fixed styling: Cyan-green linen Hanfu Dark belt Wooden hairpin Black cloth shoes A single dark steel sword A small satchel fitting the ancient style world Character temperament: Sensitive but not fragile Understands the senior sister, but can also be stung by her words Soft heart inside, trying to be restrained on the surface This segment should move from "loss" to "truly understood" V. Environmental Principles The environment must always be vivid but maintain absolute neutrality in the narrative. Environment life that can exist continuously includes: Wind Water flow Mountain mist drifting Clouds changing slowly Vegetation responding to natural wind Reflections changing with camera angle Distant subtle activities Spatial ambient sound existing continuously But must meet: Must not trigger the plot Must not solve problems Must not interrupt dialogue Must not express emotions for characters Must not create setups for jokes The environment is only responsible for: Providing a sense of real presence Providing a sense of time flow Providing spatial hierarchy Providing cinematic breath VI. Core Prop Continuity Core props of this segment: Junior Sister's small satchel Junior Sister's dark steel sword Senior Sister's unique small jade whistle given to her Requirements: The satchel is prepared from the start The sword remains the same throughout The jade whistle must be clearly visible, becoming the emotional anchor for the second and third beats Continuity of the jade whistle from being handed over to being held tight at the end must be stable All hand movements must be clearly readable and realistically restrained VII. Storyboard Script 0–5s | Full Shot or Long Shot In the departure area naturally selected according to this round's reference image real geographical structure, the Junior Sister has already packed her bags to leave. The same Junior Sister looks at the Senior Sister and asks softly: "Senior Sister, I'm going down the mountain this time, you really won't stop me?" The same white-clad Sword Immortal does not hesitate and answers: "I won't. Go further away." The Junior Sister's eyes clearly dim slightly. Camera requirements: The two are fully visible from shoes to head The space should be quiet and open, suggesting "departure" The first beat must land "Go further away" should sound like pushing away, cold, not keeping her The Junior Sister's sense of being hurt is shown only through very slight eye changes The environment runs normally but doesn't participate in this beat's emotional landing point 5–10s | Medium Shot or Cowboy Shot Maintain the same two women, same clothing, and identical geographical space. The same Junior Sister turns to leave. Before she has taken even her second step, the same Sword Immortal Senior Sister actively reaches out: Adjusts her loose shoulder strap Retightens the knot on the sword Then places the unique small jade whistle into her palm Simultaneously saying calmly: "Blow it if you get lost, don't push yourself if you're hurt." The Junior Sister looks down at the same jade whistle and looks back at her Senior Sister again. Camera requirements: The second beat must land clearly The Senior Sister's actions are more important than the dialogue Adjusting the strap, retightening the knot, giving the whistle are quiet but specific care These actions shouldn't be rushed, like she had already planned them The camera must let the hand interaction and the whistle entering the palm be clearly visible The background remains vivid but neutral 10–15s | Close-up or Extreme Close-up The same Junior Sister asks softly: "Didn't you tell me to go further away?" The same Sword Immortal Senior Sister truly makes eye contact this time, calmly answering: "Going far is to let you see your own path." Maintain a half-beat pause with emotional weight. Then adds: "It's not to make you unable to find the way back." Extreme close-up continues: Junior Sister's eyes become slightly moist But she finally smiles Slowly tightens her grip on the same jade whistle in her palm She asks again: "And when I return?" The same Sword Immortal Senior Sister shows a very slight smile at the corner of her mouth: "Bring back some of your own stories." The Junior Sister nods seriously, then turns and walks forward, not looking back this time. The Senior Sister stands calmly where she is, watching her walk away, expression not of loss but of extreme restrained relief. Camera requirements: The third beat must land truly "See your own path" is the theme sentence "Not to make you unable to find the way back" is the relationship redefinition sentence "Bring back some of your own stories" is the closing sentence, must be gentle but restrained The ending is not a farewell but an acknowledgement of growth The Senior Sister's relief shouldn't be overdone, just enough for the audience to feel she is truly letting go

I. Project Positioning Cinematic realistic texture, pure ancient Chinese xianxia aesthetics. Style requirements: Restrained emotional drama transitioning into warm, deadpan humor. Elegant classical character blocking, Arri Alexa cinematic texture, stable facial micro-details, fine film grain, natural volumetric light. Core structure: 'Three-Beat Emotional Reversal'. 1st beat: use a cold remark to create emotional drop. 2nd beat: character's actions contradict the remark. 3rd beat: a weighty line redefines the relationship. Strict rule: plot must be driven by character choices, dialogue, and action; environment remains neutral and non-intervening. II. Narrative Principles The core is the re-confirmation of the relationship, not superficial conflict. Tone: quiet, restrained, with slight misunderstanding in the first half; warm and unsentimental in the second half. Humor comes from their mutual understanding and the senior sister's blunt way of caring. III. Reference Control 1. Character Anchors: @Image 1 strictly for Senior Sister (Role A). @Image 2 strictly for Junior Sister (Role B). 2. Environment: Merge uploaded backgrounds into a unified geography. 推演Terrain, architecture, scale, materials, vegetation, weather, lighting, etc. to create a consistent new space shared across three shots. IV. Character Settings Role A (Senior Sister): Same as @Image 1. 25-30 years old, East Asian, oval face, fair skin, dark almond eyes, black hair partially pinned with white jade, tall and slender. Outfit: white embroidered silk hanfu, wide sleeves, silver belt, jade pendant, white boots. Holds a silver longsword. Role B (Junior Sister): Same as @Image 2. 20-25 years old, round lively face, braided black hair, petite. Outfit: celadon linen hanfu, dark belt, wooden hairpin, black shoes. Holds a dark steel sword. Extra: Right wrist wrapped in clean light-colored cloth (must be stable throughout). V. Environmental Principles Environment operates independently and naturally (wind, water flow, mist, reflections) but does not trigger or resolve the plot. It only provides realism and depth. VI. Prop Continuity Dark steel sword, wrist wrap, and the cloth used to wrap the sword must be consistent. The wrapping action must be detailed and sincere, as it serves as the evidence for the second beat reversal. VII. Storyboard Script 0-5s | Wide Shot: In a quiet resting space, the Junior Sister looks at her sword and asks: 'Senior Sister, what if I can no longer hold the sword?' The white-clad Senior Sister answers without hesitation: 'Then don't hold it.' Beat 1 must feel cold, creating a drop for the character and audience. 5-10s | Medium Shot: Junior Sister is visibly hurt, saying: 'You are certainly blunt.' Senior Sister doesn't explain but takes the sword and carefully wraps it in cloth. She says: 'There are so many paths down the mountain; who says you must fly with a sword?' Beat 2 shows her actions contradict her words—she is protecting her sister's future. 10-15s | Close-up: Junior Sister asks: 'Am I still your junior sister?' Senior Sister looks her in the eye: 'It's you I recognize, not this sword.' A weighty pause. Junior Sister smiles slightly and asks: 'Will you still scold me then?' Sister replies: 'Yes, as always.' Junior Sister laughs, and the Sister hides a tiny smile while continuing to wrap the sword. Beat 3 redefines the relationship with warmth and deadpan humor. VIII-XII. Performance & Technical Detailed acting cues for facial expressions, hand movements for wrapping the sword, cinematic Arri Alexa camera movement (16:9), native Mandarin dialogue sync, and high technical fidelity for silk fabrics and micro-expressions.

Seedance Prompt | Turn an Intense Duel of Gazes into a High-Level Martial Arts Showdown I. Project Positioning Cinematic realistic texture, pure ancient Chinese xianxia aesthetics. Overall style: Epic-level master duel camera grammar, restrained visual comedy, Arri Alexa cinematic quality, clear and stable facial micro-details, fine film grain, natural volumetric light, ample controlled quiet pauses, and precise rhythmic response. Core Plot: Two true masters take a mundane act of 'staring at each other' and escalate it into a contest of martial dignity. Hard Limits: No magic failures, no secret reveals, no environmental accidents, no clumsy character moments. All humor must stem from pride, competitive desire, micro-expressions, breathing, and rhythm. II. Narrative Principles Plot progression relies solely on: Intention, gaze, breathing, micro-expressions, pauses, rhythm, dialogue, and minimal physical reactions. The environment must remain a vivid, active, and neutral world with continuous physical changes like flowing water, drifting mountain fog, swaying vegetation, shifting light, and natural parallax. III. Reference Control 1. Character Anchors: @Image 1 for the Sword Immortal Senior Sister, @Image 2 for the Junior Sister. 2. Environment Reference: New uploaded backgrounds define the environment DNA, organized into a unified new space that respects identity and logic across three shots. IV. Character Settings - Character A: Sword Immortal Senior Sister (@Image 1). 25–30 year old East Asian female, oval face, white embroidered silk hanfu, silver belt, jade pendant, carrying a silver longsword (sheathed). - Character B: Junior Sister (@Image 2). 20–25 year old East Asian female, round face, green linen hanfu, dark belt, carrying a dark steel sword (sheathed). V. Core Performance Logic The scene looks like a master duel but is a battle of composure and holding back basic human reactions. Humor comes from the contrast between the small event and the grand ritual. VI. Storyboard - 0–5s: Wide/Long shot. Two women stand 3 steps apart in a quiet space, sheathed swords in hand, staring seriously. Sister: "Today we compete in composure, not swords." Junior: "How?" Sister: "Whoever looks away first loses." Junior: "Bring it." - 5–10s: Medium/Cowboy shot. Symmetrical composition, camera slowly pushes in. No head movement, locked gaze. Junior blinks slightly; Sister raises an eyebrow. Eyes become watery. Junior: "Sister, your eyes are red." Sister: "So are yours." - 10–15s: Close-up/Extreme Close-up. Focus on eyelashes trembling and water in lower eyelids. They blink simultaneously. Quiet pause. Junior: "You blinked first." Sister: "You spoke first." Pause. Junior: "Draw?" Sister: "Again." Final shot: They lean forward slightly, locking gaze again. Cut to black. VII. Performance & Movement Characters remain intelligent and powerful. No exaggerated faces or slapstick. Precise micro-expressions: squinting, eyebrow raising, stable breathing, watery eyes. Minimal head movement. VIII. Camera & Technical Requirements 16:9 landscape, 3 continuous shots, Arri Alexa look. Restrained camera movement from wide to close-up. Real volumetric depth and physical effects (hair, silk). Synchronized Mandarin dialogue. Seedance 2.0 Fast optimization for identity stability.

Seedance Prompt | Exposed Martial Styles After Swapping Swords I. Project Positioning Cinematic realistic texture, pure ancient Chinese xianxia aesthetics. A fusion of restrained relationship-based comedy, elegant martial body choreography, and classical visual comedy rhythm. Arri Alexa cinematic look. II. Narrative Principles Plot is driven by: choice, dialogue, gaze, swordsmanship, body choreography, and reaction pauses. Environment is vivid but neutral. III. Reference Control @Image 1: Anchor for Senior Sister Sword Immortal. @Image 2: Anchor for Junior Sister. Environmental DNA follows a unified new world logic. V. Core Performance Logic The humor comes from both characters being strong but secretly observing/learning from each other. When truth is exposed, they remain stubborn and composed. VI. Storyboard - 0–5s: Wide shot. They stand face-to-face. Senior: "Let's swap swords and see how you fare without your own." They formally exchange swords. - 5–10s: Medium shot. Junior (with silver sword) performs Senior's signature moves. Senior (with dark sword) responds with Junior's unique style. They stop. Junior: "How do you know my style?" Senior: "How do you know mine?" Pause. - 10–15s: Close-up. Junior looks away: "Watched occasionally." Senior: "Me too." They return swords, turn around, and both fight a light smile. They look back simultaneously, then hide the smile. IX. Technical Specifications 15 seconds, 16:9, 3 continuous shots, synchronized Mandarin dialogue. Seedance 2.0 multi-modal reference optimization.

I. Project Positioning Film-level realistic quality, pure ancient-style Chinese Xianxia aesthetics. Overall style fusion: High-end meta-narrative comedy Elegant and clear swordplay blocking Restrained deadpan performance Arri Alexa cinema camera look Clear and stable facial details Fine film grain Natural volumetric light Core reversal: The first half looks like two unparalleled masters engaged in a life-and-death duel over a complete falling out, only to find in the end that they are actually just rehearsing an overly solemn sect play. II. Narrative Principles The plot can only be driven by the characters' own: Performance Dialogue Swordplay actions Mistakes Reactions The environment is always an independently running, real world, but must stay neutral in narrative: Do not trigger punchlines. Do not change character goals. Do not create transitions for characters. Do not express emotions for characters. Do not participate in establishing the punchline. III. Reference Control 1. Character Reference @Image 1: Only as an identity anchor for Character ID A, Senior Sword Immortal. @Image 2: Only as an identity anchor for Character ID B, Junior Sister. 2. Environment Reference All newly uploaded backgrounds and location reference images in this round jointly determine the same environment DNA. Before formal composition, silently push and integrate: Real terrain Architectural language Spatial scale Material system Weather state Vegetation types Retain corresponding water bodies if present in reference images Cloud and fog movement logic Main light direction Reflective surfaces Overall color relationship Air depth Real walkable spatial relationships Then rearrange into a: Unique, complete, reasonable new scene with continuous spatial logic for this round. Requirements: Do not mechanically copy the composition of a specific reference image. Do not treat the environment as a static background. Continuously retain several identifiable environmental anchors across the three shots. Allow wind, water, vegetation, mountain fog, cloud shadows, reflections, Distant fine activities, and spatial ambient sounds to run naturally and continuously only in locations permitted by the reference image logic. IV. Character Settings Character ID A | Senior Sword Immortal Same as @Image 1. 25–30 year old East Asian female, oval face, fair natural skin tone, deep apricot eyes, black long hair half-up, fixed with a white jade hairpin, tall and slender. Always wears the same outfit: White embroidered silk Hanfu Semi-transparent layered wide sleeves Silver waist sash Jade pendant White cloth boots Holds the unique silver longsword. Character ID B | Junior Sister Same as @Image 2. 20–25 year old East Asian female, round and nimble face, black hair braided, petite stature. Always wears the same outfit: Green linen Hanfu Dark waist belt Wooden hairpin Black cloth shoes Holds the unique dark steel sword. V. Core Props Must be continuously retained and stably presented throughout the segment: Character A's unique silver longsword Character B's unique dark steel sword The same roll of bamboo paper for rehearsal Requirements: The rehearsal bamboo paper roll must be clearly visible in the middle and later segments. It doesn't appear suddenly; it was already placed with their personal props. Prop spatial relationships are clear and stable. VI. Storyboard Script 0–5s | Full Shot or Wide Shot In the main open space naturally planned according to the real geographical relationship of the reference images, the same two women confront each other five steps apart. The same white-clad sword immortal says in a deep, extremely resolute voice, as if truly saying a final goodbye: 'After today, you and I are through.' The same green-clad junior sister slowly draws her sword: 'Just as I wish.' Both people start forward simultaneously. Requirements: This segment must make the audience truly believe it is a life-and-death duel. Character aura is solemn and oppressive. The environment only serves as a grand real space, not participating in emotional drive. Both people fully and clearly visible from shoes to head. 5–10s | Medium Shot or Cowboy Shot Maintain the same two women, same clothing, same longswords, and exactly the same geographical space. The two swords first collide precisely, making a single crisp 'clang'. Then complete two clean, elegant, easily identifiable swordplay exchanges. Just as the emotion and action reach their climax, the same junior sister suddenly freezes in a half-drawn sword pose, lowers the blade slightly, and says directly: 'Wait, this line is mine.' The scene suddenly goes quiet. The same senior sword immortal still maintains a heroic dueling pose, pauses fully for half a beat, then slowly relaxes and asks: 'I said it too early again?' The junior sister nods seriously. Both people look at the same roll of bamboo paper for rehearsal that was already placed with their personal props, then exchange a tired look like they have rehearsed many times. Requirements: The punchline must be clean, clear, and sudden. The background continues its natural motion but must not participate in the humor. Comedy comes from the 'play within a play' being interrupted by the actors themselves. 10–15s | Close-up or Extreme Close-up The same two women readjust with very skillful movements: Gripping swords Stance Body posture The same white-clad sword immortal asks in a lowered voice: 'Which line should we start over from?' The same junior sister answers: 'From "through".' The senior sister takes a deep breath and instantly regains the terrifying, murderous dueling expression from before. However, before she can speak, the junior sister adds one more line: 'This time I'll be "through" first.' Extreme close-up: The senior sister's solemn expression breaks slightly for an instant, then she immediately steps aside elegantly, making a formal invitation gesture, and says: 'Please.' The junior sister tries to hold back a smile, raises her sword again, and begins in an incredibly sorrowful tone: 'After today —' Just before she finishes, cut to black precisely. Requirements: The end is not a full laugh, but a highly restrained break in character. Strong meta-narrative comedy feel, but characters still have dignity. The cut to black should be sharp, leaving an aftertaste.

A glitch video featuring a rhythmic stop-motion-like approach, followed by intense morphing between entirely different dimensions such as 3DCG, live-action, 2000s anime, and pixel art. It ends with a browser-crashing-style screen collapse. Concept: Completely eliminate smooth camera movements; use a rhythmic jump-cut approach to get closer to the face. The girl's face does not need to maintain its original features. The entire screen should repeatedly switch between 3D and 2D styles. Camera/Editing: Smooth transitions are strictly prohibited. 0.0-3.0s: Stop-motion approach with jump cuts every 0.5s accompanied by heavy chromatic aberration. 3.0-8.0s: Dimensional collapse at close-up, switching styles every 0.5-1s. 8.0-10.0s: Total collapse with error windows and block noise overlapping the screen until it becomes a chaotic collage. Constraint: Do not attempt to maintain the silhouette; focus on hard-cut transitions between different rendering styles.

Create a 3D claymation style animation of a cute caricature of Odysseus, wearing weathered bronze armor, a deep red cape, leather sandals, and windswept hair, sailing alone across a miniature stormy sea in a tiny wooden ship. A gigantic Cyclops suddenly rises from the water and reaches toward him. Odysseus calmly pulls out an enormous olive, launches it into the Cyclops’s mouth with a tiny catapult, and watches as the confused giant begins chewing happily. The storm instantly clears, the Cyclops gives him a thumbs up, and Odysseus takes a triumphant selfie with the giant in the background. Smooth expressive clay animation, epic orchestral music mixed with playful sound effects, handcrafted textures, cinematic lighting, mythological scale, absurd humor, viral energy.

Create a 3D claymation-style animation of a cute caricature of Erling Haaland wearing Norway's red number 9 football kit, white shorts, red socks, and red boots, with his signature blonde ponytail, walking across a simple muted green background while carrying a giant leek on his shoulder and a bucket of harvested leeks. He stops at a dirt mound, digs a deep hole with a hand trowel, plants the giant leek firmly into the soil, pats it down, then takes out his smartphone, flashes a peace sign with a playful wink, and captures a celebratory selfie. Smooth, expressive character animation with whimsical, humorous, cinematic lighting.

Create a whimsical, cinematic claymation-style animation of a playful black-and-white cat transforming a grayscale world into a vibrant paradise. As the cat flips, runs, and lands, each step splashes rainbow paint and colorful confetti, bringing grass, flowers, rivers, trees, houses, mountains, and the sky to life. The camera follows with smooth dynamic movements, revealing a charming floating island filled with smiling celestial characters, lush vegetation, and colorful details. End with the cat proudly holding a red flower in a fully transformed, cheerful miniature world. Pixar-quality lighting, handcrafted clay textures, smooth animation, vibrant colors, and magical storytelling.

Style: STOP-MOTION ANIMATION — stepped, frame-by-frame motion brought to a HAND-PAINTED 2D . True 12 frames per second, ANIMATED ON TWOS: 12 distinct hand-painted drawings per second, each pose held two frames then snapping to the next, never gliding. Constant painterly BOIL — brushstrokes and outlines subtly alive frame to frame. NO smooth interpolation, NO motion blur, NO morphing, real frame-by-frame animation not AI slop. Style from <<<image_1>>>, the yurt camp from <<<image_2>>>, the wolves from <<<image_3>>> — lean steppe wolves, coal-black with cold sheen, pale eyes. THE HERDSMAN from <<<image_4>>> (on foot). RIDER 1 from <<<image_5>>> riding HIS horse from <<<image_6>>> — keep this man and this horse together as one consistent pair. RIDER 2 from <<<image_7>>> riding HIS horse from <<<image_8>>> — a second consistent pair. Never mix the riders onto each other's horses. Atmospheric motion (blowing snow, blizzard haze) moves SMOOTHLY; figures, horses and wolves step on twos with secondary action. DIRECTOR'S NOTES: 1. THE SCENE — dusk, a blizzard rising over the Kazakh yurt camp. Wolves have come down on the herds. Panic erupts: THE HERDSMAN screams the alarm, men run, RIDER 1 is the first into the saddle and the first to tear out into the storm, RIDER 2 right behind him shouting orders. 2. RISING ENERGY — the sequence builds from dread to chaos: the dark wolf shapes at the herd, then the scream, then the explosive mounting and gallop. Each shot more energy than the last. 3. RIDER 1 IS THE LEAD — give RIDER 1 a clearly readable beat: he is the first to vault into the saddle of his horse (<<<image_6>>>) and the first to spur out of the camp, leading the charge, grim and silent, his face readable for a moment so the audience will know him again. RIDER 2 mounts shouting and follows. 4. AGGRESSIVE HANDHELD — the camera is in human hands in a panic: constant restless motion, jerky corrections, vertical bounce from running steps, buffeted by wind gusts, the horizon never level, never gimbal-smooth, never tripod-locked. Raw documentary chaos. 5. REAL PHYSICS — horses wheel and rear with weight, men vault into saddles with effort, snow is kicked up by hooves and boots as bursts of loose powder (drawn snow-spray effects on twos), cloth and manes whip in the wind with follow-through. 6. SECONDARY ACTION on twos on everything: chapan skirts and fur hats whipping, horse manes and tails streaming, breath-vapor of men and horses torn off by wind, harness straps swinging. 7. LIGHT — failing dusk in a blizzard, dim cold blue-grey storm light, soft and diffuse, no rays, no beams, no god rays; one or two struggling torch flames as small warm accents whipping in the wind. Correct neutral white balance, NOT a blue filter, muted desaturated, the wolves near-black masses. SHOT 1 — WIDE, ~35mm, aggressive handheld, the yurt camp at dusk in driving snow. COMPOSITION: the pale domed yurts low across the LOWER-RIGHT third, the churning panicked herd a dark restless mass on the LEFT third — and along the far edge of the herd, low dark wolf shapes from <<<image_3>>> flowing fast between the snow-veils, barely readable, a dark current, varied in tone and stride, never mirrored, never cloned. Blizzard haze drifts smoothly, figures and animals on twos. The herd wheels and screams. Energy rising. HARD CUT to SHOT 2 — MEDIUM, ~35mm, aggressive handheld running with him: THE HERDSMAN (<<<image_4>>>) bursts between the yurts toward camera, stumbling in the snow, waving an arm, his chapan and fur hat whipping with follow-through, breath-vapor tearing away, screaming raw over the wind, lips on twos: "Böri-i-iler!! BÖRILER!!!" Figures scramble behind him, a torch flame struggling and whipping. Snow bursts from his boots as drawn powder on twos. HARD CUT to SHOT 3 — MEDIUM, ~50mm, aggressive handheld in the chaos, on RIDER 1: he reaches his horse (<<<image_6>>>) at a run, the horse wheeling with real weight, and VAULTS into the saddle in two hard stepped poses, grim and silent, his weathered face readable for one beat in the torch-glow — then he wrenches the horse around and spurs it, the first one driving out into the storm, snow bursting from the hooves, his chapan and the mane streaming with follow-through. Behind him RIDER 2 (<<<image_7>>>) swings up onto his own horse (<<<image_8>>>), shouting hoarse over the wind, lips on twos: "Attardy qutqar!!" HARD CUT to SHOT 4 — WIDE, ~35mm, aggressive handheld, the edge of the camp: RIDER 1 already a dark mass at full gallop tearing toward the UPPER-LEFT into the blizzard, leading; RIDER 2 and one or two further silhouetted riders breaking after him from the LOWER-RIGHT, hooves drumming, powder bursting, manes tails and chapans streaming on twos, the pale yurts falling behind in the snow-haze, the dark storm ahead as the negative space they ride into. Camera buffeted, horizon tilting. The hoofbeats swallowed by the wind. End. Audio: NO MUSIC — the roar of the blizzard wind, panicked horses screaming and stamping, the herdsman's raw cries, men shouting, drumming hooves fading into the storm, distant wolf snarls under the wind. No subtitles. Natural diegetic sound only, absolutely no music. Constraints: stop-motion stepped cadence on twos at 12fps with painterly boil, figures horses and wolves stepping pose to pose never gliding, NO smooth interpolation NO motion blur NO morphing, blizzard haze and blowing snow smooth while all living things and drawn snow-spray effects step on twos, hand-painted oil look from <<<image_2>>> not photoreal not 3D, four shots with hard cuts rising in energy — wide wolves-at-the-herd, medium the screaming alarm "Böriler!!" from THE HERDSMAN <<<image_4>>>, medium RIDER 1 <<<image_5>>> vaulting onto his horse <<<image_6>>> grim and silent his face readable one beat and spurring out FIRST while RIDER 2 <<<image_7>>> mounts his horse <<<image_8>>> shouting "Attardy qutqar!!", wide the riders tearing out into the storm RIDER 1 leading — each rider stays on his own horse the pairs never mixed, AGGRESSIVE HANDHELD constant restless jerky motion vertical bounce wind-buffeted horizon never level never gimbal-smooth, REAL PHYSICS horses wheeling and rearing with weight men vaulting with effort snow kicked as loose powder bursts, SECONDARY ACTION on twos chapans fur hats manes tails harness breath-vapor all whipping with follow-through, dusk blizzard dim cold storm light soft no rays no beams torch flames as small warm whipping accents, correct neutral white balance not a blue filter muted desaturated, wolves from <<<image_3>>> varied never mirrored never cloned, spoken Kazakh in Latin transliteration pronounced as written not Russian-accented, NO MUSIC only storm and voices.

Style: STOP-MOTION ANIMATION — stepped, frame-by-frame motion brought to a HAND-PAINTED 2D look, a moving oil painting, NOT clay, NOT puppets, NOT 3D. True 12fps, ANIMATED ON TWOS: 12 distinct hand-painted drawings per second, each pose held two frames then snapping to the next, never gliding — even at full gallop the action steps pose to pose with the strobing cadence of hand-drawn animation. Constant painterly BOIL. NO smooth interpolation, NO motion blur, NO morphing, real frame-by-frame animation not AI slop. Style from <<<image_1>>>, the wolves from <<<image_2>>>, RIDER 1 from <<<image_3>>> riding HIS horse from <<<image_4>>> — the same lead rider and the same horse as one consistent pair throughout, never another horse, never another man. The dusk snowfield from <<<image_5>>>. Blizzard haze and blowing snow drift SMOOTHLY; figures, horses, wolves and drawn snow-spray effects step on twos with secondary action. DIRECTOR'S NOTES: 1/ THE SCENE — full-gallop chase in the dusk blizzard, brutal and fast. RIDER 1, the lead rider, chases a sprinting wolf, leans low off the saddle in the manner of a kok-boru player reaching for the ulak, seizes the wolf by the scruff — and the wolf twists and savages his arm. A second wolf hits him from his blind side. He is torn from the saddle at full speed and the pack swarms him. Savage, dynamic, with REAL PHYSICS and a visible causal chain — every wound has its on-screen cause. 2/ REAL PHYSICS AND CAUSALITY — the horse (<<<image_4>>>) gallops with true four-beat weight; RIDER 1 hangs low off the side of the saddle, one hand braced, reaching down; the wolf is a heavy animal — when seized it twists its WHOLE body mid-stride, and that twisting weight plus the clamping bite WRENCHES the rider off balance; the bite is shown on screen: jaws clamp and crush onto his forearm, tearing sleeve and flesh, dark blood — cause before effect, always; the second wolf launches from his blind side and slams full-body into his chest like a missile; torn from the saddle at gallop speed he hits the snow HARD and TUMBLES with momentum, rolling over and over, loose powder snow bursting around each impact (drawn snow-spray effects on twos), the riderless horse galloping on; the pack converges as a fast dark stream and swarms over him, a savage thrashing dark mass. 3/ BRUTALITY STAGED, NOT LINGERED — the violence is fast, hard and physical: the crunching bite, the dark blood across the snow, the swarming pack. Show the savagery through motion, dark mass and sound, not slow anatomical detail. 4/ AGGRESSIVE HANDHELD CHASE CAMERA — racing alongside at gallop, violently jolting with the speed, whipped by wind, jerky corrections, the horizon tilting and never level; on the fall a hard DUTCH TILT as the world goes over with him. Never gimbal-smooth, never tripod-locked. 5/ SECONDARY ACTION on twos — RIDER 1's chapan skirts and fur hat whipping with follow-through, the horse's mane and tail streaming, the wolves' fur rippling along their backs, harness swinging, breath-vapor of horse and man tearing off in the wind, snow bursting from hooves as drawn powder on twos. 6/ LIGHT — deep dusk in a blizzard, dim cold blue-grey storm light, soft, no rays, no beams, no god rays, the figures dark masses against the pale snow. Correct neutral white balance, NOT a blue filter, muted desaturated; the blood a dark muted red, stark on the snow but never glossy, never bright. SHOT 1 — FULL-GALLOP TRACKING, ~35mm, aggressive handheld racing alongside. COMPOSITION: RIDER 1 on his horse a large dark mass driving in from the RIGHT third, the sprinting wolf low ahead of him on the LEFT third, both tearing diagonally through the frame left-and-deeper — the diagonal of the chase as the line of dynamics; the pale storm-lit snowfield from <<<image_5>>> streaking past as negative space. RIDER 1 drops LOW off the side of the saddle in a kok-boru lean, one hand braced on the saddle, the other arm stretching down for the wolf's scruff, his chapan and the horse's mane whipping on twos, hooves throwing bursts of powder. He bares his teeth, hoarse over the wind, lips on twos: "Ustadym!.." His fist closes on the wolf's scruff — HARD CUT to SHOT 2 — CLOSE DYNAMIC, ~50mm, aggressive handheld slammed in tight: the wolf, seized, TWISTS its whole heavy body mid-stride in one violent stepped motion — and its jaws clamp CRUSHING onto RIDER 1's forearm, on screen, tearing through sleeve and flesh, dark blood whipping across the snow and the wolf's muzzle. His raw scream tears over the wind. The wolf's twisting weight and the bite WRENCH him sideways off his balance, his body dragged half out of the saddle, his fur hat ripping away with follow-through. Cause and effect brutal and readable, all stepping on twos. HARD CUT to SHOT 3 — MEDIUM WIDE with a hard DUTCH TILT, ~35mm, aggressive handheld: from his blind side a SECOND wolf launches — a dark missile — and slams full-body into his chest. Torn from the saddle at gallop speed RIDER 1 hits the snow HARD and TUMBLES, rolling over and over with real momentum, powder snow bursting at each impact as drawn effects on twos, the horizon tilted and reeling, his riderless horse (<<<image_4>>>) galloping on into the storm. And then the pack pours in — a fast dark stream of wolves from <<<image_2>>> out of the storm, varied coats, converging from all sides and SWARMING over him, a savage thrashing dark mass on the pale snow, dark blood spreading, snow scattering, his cries swallowed by the blizzard and the snarling. The dark mass of the pack traps and frames him. Hold one brutal beat in the howling wind. End. Audio: NO MUSIC — the roar of the blizzard, pounding gallop, the wolf's snarl and the wet crunch of the bite, the rider's raw scream, the heavy tumbling impacts, the converging snarls of the pack, the wind swallowing everything. No subtitles. Natural diegetic sound only, absolutely no music. Constraints: stop-motion stepped cadence on twos at 12fps with painterly boil even at full gallop, every action stepping pose to pose never gliding, NO smooth interpolation NO motion blur NO morphing, blizzard haze smooth while all figures animals and drawn snow-spray effects step on twos, hand-painted oil look from <<<image_1>>> not photoreal not 3D not glossy, three shots with hard cuts — full-gallop tracking with the kok-boru lean and the grab "Ustadym!..", close dynamic of the twisting wolf and the crushing on-screen bite with dark blood and the scream, dutch-tilted wide of the second wolf's full-body slam the hard tumbling fall at speed and the pack swarming him as a savage dark mass — RIDER 1 from <<<image_3>>> and his horse from <<<image_4>>> the SAME consistent pair in every shot never swapped, REAL PHYSICS AND VISIBLE CAUSALITY every wound caused on screen the wolf's twisting weight wrenching the rider the bite before the blood the slam before the fall the momentum carrying the tumble, AGGRESSIVE HANDHELD racing jolting wind-whipped horizon never level hard dutch tilt on the fall never gimbal-smooth, SECONDARY ACTION on twos chapan hat mane tail wolf fur harness breath-vapor all whipping with follow-through, wolves from <<<image_2>>> varied in coat tone size and stride never mirrored never cloned, deep dusk blizzard dim cold storm light soft no rays no beams, correct neutral white balance not a blue filter muted desaturated dark muted blood never bright never glossy, brutality fast hard and physical staged through motion dark mass and sound not lingering gore, spoken Kazakh in Latin transliteration pronounced as written, NO MUSIC only storm hooves snarls and screams.

A clay pirate captain duck with a tiny hat, wooden sword, eye patch and dramatic pose Sails across a bathtub ocean, battling giant soap waves and riding a sponge ship toward a rubber duck treasure island Bathroom fantasy world with bubbles, towels as cliffs and warm light reflecting on water 3D clay , Pixar-style playful adventure, soft clay textures, rounded props, dynamic camera glide over water, bright cheerful atmosphere, ending with the duck discovering a golden bath plug treasure.

Style: STOP-MOTION ANIMATION — stepped, frame-by-frame motion brought to a HAND-PAINTED 2D look, a moving oil painting, NOT clay, NOT puppets, NOT 3D. True 12 frames per second, ANIMATED ON TWOS: 12 distinct hand-painted drawings per second, each pose held two frames then snapping to the next, never gliding. Constant painterly BOIL — brushstrokes and outlines subtly alive frame to frame. NO smooth interpolation, NO motion blur, NO morphing, real frame-by-frame animation not AI slop. Style from @[Image 1](image_1), ANA from @[Image 2](image_2), UMAI from @[Image 3](image_3), THE WOLVES from @[Image 4](image_4) — lean steppe wolves, coal-black with cold sheen, pale eyes, the snowy cliff from @[Image 5](image_5). THE INFANT is not a separate reference — render from description: a tightly swaddled baby wrapped in a thick DARK BLUE wool blanket, PRESSED AGAINST ANA'S CHEST in one arm as she clings to the cliff, only a small dark-blue bundle, face barely visible, stirring faintly; NOT a second active child. Heavy weather: drifting FOG, falling and blowing SNOW, gusting WIND. Atmospheric motion (fog, falling and blowing snow, wind-haze, breath-vapor) moves SMOOTHLY; figures, falling rock, wolves and drawn snow-spray step on twos. DIRECTOR'S NOTES: 1. THE SCENE — the catastrophe, the heart of the whole story, and it MUST READ through a precise CHAIN OF CAUSE AND EFFECT, with composition and camera telling the cruel story. As ANA climbs one-armed with her infant, her foot dislodges a rock slab; the slab falls toward little UMAI below; and a WOLF leaps and SHOVES UMAI clear WITH ITS BODY an instant before the slab hits — the wolf SAVES her. Then the pack streams in and carries UMAI away. The wolves do not attack — the first saves her, the pack takes her. This reversal is everything. 2. THE CAUSAL CHAIN — stage each link on screen, cause before effect, on twos: (a) ANA's foot comes down on a ledge high on the cliff, the ledge CRACKS and BREAKS under her weight; (b) a heavy SLAB breaks loose and FALLS straight down toward UMAI — the falling rock the through-line; (c) UMAI stands directly below, looking up; (d) from the side a WOLF LAUNCHES and SLAMS its shoulder and body into UMAI — NOT its jaws, NOT a bite — knocking her sideways clear; (e) the slab CRASHES into the snow exactly where she stood, a burst of powder snow on twos; (f) UMAI tumbles unhurt among the wolves. Every cause visible. 3. THE WOLF SAVES WITH ITS BODY — CRITICAL: the first wolf strikes UMAI with its shoulder/flank, a SHOVE, mouth NOT on her, no bite, no seize — a rescue read as a body-slam. For one instant it LOOKS like an attack — then the rock smashes the empty snow and we understand. Play the shock-reversal. 4. THE PACK TAKES HER — the pack pours in as a fast dark stream from @[Image 4](image_4), varied coats, and SWEEPS UMAI up among them, NOT tearing her, carrying her in the current as they stream away into fog and blizzard. She is small in the dark flowing mass, swept away into the white. The beginning of her life among them. 5. ANA ABOVE SEES AND SCREAMS — high on the cliff, helpless, clinging one-armed with her baby, she SEES and reaches out and SCREAMS her daughter's name. FACIAL ACTING: her held composure SHATTERS — her face breaks open, eyes blown wide in horror, mouth tearing open, all the controlled tenderness from before exploding into raw terror, on twos as snapping held poses of a face coming apart. 6. CAMERA LAW (angle and height tell the cruelty) — SHOT 1 high WITH ANA then a hard VERTIGO TILT DOWN following the falling rock — the height is the cruel mechanism, she is the unwitting cause from above. SHOT 2 LOW at UMAI's level or below — we share the child's helplessness, the rock and the wolf bearing down on us. SHOT 4 the BIG-VERSUS-SMALL angle — ANA tiny and powerless high on the vast cliff, the storm dwarfing her, never powerful, only helpless. Aggressive dynamic handheld throughout, the horizon reeling on the scream, never gimbal-smooth, never tripod-locked. 7. COMPOSITION LAW (the frame tells the story — RUPTURE, the opposite of the vow's affinity) — LINE: the cruel CROSS of forces in the save — the VERTICAL line of doom (the falling rock from above) intersected by the HORIZONTAL line of salvation (the wolf from the side); their crossing on the tiny child IS the story. The pack's flow a DIAGONAL line of dynamics sweeping her off. The vertical gap between mother above and child below now PERMANENT and vast. SHAPE: the angular rock and angular lunging wolf (both read as threat) converging on the small rounded child. TONE: dark masses (rock, wolf, pack) on pale snow and fog, the child the focal point. MOVEMENT: contrast and high intensity — fast violent vertical fall, horizontal slam, diagonal sweep, against the slow helpless reach. CONTRAST & AFFINITY: maximum CONTRAST and PEAK visual intensity — the rupture the vow's affinity has been saving for. SPACE: deep vertical, the cruel distance. 8. FRAMED INK STAGING (read as masses) — FRAME-TRAP the small child between the falling rock above and the lunging wolf from the side, two dark angular masses closing on her. CONCEAL: fog half-swallows the catastrophe, and swallows UMAI as the pack carries her into the white. CAMERA HEIGHT: low and helpless at the child; tiny and powerless at the mother. LARGE VS SMALL: the small child and small distant mother against the vast cliff, the storm, the dark flowing pack. 9. WEATHER — FOG, SNOW, WIND, all smooth: low drifting fog across ground and cliff base, thick snow blown sideways in gusts, gusting wind dragging fog and snow and tearing breath-vapor, the catastrophe half-veiled and swallowed by the storm. 10. SECONDARY ACTION on twos — ANA's and UMAI's clothes hair and breath-vapor whipping, the wolves' fur rippling, the dark-blue swaddled bundle shifting, snow bursting from the rock's impact and the wolves' running as drawn powder on twos; fog and blowing snow smooth. 11. LIGHT — cold grey stormlight in fog and snow, flat soft directionless, no sun, no rays, no beams, no god rays, faces readable. Correct neutral white balance, NOT a blue filter, muted desaturated, snow soft storm-grey white, fog pale grey; nearly BLOODLESS — a rescue not a mauling, any blood dark muted and minimal. SHOT 1 — ON THE CLIFF, ~50mm, aggressive handheld, high on the rock face in fog and blowing snow. COMPOSITION: ANA climbing one-armed in the UPPER frame, the dark-blue swaddled infant pressed to her chest, her boot reaching for a ledge — framed HIGH WITH HER so we feel the drop below. Her foot comes down — the ledge CRACKS and BREAKS away on twos, a SLAB breaking loose. Hard VERTIGO TILT DOWN following the slab as it FALLS down the cliff through the fog toward the tiny figure far below — the vertical line of doom, the height the cruel mechanism. Cut on the falling rock. HARD CUT to SHOT 2 — BELOW THE CLIFF, ~35mm, aggressive handheld, framed LOW at UMAI's level. COMPOSITION (the cross of forces): little UMAI small in frame looking up, fog and snow blowing past — the SLAB falling INTO frame from the TOP straight down at her (vertical line of doom) — and from the SIDE a WOLF LAUNCHES and SLAMS its shoulder and body into her (horizontal line of salvation), the two lines crossing on the tiny child, knocking her sideways clear (jaws NOT on her, a shove not a bite) — and the slab CRASHES into the snow exactly where she stood, a burst of powder on twos. For one beat the two dark angular masses closing on her read as ATTACK — then the empty crater shows the wolf saved her. UMAI tumbles unhurt among arriving wolves. Real weight and impact, all on twos. HARD CUT to SHOT 3 — BELOW, WIDE, ~35mm, aggressive handheld. COMPOSITION (the diagonal sweep): the pack pours

Create an 8 second square 1:1 experimental art film featuring a small stack of handmade playing cards placed in the exact center of a deep crimson red textured surface. Use a perfectly locked top down overhead camera with a clean, symmetrical composition. The frame should feel balanced and minimal, with the deck centered and enough negative space around it. No camera movement, no hands, and no surrounding objects. The cards are warm off white with slightly uneven handmade edges, softly bent corners, subtle paper fibers, realistic thickness, delicate contact shadows, and a faint shadow beneath the deck. The background is a rich dark red textured surface, resembling paper, fabric, or velvet, with soft studio lighting, a gentle vignette, and darker edges. Animate the top card using rapid rhythmic stop-motion jump cuts. Replace the card approximately every three frames, creating around eight sharp transformations per second. Each new card should remain briefly frozen before instantly changing into another design. Add extremely subtle deck wobble, tiny positional shifts, slight changes in rotation, natural paper vibration, and occasional one frame motion smears during transitions. Each card contains a different bold editorial illustration created only with black, white, and vivid scarlet-red ink. Cycle through surreal minimalist artwork such as: horse riders, galloping horses, elegant female silhouettes, dancers, abstract human faces, expressive eyes, red lips, flowers, fashion portraits, masked figures, brush painted bodies, geometric facial fragments, mysterious black shadows, and poetic botanical symbols. Use rough ink strokes, screen printed textures, imperfect paint edges, dry brush marks, vintage European graphic design, contemporary fashion branding, handmade collage aesthetics, and refined artistic playing card composition. Include small abstract playing card rank and suit markings in opposite corners, but no readable sentences, logos, captions, or watermarks. Keep the white deck consistently centered while only the illustration, card markings, slight card shape, and tiny rotation change. Synchronize every transformation to a fast sequence of crisp mechanical clicks, card tapping sounds, and minimalist percussion. Build a hypnotic rhythm that becomes slightly more intense toward the end. Finish on a clean card featuring a graceful black human silhouette with one strong red accent. Hold the final image sharply for a brief moment before ending with a direct cut. Visual quality: premium studio photography, realistic paper texture, sharp graphic details, deep red and black contrast, soft cinematic shadows, polished art direction, subtle film grain, high end editorial branding film, smooth 24 fps output while preserving the deliberate 8 fps stop motion transformation rhythm. Avoid: moving camera, zooming, human hands, floating cards, cards flying awa

Using the reference image as the sole visual basis, generate a 5-second, 9:16, silent image-to-video. Animation requirements: Maintain a paper cutout / scrapbook / stop-motion collage style. All elements (characters, maps, word cards, copper coins, cart tracks, arrows, etc.) move as independent cutouts. Movements should have: frame-dropping feel, slight paper jitter, segmented displacement, and sticker bounce.

[CONDITION DEFINITION] 1:1 square, 15 seconds, no reference images. A cinematic high-end miniature diorama formation montage where a port town is rapidly assembled across a vast island. The direction is not a cheap toy, but the texture of a high-end handmade diorama, luxury collector toy, or a precision model for movie shooting. While maintaining cuteness, organized beauty, and the fun unique to models, it creates a sense of grandeur as the entire island expands. Shows bright clear daylight, blue sea, white waves, red/blue/white roofs, wooden piers, stone wharfs, lighthouses, bridges, canals, streets on hills, and multiple port sections. The main character is not a person, but the giant harbor island diorama itself. No text, logos, watermarks, or subtitles required. [SHOT/FLOW] 0.0–3.0 seconds: Start with a high wide-angle overhead shot. Not a small garden, but showing the entire large island surrounded by the blue sea. On the empty island terrain, the outlines of coastlines, bays, breakwaters, cliffs, hills, harbor mouths, canals, multiple wharfs, bridges, main roads, and terraced city blocks appear at high speed, and the large framework of the island and harbor is completed all at once. From the beginning, show the expanse of the open sea, bay, hills, and harbor, avoiding a small impression. 3.0–6.5 seconds: While maintaining a wide angle, the camera moves forward toward the harbor side while gently descending. Wooden piers, cobblestones, stairs, railings, warehouses, fishing huts, buildings along the canal, rows of houses with red/blue/white roofs, and narrow paths leading to the hills are assembled rhythmically as if clicking into place. The composition always shows the depth of the harbor, hills, and open sea without being too close. 6.5–10.0 seconds: Insert a natural scene transition and move to a medium-range shot of the harbor. Details such as multiple piers, small boats, sailboats, mooring ropes, wooden boxes, small cranes, streetlights, flags, bridges, canals, stone joints on the wharf, and pier planks are completed in succession. This is the scene where the pleasure of formation is shown most strongly, but it is not just close-ups of small objects, but shows the entire wide harbor section gaining density all at once. 10.0–12.5 seconds: Show lighthouses, streets on hills, multiple harbor sections, and bays leading to the open sea. The lighthouse is completed, and roofs, windows, bridges, streetlights, ships, and flags are arranged as a final finish, bringing the vast harbor island diorama close to its final form. Here, show the sense of scale of the entire island strongly once more. 12.5–15.0 seconds: Life is breathed into the completed harbor island. The water surface sparkles, small boats sway slightly, flags and ropes flutter in the wind, the lighthouse light slowly rotates, and windows and streetlights light up small. The finale is a magnificent panoramic view pulled back widely with a wide angle. Ending with the lingering sound of the large luxury diorama island, including the harbor, hills, lighthouse, multiple piers, and even the open sea, starting to come to life cutely. [CAMERA/EDITING] Emphasis is placed on wide-angle camera work, always making you feel the expanse of the entire island. Starts with a high bird's-eye view, followed by a descending advance, a close pass along the harbor, a light rotation around the lighthouse, and finally a grand panorama pulling back. Insert 2-3 natural scene transitions within 15 seconds. Close-up cuts may be included, but avoid getting too close so it doesn't just look like a small toy. Layer the foreground, middle ground, and background to create a composition where the harbor, hills, lighthouse, and open sea are visible simultaneously. The tempo is fast, emphasizing the pleasure of formation. Avoid monotonous linear movement or long pauses. [SOUND] With BGM and sound effects. The BGM is bright, refreshing, cinematic music with a slight sense of adventure. It supports the excitement of the luxury diorama being completed, but is not too flashy. Sound effects include small assembly sounds, wood or stone clicking into place, light clicking sounds, water sounds, rustling waves, sounds of ropes and flags, creaking of ships, distant seagulls, and small mechanical sounds of the lighthouse. No dialogue, no narration. [NEGATIVE] Avoid islands that are too small, small box gardens, cheap toy feel, cheap plastic feel, rough craft feel, narrow ports, towns with only a few houses, compositions without scale, dark cloudy weather, muddy sea, life-sized giant humans, making characters the main focus, text, logos, watermarks, subtitles, monotonous linear cameras, and endings that just stay completely still after completion.

Maintain the Qilin's appearance, body proportions, limb positions, antler structure, scale distribution, and overall environment exactly as shown in the first frame. 10-second continuous natural documentary shot, no cuts. A massive Qilin stands quietly on a prehistoric valley riverbank, staying in place, not walking or turning. it breathes slowly and heavily, with very slight rises and falls in the chest and abdomen; the head has only tiny natural movements, ears occasionally twitch slightly, and eyes blink slowly once. Long mane and neck hair sway slowly in the valley breeze, and the deep green scales on the back show tiny muscular ripples with each breath. Small animals by the river move naturally, some lowering their heads to drink, others taking a few slow steps, but never running in panic. The water surface flows continuously and slowly, producing tiny ripples and realistic reflections. Mist in the valley drifts slowly between the woods and the Qilin's legs, and distant leaves sway slightly in the breeze. The camera is at a low angle in the distance, using a 70mm natural documentary telephoto lens, moving extremely slowly towards the Qilin, with only a very slight rightward pan, creating restrained parallax that reveals the depth between the Qilin's massive body and the distant valley. The Qilin remains the absolute single subject of the frame, maintaining the weight of a real large creature, without exaggerated movements, no glowing, and no magical effects. Real wildlife documentary, BBC / IMAX natural history photography quality, realistic biological movement, realistic muscle weight, atmospheric perspective, morning soft light, subtle wind-blown hair, 70mm IMAX film, solemn, quiet, mysterious.

Seedance 2.0 / 2.5 Direct Delivery Version 15-Second Chinese Xianxia Heavenly Palace Cinematic Short Film | 5-Shot Continuous Hard-Cut Edition Generation Goal: Generate a complete, continuous 15-second Chinese Xianxia heavenly palace cinematic short film. The entire film consists of 5 clear cinematic shots, each approximately 3 seconds, connected by clean, direct cinematic hard cuts. Strictly Use: @Image 1 @Image 2 @Image 3 @Image 4 @Image 5 as independent visual and spatial anchors for each of the 5 consecutive shots. I. Core Constraints 1) Rules for Using the Five Images - The 5 reference images have completed art design. - Each shot must strictly correspond to its respective reference image. - Within each shot, the following elements from the reference image must be maintained: building structure, number of characters, character clothing, trees, water systems, sea of clouds, bridges, railings, circular gates, dragon pillars, palace proportions, spatial perspective, light direction, and overall composition. - Do not redesign these scenes. - Do not merge the five images into the same scene. - No AI morph transitions. - No building deformation during shot switches. 2) Overall Worldview and Visual Quality Maintain a unified Eastern Xianxia civilization visual DNA: authentic Chinese ancient architectural structures (vermilion wood structures, dark tile roofs, dark gold carvings, stone foundations), high-altitude sea of clouds, authentic natural light (warm gold low-angle sunlight, deep shadows, aerial perspective), restrained film texture, grand but quiet, natural but continuously living, authentic cinematic photography feel. II. Five-Shot Storyboard Shot 1 | 0–3s | @Image 1 | Corridor Bridge and Ancient Pine - Shot Type: Long-focus distant cinematic shot, approx. 135mm telephoto compression. - Camera: Very slow, steady slight push forward along the central axis. - Spatial Relationships: Foreground eaves, pillars, and railings have slight parallax; middle-ground four characters maintain natural spatial relationships; distant giant ancient pine and sea of clouds move the least. - Character Requirements: Four characters must be living from the first frame. They maintain unsynchronized natural micro-behaviors: two on the left move a few steps slowly along the corridor, one slightly turns to look at the central ancient pine, the two on the right continue walking slowly or turn heads naturally while talking. Paces, breathing, shoulders, body weight, cuffs, hems, and a small amount of hair all change naturally. No collective synchronized movements or statue-like stillness. - Environmental Requirements: Constant high-altitude wind; ancient pine trunk is stable but twigs and needles sway continuously; sea of clouds is not a static background—it must have clouds at different depths drifting slowly, local expansion, dissipation, and reforming, with a few thin clouds slowly passing distant peaks. Shot 2 | 3–6s | @Image 2 | Heavenly Palace Over the City - Transition: Direct hard cut. - Shot Type: Distant long-focus establishing shot, 135–200mm cinematic telephoto observation. - Camera: Extremely slow zoom out. - Goal: Gradually complete the scale relationship between the ancient city below, walls, rock formations, and the giant heavenly palace. - Building: Giant main palace structure is absolutely stable with no deformation or drift. - City and Background Life: The city below cannot be a static texture; it must be alive: slow-flowing low-level mist and moisture between city parts, clouds passing through buildings, distant cloud silhouettes evolving slowly. - Birds and Characters: Existing cranes and birds fly naturally (unsynchronized, non-looping); bottom small characters have readable robe movement and natural body motion. Lighting on roofs and clouds changes continuously as clouds pass the sun. Shot 3 | 6–9s | @Image 3 | Waterfall Immortal Palace - Transition: Direct hard cut. - Shot Type: Restrained long-focus observation shot. - Camera: Nearly locked, only slight horizontal drift. - Dynamics: Water, wind, and clouds. Waterfalls must fall naturally with weight and speed variance, creating mist that blends with clouds. Water surface has small natural ripples. Reflections of characters and buildings must change subtly with water and light; reflections must not freeze. Weeping willow twigs sway naturally in the wind. Characters show authentic asynchronous behavior (adjusting posture, moving forward, looking at the waterfall). Shot 4 | 9–12s | @Image 4 | Dragon Pillar Corridor - Transition: Direct hard cut. - Shot Type: 135mm long-focus cinematic tracking shot. - Camera: Very slow horizontal right movement along the suspended corridor, synchronized with character movement. - Character Walking: Four characters walk naturally with realistic human gaits (slight variation in step length, unsynchronized, shifting center of gravity). Robes have weight, and hair sways in high-altitude wind. One person may briefly look at the sea of clouds. - Parallax: Foreground railings and pillars move fastest; characters and corridor have medium displacement; distant mountains and waterfalls move the least. - Limitation: The circular gate is a solid structure and must not become a glowing portal. Shot 5 | 12–15s | @Image 5 | Interior Circular Gate Final Shot - Transition: Direct hard cut. - Shot Type: Solemn long-focus liminal space observation, 135mm lens compression. - Camera: Very slow diagonal retreat within the giant hall, slightly moving left. - Parallax: Foreground dragon pillars and floor reflections produce most parallax; two characters in the middle ground; circular gate and distant palace move the least. - Ending: Two characters continue walking naturally (unsynchronized, realistic weight). Distant environment remains alive with drifting clouds and falling waterfalls. Camera continues quiet retreat at the same speed until a stable cinematic ending is formed. III. General Rules for Natural Movement - Core Principle: Characters and environment must simultaneously have continuous life. No frozen backgrounds or statue-like characters. - Character Rules: Natural breathing, walking, posture adjustment, gaze shifts, and hair/fabric movement. No mechanical synchronization. - Wind Rules: Same high-altitude wind affects hair, sleeves, pine needles, and mist differently based on material weight. - Clouds: Overall slow displacement, local shape evolution, and realistic occlusion of distant buildings.

A cinematic view of a high-tech factory floor where advanced AI robotic arms work with precision. Automated systems manage the assembly line, and glowing data streams visualize a seamless digital integration powering the industrial revolution.

Create a 15-second premium Japanese technology commercial announcing the arrival of Seedance 2.5 on Tasky. IMPORTANT: GENERATE ABSOLUTELY NO TEXT OR TYPOGRAPHY ANYWHERE IN THE VIDEO. All titles, captions, product names, brand names, and Japanese copy will be added later during

{ "duration": "20 seconds", "aspect_ratio": "9:16", "style": "ultra photorealistic, raw handheld smartphone video, natural lighting, digital noise, realistic motion blur, found footage aesthetic", "camera": { "type": "handheld smartphone", "movement": "shaky human hand movement, increasing panic shake, sudden tilts and jerks", "perspective": "first-person from inside the dense crowd", "lens": "wide phone lens, natural distortion, lens flare" }, "scene": "Large excited crowd on a cliff in northern Spain during the August 12 2026 total solar eclipse. Late evening.", "sequence": [ { "time": "0-5s", "action": "Totality begins. Normal black moon with bright white corona. People cheering and filming. Camera pans across the crowd." }, { "time": "5-9s", "action": "Unexpectedly the black moon starts turning deep blood red. The corona shifts to glowing crimson. Crowd goes from cheering to confused gasps." }, { "time": "9-14s", "action": "A circular glowing portal tears open around the red moon, expanding outward with swirling energy and light. Some people in the crowd begin slowly floating upward toward it." }, { "time": "14-20s", "action": "Full chaos. People screaming and running. More bodies getting pulled into the air toward the red portal. Camera becomes extremely unstable as the filmer panics and tries to escape while still filming the red moon and portal." } ], "audio_cues": "excited crowd noise → confused gasps → loud screaming, wind, running, heavy breathing, phone mic distortion and clipping", "negative_prompt": "smooth camera, cinematic color grade, CGI look, text, watermarks, perfect focus, professional film, cartoon" }

A young East Asian woman with a short bob haircut sits on a wide windowsill in a messy Brooklyn-style apartment bedroom. She wears a black long-sleeve top, red collared shirt with a dark tie, gray cargo pants with a light blue sweater tied around her waist, pink-and-red arm sleeves, bright white socks, and teal sneakers. Purple over-ear headphones rest around her neck. The room is cluttered with anime posters (including Dragon Ball), comic books, a Monster Energy can, Funko Pops, clothes scattered on the floor, and a bed to the right. Outside the large open window is a cloudy New York City skyline of red-brick buildings. She looks at the camera with a slight smile, puts the headphones on, leans back, then suddenly swings out the window on a thin white web line. Dynamic tracking shots follow her web-slinging through the streets of New York: she flies between brick apartment buildings with fire escapes, over busy intersections filled with yellow taxis and pedestrians, past corner stores and traffic lights, diving and twisting acrobatically. The camera moves with her — low angles looking up, high angles looking down, fast motion blur on the city. She lands on a rooftop, stands with arms outstretched in triumph, hair blowing, looking out over the golden-hour skyline. The view includes water towers, dense rooftops, the East River, and the distant One World Trade Center glowing in the soft sunset light. Cinematic, realistic live-action style, vibrant colors, energetic movement, inspired by superhero web-slinging sequences.

Pilot = a fighter pilot around 35, rugged focused face with light stubble, sharp determined eyes, wearing a helmet with visor and oxygen mask, full flight suit. Intense and battle-focused. Face per reference. Appearance only. Seabaycity = a modern coastal city on a sea bay, skyline of skyscrapers along a waterfront bordering the open sea, harbor, bridges, dramatic overcast sky, buildings smoking and crumbling, fires and dust clouds rising, a city under attack. Moody desaturated blue-grey palette with orange fire glow. Environment only. Monster = a colossal towering lizard-like kaiju, skyscraper-sized reptilian creature with dark grey-green armored scales, huge clawed limbs, a long heavy tail, a fanged head with glowing eyes, spines down its back, standing in the sea bay smashing buildings. Terrifying and destructive. Appearance only. Jet = a modern military fighter jet, sleek grey afterburners glowing, missiles mounted, screaming through the sky at high speed. Vehicle only. SCENE: a cinematic kaiju action sequence. A colossal lizard monster rampages through a coastal city's sea bay, destroying skyscrapers along the waterfront. A lone fighter jet approaches from the distance across the sea, the pilot locking on with fierce focus, and fires missiles that strike the monster, driving it crashing down into the bay. Epic, intense, blockbuster disaster-film scale. TECHNICAL: 16:9, cinematic anamorphic-style lens, epic blockbuster scale, dynamic camera — sweeping aerial shots, fast jet tracking, dramatic low angles on the monster — moody desaturated blue-grey grade with orange explosion glow, heavy atmosphere, smoke, dust, lens flare, motion blur, fine film grain, photorealistic disaster-film quality. CUTS: CUT 1 (0-4s): Establishing shot — the colossal lizard monster towering over the sea-bay city, roaring and smashing a skyscraper along the waterfront, buildings crumbling and smoking, debris raining into the bay. Dramatic low angle conveying its massive scale. CUT 2 (4-7s): Wide shot across the sea — a lone fighter jet screams into frame from the distant horizon, low over the water, approaching the city and the monster at high speed. Fast tracking shot conveying speed and purpose. CUT 3 (7-10s): COCKPIT FACE ZOOM — cut inside the cockpit, camera pushes in on the pilot's intense focused face behind his visor and mask, glowing instruments around him, his eyes locked on the target. He grips the controls and fires. CUT 4 (10-13s): Missiles streak from the jet toward the monster and slam into it in a burst of fire and explosions, the creature reeling from the impact, staggering amid smoke and flame. CUT 5 (13-15s): The monster crashes down into the sea bay in an enormous explosion of water and debris, defeated, as the jet roars past overhead. Epic final beat, huge splash and spray.

Generate a 20-second, 16:9 widescreen, 720p, high-realism live-action urban street chase action short. Employ tense, fast, but spatially clear standard action movie editing. A rain-slicked commercial street at twilight, in the blue hour; most shops are closed, with metal rolling shutters, old brick walls, narrow alleys, external fire escapes, and scattered warm yellow shop lights forming a realistic urban environment. The wet asphalt reflects streetlights and neon, with no recognizable brands or garbled text. [Full Protagonist Design] The lead is a 22-year-old East Asian female, approx. 168cm, slender, firm, with an athletic build from daily running; agile but not a professional spy or superhero. Warm natural skin tone with real pores and texture; narrow oval face, clear brow ridge, thick straight eyebrows, sharp eyes, natural straight nose, clean jawline, and a very faint small scar at the end of the right eyebrow. Bust is full with clear contours. Black collarbone-length hair tied in a loose low ponytail. The ponytail, stray hairs, and clothes sway continuously with speed and turns; as stamina depletes, some stray hairs stick to the forehead and cheeks. Clothing is fixed as: Dark green matte lightweight windbreaker, Dark grey U-neck tight T-shirt, Black straight work pants, Dark grey non-slip running shoes, Black nylon crossbody bag with a burgundy strap crossing from right shoulder to left waist. Maintain the same face, build, hair, clothing, and bag throughout. The bag belongs strictly to the lead and cannot transfer to pursuers. [Pursuers] Fixed as 3 adult males, unarmed: Pursuer A (Tall, charcoal hoodie, black pants), Pursuer B (Stout, blue denim jacket, yellow T-shirt), Pursuer C (Thin, brown old work jacket, grey beanie). Identities and clothing must not swap; count remains 3. [00:00–00:02.5 | Setup & Discovery] Fixed side panorama. Lead sprints out of an alley on the left, checks behind. A, B, and C emerge in sequence. Lead accelerates towards the right. Music starts with low pulses and rhythmic footsteps. [00:02.5–00:06 | Side Track Follow | Distance Closes] High-speed track shot parallel to the lead. Lead sprints, bag bouncing. Pursuer A closes the distance. Lead steps into a puddle, water splashes realistically; no slipping. [00:06–00:09.5 | Low-Angle Follow | Obstacle Cross] Low-angle wide follow from the front. A tipped-over delivery bike blocks the path. Lead adjusts, steps firmly with the left leg, and jumps over the frame. Lands on the right foot, cushioning with bent knees, then re-accelerates. A hurdles the bike, B bypasses left, C bypasses right. [00:09.5–00:13.5 | Long Lens Pressure | Dead End] Frontal long lens from deep in a narrow alley. Lead turns into the alley; lens compression makes pursuers look closer. Lead discovers the alley ends in a brick wall. Speed drops to rapid deceleration steps. She panics briefly, then spots a metal tool chest and a fire escape ladder 2 meters above. [00:13.5–00:17 | Low-Angle Medium | Leap to Fire Escape] Lead sprints to the tool chest, steps on the metal edge (clanging sound), and leaps. Hands grab the lowest rung. Arms bear her weight. Pursuer A lunges for her right foot but misses by 20cm as she pulls up. [00:17–00:20 | Low-Angle Wide | Temporary Escape] Lead pulls herself onto the first fire escape platform. Chest, belly, and knees cross the edge sequentially. Pursuers A, B, and C reach the ladder and look up. Lead stands and climbs higher, looking down with alarm rather than triumph. Camera tilts up following her. Ends with the vibrating ladder, pursuers below, and a heavy musical strike.

# Seedance 2.5 — 30-second Korean Bus Romantic Comedy / Sudden Brake Catch / Romantic Long Kiss / Bus Exit Cheek Kiss Use exactly 2 uploaded image assets. image1 = The sole and highest priority reference for the adult female lead's identity, face, hairstyle, body type, school uniform, and accessories. Strictly maintain the recognizable face, skin tone, light blue short hair, white cloud hair clip, blue fish hair clip, petite proportions, and the exquisite deep blue school uniform, white shirt, bow, plaid skirt, socks, and shoes from the reference image. image2 = The sole and highest priority reference for the adult male lead's identity, face, hairstyle, body type, school uniform, and accessories. Strictly maintain the recognizable face, black hairstyle, tall and slender proportions, and the Korean-style male school uniform from the reference image. All main characters and recognizable background passengers are clearly depicted as adult actors over 20 years old, using only school-style uniforms and school bus themes. Generate a 30-second, 16:9 widescreen, 4K, 24fps, hyper-realistic live-action Korean romantic short film. Warm afternoon natural light, realistic skin, realistic human inertia, realistic bus suspension and braking physics, delicate eye contact and breathing, natural motion blur, slight film grain. The overall look must be like a real Korean romantic film, not a music video, animation, game CG, or plastic AI skin feel. [Characters and Scene] The female lead is about 156cm, distinctly petite; the male lead is about 180cm, with a natural and obvious height difference when standing together. At the start, the female lead stands in the bus aisle holding a hand strap; the male lead is clearly seated in the aisle seat next to her. The bus is moderately crowded with adult student actors of different faces, hairstyles, and poses. Warm afternoon sunlight shines outside the window. [0–4s | Peaceful After-School Atmosphere] A 35mm aisle wide shot establishes the bus interior, with the camera slowly pushing toward the two leads. The female lead holds a strap with one hand and a school bag with the other; the male lead sits quietly beside her. Outside objects move naturally, and the bus has subtle road vibrations. People around are looking at phones or chatting, overall relaxed and natural. [4–8s | Sudden Brake and Loss of Balance Must Occur Simultaneously] The driver suddenly slams on the brakes. First, clearly hear the sharp, realistic sound of brake friction: “Kkiii-ik——!” At the same instant the braking starts, the entire bus immediately generates strong forward inertia, the front dips, and the vehicle vibrates significantly. All physical reactions must happen at once: Hand straps swing violently forward; Bags, clothing edges, and hair all sway forward; Seated passengers' upper bodies jerk forward instantly; Someone instinctively grabs a handrail; The camera is also jolted forward by inertia with a slight recoil, creating a realistic handheld jolt. The female lead must also immediately lose her balance at the same instant the braking starts. The arm holding the strap is suddenly pulled straight, her center of gravity instantly shoots out of her foot support range, and her feet are forced to take several frantic, messy corrective steps forward. The action must be a continuous physical chain: Braking starts → Body simultaneously leans forward violently → Feet take quick corrective steps → Upper body is still carried by inertia → Fingers slip from the strap → The person naturally falls toward the male lead. Do not show the female lead suddenly losing balance after the bus has already stopped, and do not let her regain her footing midway. [8–12s | Male Lead Catches Her While Vehicle is Still Shaking] Use a 3/4 medium shot from the female lead's front-right, clearly showing the male lead's seat and the female lead's path of falling. The female lead is already in the process of falling, and the bus still has obvious braking inertia, not yet fully stabilized. The male lead is clearly seated at the very beginning. Upon seeing her fall toward him, he reacts immediately: Hips leave the seat → Legs quickly straighten → Takes a half-step into the aisle → Reaches out both hands to her waist and back. Just as the female lead is actually about to fall, the male lead catches her in time. The female lead actually crashes into his arms due to forward inertia, rather than being in a pre-arranged embrace pose. The two are carried by the remaining vehicle inertia, wobbling for a half-step; the male lead must also adjust his feet and center of gravity before finally steadying them both. The female lead instinctively grabs the front of the male lead's uniform or shoulders. The correct causality must be clear: Braking → Female lead simultaneously loses balance → Falling → Male lead rises → Catch → Both continue to wobble together → Finally stabilize gradually. [12–15s | First Heartbeat] The bus and camera gradually recover from shaking to stability. A medium two-shot slowly pushes in to a close two-shot. The male lead still supports the female lead's waist and back, and she still grips his clothes. The two realize their faces are only a few centimeters apart. The female lead is startled first, then her breathing becomes slightly erratic, her gaze shifts away then can't help but return, her cheeks gradually flushing. The male lead first checks if she is hurt, then his gaze slowly turns tender. The female lead tilts her head up slightly, and the male lead naturally tilts his head down. Maintain realistic eye contact, do not kiss immediately. [15–17s | Surrounding Students Become Excited] Nearby students notice their pose; some cover their mouths, nudge friends, peek, or open their eyes wide. The speech must sound like someone truly seeing a heart-fluttering scene and unable to hold back excitement, not like reading a script: Slightly fast speech, lowered voice, natural rising intonation at the end of sentences, with laughter and breathy tones; multiple voices can slightly overlap. “What's this, what's this?” (in Korean) The second one is higher-pitched. “Heol, did you see that?” (in Korean) “Heol” is short like a gasp, “Did you see that?” rises excitedly. “Daebak…” (in Korean) Whispered giggling, in disbelief. No announcer-style delivery, no mechanical one-word-at-a-time. [17–19s | Female Lead Shyly Tries to Back Away] The female lead realizes everyone is watching and tries to back out of the male lead's arms shyly, hugging her bag tightly and not daring to look around. The male lead does not sit back down. Just as she tries to turn and hide, the male lead gently grabs her hand or forearm, making her stop. The female lead turns back in surprise. [19–24s | Embrace + Romantic Kiss] The male lead takes the opportunity to gently bring her a half-step closer, one hand naturally encircling her lower back. The female lead freezes for a second, then does not continue to back away, her hands lightly grabbing the male lead's lapels. The camera very slowly pushes in from a 30° side-angle close two-shot. The two first maintain about 0.5 seconds of eye confirmation. The male lead naturally tilts his head down, and the female lead slightly tilts hers up. Then complete a clear, gentle Korean drama-style kiss lasting about 3–4 seconds. Do not just touch and separate. A gentle initial touch, then both bodies naturally relax into an embrace. The female lead gradually relaxes from initial tension, and the male lead's movements remain restrained and gentle. No exaggerated deep kissing, no intense movements. Try to maintain a stable continuous shot, only allowing a very slight arc movement. Warm light by the window naturally outlines their faces, hair strands, and uniform silhouettes. [24–26s | After the Kiss / Arrival at Stop] The two naturally separate but remain very close. The female lead lowers her head shyly, realizes she's still holding the male lead's clothes, lets go slightly, but can't help but reveal a small smile. The male lead also smiles tenderly. Nearby students lower their voices even more excitedly: “Heol…” “Did they really do it?” “Crazy…” (in Korean) Must be excited, laughing, incredulous youth slang tone. “Crazy…” is not anger, but an exclamation after seeing a heart-fluttering scene. Simultaneously, the bus begins to decelerate normally and gently toward the stop. [26–28s | Natural Exit] The bus door opens with a realistic “chick—” air pressure sound, and warm afternoon light enters the cabin. The female lead walks toward the door first, with the male lead following closely behind. The camera tracks from a 3/4 rear-side angle following them. Must clearly see sequentially: Bus stops steady → Door opens → Female lead steps down → Male lead follows off the bus. Do not teleport directly from inside the bus to the street. [28–30s | Female Lead's Active Cheek Kiss / Ending] After getting off the bus, the female lead doesn't leave immediately but turns at the door to wait for the male lead. Just as the male lead steps off the last step, the female lead looks up at him. Her expression changes from shy to a bit playful and brave. Due to the obvious height difference, the female lead naturally stands on her tiptoes, leans in gently, and gives a short, clear kiss on the male lead's cheek. This is clearly a light kiss on the cheek, not mouth-to-mouth. Do not suddenly cut to a close-up. The camera originally maintains a medium two-shot, naturally and softly pushing in as the female lead stands on tiptoes and leans in, clearly showing: Standing on tiptoes → Active approach → Cheek kiss → Male lead's surprised expression → Female lead's shy yet happy smile. The female lead drops back to her heels, face slightly flushed, smiles, and turns to walk forward. The male lead is stunned for a beat before truly smiling and immediately catches up to her. Finally, the camera does a slight 3/4 tracking from behind: The two walk naturally side-by-side on the warm afternoon street, shoulders getting closer, the male lead reaches out, and the female lead naturally takes his hand. The bus slowly departs in the background. The two walk and laugh together. CUT TO BLACK. [Cinematography Core] 0–4s: Stable wide

0:00-0:02 Extreme low-angle tracking inches from the spinning front tire. Dolly chase synchronized with wheel rotation. The motorcycle bursts through dense dust, suspension compresses aggressively, front fork oscillates, tire throws gravel toward lens, sparks scrape asphalt, exhaust pulses, scarf whips violently, jacket flutters, armor vibrates, goggles shimmer from reflected sunlight, subtle breathing visible through chest movement. 0:02-0:04 High-speed lateral tracking beside the motorcycle. Rider drifts between burned vehicles with controlled counter-steer. Rear tire flexes under load, sand rooster-tail curls outward, handlebars receive tiny steering corrections, left boot briefly skims asphalt, fingers tighten on throttle, elbows absorb vibration, cloth layers ripple independently, spent sand particles strike lens. 0:04-0:07 Rear pursuit camera. Convoy trucks emerge from dust. Mounted gunners fire in disciplined bursts. Bright muzzle flashes illuminate drifting smoke. Bullet strikes erupt around the rear tire. Rider performs a fast shoulder check, head rotates naturally, torso twists independently from hips, one hand keeps the motorcycle stable while the other smoothly draws and fires the worn shotgun then revolver. Visible recoil, shell ejection, smoke trails, realistic weapon cycling, motorcycle remains balanced through suspension correction. 0:07-0:10 Dynamic crane-to-follow transition into jump. Motorcycle launches across broken guardrails and twisted debris. Dramatic speed ramp at apex. Wheels continue rotating naturally, suspension fully extends, scarf and straps float with believable inertia, tiny debris spins through air, exhaust emits glowing sparks, rider adjusts body posture mid-flight with subtle ankle and knee corrections before landing. 0:10-0:12 Heavy landing compresses suspension. Rider accelerates beside the nearest armored truck, smoothly retrieves the satchel charge, throws it onto exposed armor while maintaining one-handed steering. Fingers release naturally, satchel tumbles with realistic weight, truck body rolls from uneven terrain, dust envelopes both vehicles. 0:12-0:15 Massive practical explosion erupts behind the rider. Expanding fireball, dense black smoke, flying steel panels, burning debris, shockwave pushes dust outward across the highway. Convoy vehicles brake, collide, scatter and lose formation. Camera pulls into a wide aerial tracking shot as the Rider accelerates toward the blazing sunset, dust trail glowing gold beneath long shadows. CINEMATOGRAPHY Ultra-photorealistic practical action filmmaking, Panavision anamorphic, 35mm and 50mm lenses, alternating Ru

Hundreds of colossal stone giants emerge from beneath the earth exactly at sunrise. Mountains crumble from their shoulders while forests slide from their backs. Tiny armies gather at their feet. The title 'MORNING' appears carved across the chest of the central giant in enormous ancient cinematic serif typography. Unreal scale, golden sunrise, volumetric lighting, IMAX, photorealistic masterpiece.

Create a high-end cinematic sports commercial set on a professional outdoor tennis court during golden hour. A confident professional female tennis player wearing a premium modern tennis outfit, visor, wristbands, and performance shoes prepares to serve. Begin with a smooth cinematic tracking shot following her from a medium angle as she tosses the ball and starts her service motion. The exact instant the racket strikes the tennis ball, transition seamlessly into a dramatic overhead top-down shot. The tennis ball fills the foreground, appearing extremely close to the camera, while the player remains directly beneath it, perfectly frozen at the peak of her serve. Time completely freezes. As the world remains suspended, the camera moves fluidly around the frozen scene, capturing premium commercial-style macro close-ups of the apparel, fabric texture, stitching, logo, visor, wristbands, racket strings, shoes, and other athletic details. Use luxurious product lighting, shallow depth of field, crisp textures, smooth cinematic camera movement, and polished sports-commercial cinematography. Time resumes naturally. She completes the serve, lowers her racket, and walks forward with quiet confidence. The camera instantly whip-pans across the court, following the tennis ball's trajectory. On the opposite side of the court, dozens of tennis balls rapidly fly into place, precisely arranging themselves to spell **"COMTEX"** in large uppercase letters. End with a clean overhead hero shot showcasing the completed brand name made entirely of tennis balls. Use premium commercial pacing, seamless transitions, realistic ball physics, dynamic camera work, cinematic sound design, HDR lighting, natural motion blur, ultra-realistic textures, and luxury sports-ad production quality throughout.

Keep the exact character from the reference. Photorealistic face only; everything else in premium anime style. She wears a futuristic white astronaut suit with yellow accents. \n\nPlace the provided Nemo logo as a clean embroidered mission patch centered on her chest, integrated naturally into the suit. She stands on the Moon, gazing into her helmet visor as if it were a mirror. She never looks at the camera, never smiles, maintaining a calm expression of quiet awe. \n\nOnly subtle breathing, tiny head movement, gentle hair motion, and at most one slow natural blink near the end. Crystal-clear visor with realistic reflections. Behind her is a cinematic deep-space sky with thousands of bright stars slowly drifting and twinkling, multiple graceful shooting stars leaving glowing trails, and Earth glowing in the distance. \n\nSmooth cinematic dolly-in with a subtle orbit, HDR lighting, ray-traced reflections, shallow depth of field, IMAX, masterpiece, ultra-detailed. \n\nNegative Prompt: Cartoon face, anime face, smiling, talking, looking at camera, exaggerated blinking, exaggerated motion, camera shake, distorted anatomy, extra fingers, warped helmet, blurry face, low quality, static stars, distorted logo, oversized logo, watermark, text, artifacts.

Extreme close-up of a white spacesuit shoulder inside a dim spacecraft cabin, slowly rotating into frame. A circular embroidered mission patch stitched onto the shoulder reads "HIGGSFIELD" in clear capital letters across its face, the thread texture and raised edge of the badge sharply visible. The camera pulls slowly back and racks focus to the astronaut's helmet visor, where the curved blue edge of a planet slides across the reflective glass. Slow dolly-out with a rack focus from the patch to the visor, close-up opening to medium close-up. Cool blue instrument light from below mixed with hard white sunlight cutting in from a porthole. Photorealistic hard-science-fiction look, ultra-detailed fabric and glass, shallow depth of field. Audio: low hum of life support, soft mechanical switch clicks, amplified breathing inside the helmet, a sparse sustained low string note.

Create an authentic 15-second analogue home-video recording filmed with a handheld Hi8/VHS-C camcorder. The camcorder is never visible. The viewer only sees the recorded footage, as if watching a digitized family videotape. This is not a vlog, commercial, or cinematic film, just a friend casually filming another friend. Use genuine analogue camcorder aesthetics: soft SD resolution, VHS grain, faint interlacing, analogue colour bleed, mild tracking noise, autofocus breathing, slight exposure fluctuations, imperfect handheld shake, natural motion blur, washed-out colours, and subtle tape hiss. A young Korean woman in her early twenties wears a pastel fitted crop top, mid-rise vintage blue denim shorts, white ankle socks, classic white sneakers, a lightweight windbreaker tied around her waist, and carries a simple canvas shoulder bag. The scene takes place beside the Han River on a warm summer afternoon. She walks naturally toward the riverside railing before stopping to admire a massive floating HIGGSFIELD installation in the river. The word HIGGSFIELD is built from giant three-dimensional block letters, mounted on a realistic floating platform. The camera gently pans between her and the installation, keeping both in frame. She smiles naturally, takes a few steps closer, then notices the unseen camera operator. She laughs softly, gives a warm wave, points excitedly toward the giant HIGGSFIELD sign, gives a subtle thumbs-up, then looks back at the river with a smile before continuing her walk. One continuous 15-second recording with no cuts or transitions. The HIGGSFIELD installation must stay prominent, crystal clear, and naturally integrated into the environment. Include realistic ambient sounds such as wind, birds, distant conversations, bicycles, and water.

0–2s: Close-up of a sizzling steak on the grill. Text on screen: "Hungry?" 2–5s: Quick cuts of a juicy burger, fresh pasta, crispy fries, and a colourful dessert being served. Voice-over: "Every bite is made fresh, packed with flavour, and worth coming back for." 5–8s: Friends laugh, toast drinks, and enjoy their meals in a warm, inviting restaurant. Voice-over: "Great food. Great moments. Every day." 8–10s: Restaurant logo appears with a hero shot of the signature dish. Voice-over: "Taste the difference. Visit us today!" Text on screen: "Fresh • Delicious • Unforgettable"

10-second vertical 9:16, American street realistic twist advertising short drama. iPhone native camera first-person tracking shot, the photographer chases character @1 like a passerby from behind. Auto exposure, auto white balance, auto focus, uncolored, no filters, no beauty enhancement. Realistic walking ups and downs, tracking jitter, breathing sensation, turning motion blur and slight focus delay. Exterior of an ordinary American city commercial building, gray metal walls, red brick footings, narrow sidewalks, gravel landscaping belts and lawns. Background remains clean, no brand signs appear. [Character Lock] The only live person on site is the preset character @1 Ms. Clothing: Image_20260719093342_1234_183. Face, hairstyle, figure and full set of preset clothing strictly refer to @1, remaining consistent throughout. Character @1 always walks forward along the sidewalk in the first half. When she is patted on the shoulder and given a warning for the first time, she cannot stop completely; she can only slow down slightly and turn her upper body to look back. She can only officially stop after the second shoulder pat. The photographer's face is never shown; only a right hand of an adult passerby is allowed to briefly enter the frame. The palm can only lightly tap @1's upper back or shoulder. [Handheld Signboard] The signboard is not a wall sign but a horizontal handheld hard board that can be lifted with both hands. Dimensions approx 60cm wide, 42cm high, 5mm thick, black lightweight rigid foam board, significantly smaller than @1's upper body width, and can be completely hidden behind her. The front strictly uses the black-gold advertising reference image provided by the user hf_20260723_023253_17d01f18-b189-45de-be9c-f30db6d27116. The character in the advertisement is only a flat printed portrait and cannot become a live person. Her left hand is naturally hidden behind her back, secretly holding the signboard, but the signboard is completely covered by her body and cannot reveal any edges, golden patterns or text in advance. [Strict Time Split-screen] 00:00—00:01.30 @1 Ms. faces away from the camera, walking forward continuously along the sidewalk beside the building exterior wall. Her steps are natural, and her hair and clothes sway slightly with her steps. The first-person photographer speeds up to chase from behind. The camera shows obvious but realistic ups and downs, and the sound of accelerating steps and breathing can be heard. The photographer catches up behind her, and the right hand extends from the bottom right of the frame, lightly tapping her right shoulder. 00:01.30—00:03.50 @1 Ms. does not stop. She only slows down slightly, her body continues to move forward, while turning her upper body from the right shoulder direction to stare at the camera. Her eyebrows tighten, and she uses her right index finger to point back at the chasing photographer, warning seriously in Mandarin while walking: "Touch me again, and you'll know who I am!" She cannot stand still; her feet must continue to step forward alternately. After the warning ends, she withdraws her finger, turns her head back to the front, and resumes her original walking speed. 00:03.50—00:05.10 @1 Ms. continues to walk forward with her back to the camera, not waiting for the photographer or looking back. The photographer pauses for half a second and then speeds up again to catch up. The camera produces more obvious chasing jitter, and the right hand extends to the same shoulder position again, lightly tapping a second time. 00:05.10—00:06.40 After being patted the second time, @1 Ms. takes one more step forward and then suddenly stops one foot firmly, coming to a complete stop for the first time. She quickly turns to face the camera. The photographer is startled and instinctively retreats two steps, causing obvious camera shake and expanding to a medium full-body shot. At the same time she turns, @1 Ms. suddenly pulls out the horizontal signboard from behind. The signboard must realistically slide out along her left waist, not appear out of thin air. Her right hand immediately catches the bottom right corner of the signboard, and both hands quickly flip it, holding the front of the signboard towards the camera at chest level. Pulling, flipping, and catching with both hands are continuous and fast. The hard board has realistic weight and inertia, shaking once slightly when it stops. 00:06.40—00:08.30 @1 Ms. holds the signboard steadily with both hands. Her previously serious expression suddenly turns into a confident, friendly, and slightly smug commercial smile. She looks at the camera from above the signboard and announces loudly in Mandarin: "I am the official brand ambassador for JOHN 87445528!" While speaking, the signboard always faces the camera, does not rotate left or right, and does not block the brand text. 00:08.30—00:10.00 @1 Ms. remains standing, slightly handing the signboard towards the camera, forming a natural advertising display motion. The camera moves slowly from her face down to the signboard, auto-focusing and locking on the black-gold pattern and "JOHN87445528". The signboard remains complete and clear for at least the last 1.5 seconds, with her showing a satisfied smile from above the signboard as the scene ends. [Sound] Only real ambient sound: outdoor breeze, distant vehicles, continuous walking, photographer's accelerating steps, slight breathing, two shoulder taps, clothing friction, sound of pulling and flipping the signboard, and @1's original Chinese voice. No background music, no narration, no subtitles, no laughter effects. [Hard Constraints] After the first shoulder pat, @1 Ms. cannot stop or say lines standing still; she must turn back to warn while walking; after the warning, she must turn her head and continue walking forward. Only after the second shoulder pat can she stop completely, turn around and pull out the signboard. The signboard cannot appear in advance, cannot be generated out of thin air, and cannot become a wall sign, electronic screen, paper or oversized billboard. Dogs, cats and other animals are strictly prohibited. The only live person in the entire scene is character @1. No subtitles, titles, watermarks, media logos, platform interfaces or extra text; except for "JOHN87445528" on the signboard.

15 seconds, vertical screen 9:16, iPhone native camera texture, single continuous shot, one-take, no editing, no jump cuts, no transitions. The reference video is only for referring to the rhythm and camera movement of "front-side floating performance in the first half, right-circling reveal of the mechanism in the second half," not copying the venue, landing stand, top subtitles, watermark, or characters from the reference video. The scene is an outdoor city sidewalk during the day, bright but not overexposed, with soft natural daylight, a real smartphone snapshot texture. Two adult women, like close friends, are playing a light prank on the roadside. Sidewalk is smooth, clean, and open; the main background is a non-reflective light-colored building wall and a small amount of natural greenery, no street-facing glass windows, mirrored doors, or reflective metal walls. Occasionally one or two ordinary adult passersby walk naturally across in the distance, at least 4 meters away, not stopping, not watching, not looking at the camera, not imitating actions, and not crossing the core performance area of the coffee cup, boom arm, @1, and @2. The core props of the frame are always only one coffee cup, one gray horizontal boom arm, and one white slip of paper. [The only physical mechanism set up from the first frame] Exactly one rigid straight gray horizontal boom arm, parallel to the floor, extends forward from @2’s waist toward the camera, with the coffee cup fixed at its front tip. The same boom already exists at frame 1 and never appears, grows, extends, changes direction, or changes shape later. The only gray boom arm is about 80 cm long and 2.5 cm in diameter, always straight and parallel to the ground, maintained at height between @2's waist and chest. The front end of the boom arm only connects to the exact center of the back of the coffee cup, with the connection point completely hidden behind the cup; no short poles or joints protrude from the left or right sides of the cup; the back end is completely covered and fixed by @2's forearm and sleeve held close to the abdomen. The boom arm does not touch the ground, wall, or ceiling. The space between the bottom of the cup and the ground is always clearly visible and empty. At the opening, the camera optical axis, the center of the coffee cup, the long axis of the gray boom arm, and @2's torso center are strictly on the same depth-of-field straight line. The coffee cup completely hides the boom arm behind it from the front, so even though the mechanism exists from the first frame, it is completely invisible from the front; it is only gradually seen later due to parallax as the camera circles to the right, not because a new pole is generated midway. A medium blue and white takeout coffee cup, royal blue body, white lid, only the clear white English word "JOHN" in the center of the cup body. No real brands or antler designs. The cup stays fixed in the same spatial position throughout, remaining vertical, same size, not drifting, rotating, or deforming. There is only this one cup in the entire film. The white slip of paper is 14 cm wide and 10 cm high, already fixed to the bottom side of the middle section of the boom arm with two short pieces of transparent tape from the first frame. The top edge of the paper is flush with the arm, and the vertical plane of the paper is parallel to the long axis of the boom arm: from the front at the start, only the very narrow edge of the paper is visible; as the camera circles to the right, the front of the paper naturally unfolds with parallax. The paper has only two lines, the first line must not be broken: Your Ad Space hf_20260723_023253_17d01f18-b189-45de-be9c-f30db6d27116 @1 wears @1 reference clothing, always stands in the left third of the frame, feet stationary, body and head never crossing the center line of the screen, never walking in front of the coffee cup or between the cup and @2. @1 only extends her hands from the left toward the top and bottom of the cup for performance; her palms are always about 10 cm from the cup, not touching it. She is playful, lively, smart, and agile like she's teasing a friend on the street: naturally relaxed with a slight side turn, soft wrist flick, fingers slowly spreading. Her eyes move naturally between the cup, @2, and the camera, occasionally raising an eyebrow, showing a mischievous smile she can't help but let out, and giving the camera a short playful wink when successfully demonstrating "levitation." Not a formal commercial pose, not stiff, not vulgar, no exaggerated twisting, no funny faces. @2 wears @2 reference clothing, feet stationary, always standing on the same depth-of-field axis directly behind the coffee cup. @2 hides the back end of the boom arm with her forearm and sleeve held close to her abdomen, keeping the arm fixing the mechanism, her body, and position stationary throughout; however, in the first 5 seconds when her face is still clearly visible, she acts like she's playing along with her friend's prank: first staring seriously at the cup, then eyebrows raised, eyes slightly widened, pretending to be surprised despite knowing the mechanism; briefly pursing her lips to hold back a laugh when seeing @1 blow air, then looking at @1 and the camera with an innocent humorous gaze that says "almost got me." There is a familiar, relaxed rapport between them. @2 only performs facial and slight eye changes, not speaking, not opening her mouth wide, not shaking her head, not shrugging. After 5 seconds, the camera zooms in, and @1 and @2's faces are naturally cropped out of the frame. [Strict continuous timeline] 0—2 seconds: Front medium wide shot, no camera horizontal movement. The cup is in the center of the frame, @1 is fixed on the left, and @2 is directly behind the cup. The complete sidewalk pavement under the cup is clearly visible. @1 has one hand above the rim and one below the bottom, no contact; she first looks seriously at the cup, then looks at @2 and the camera with the corner of her eye, raising an eyebrow with a playful expression of holding back a laugh. @2 cooperates with her friend, first staring seriously at the cup, then slowly raising eyebrows and widening eyes, putting on a humorous reaction of "wow, it actually floated." Neither of their bodies or feet moves. A passerby can naturally pass in the distance without looking at them or entering the main axis. 2—5 seconds: The camera remains in the same front position. @1's feet and body don't move, only performing three sets of clear, slow hand movements: the right hand passing horizontally over the cup rim; the left hand passing horizontally under the cup bottom; hands spreading out to show they are empty. She acts as if intentionally teasing @2, with the corners of her mouth starting to turn up; then she leans in from the left to blow air gently on the cup rim, looks at @2, then gives the camera a short playful wink, and finally frames the cup with both hands (one above, one below) with a small proud smile that says "How about that, pretty good, right?" @2's body and the arm fixing the boom stay absolutely still, only responding with expressions: first looking back and forth between the cup and @1, pursing lips to hold back a laugh when @1 blows air, and finally looking at @1 with an innocent humorous gaze that says "I'm just quietly watching you perform." They act like familiar friends naturally joking, not actors reciting lines. Throughout the 3 seconds, the cup, @2, and the camera optical axis remain collinear, with the boom completely hidden by the cup. 5—7 seconds: The camera moves slowly forward along the optical axis, no horizontal movement, no rotation, pushing from medium wide shot to a close-up about 45 cm from the cup. Autofocus locks on "JOHN." @1 and @2's faces gradually leave the frame, leaving only the cup, @1's hands, a small amount of clothing edges, and the empty ground under the cup. The boom is still invisible due to collinear obstruction. 7—10 seconds: The camera rotates slowly and evenly to the right about 70 degrees around the coffee cup as a fixed center at the same height and distance, taking a full 3 seconds, no camera whip. Neither the cup nor the boom moves. At 7—8 seconds, only a small section of the gray crossbar behind the cup is seen due to slight parallax; at 8—9 seconds, about half is seen; at 9—10 seconds, the same full boom arm is seen extending straight from the back of the cup toward @2's abdomen. It appears as a horizontal line in the frame, and the paper also gradually unfolds from its narrow edge. Faces no longer in frame. 10—12 seconds: Maintaining the same shot, the camera moves slowly along the side of the boom arm from the front of the cup toward the middle section. The coffee cup naturally retreats to the left edge, the only gray crossbar runs through the center of the frame, and the front of the paper gradually turns toward the camera. The focus shifts smoothly from the bar to the paper; the paper doesn't appear suddenly. 12—15 seconds: The camera reaches a close-up of the paper and stops moving. The paper occupies about 50% of the frame width, the gray boom arm runs horizontally behind the paper, and only a small part of the coffee cup remains on the left edge. The two lines of text are stable, upright, and clearly legible: "Your Ad Space / hf_20260723_023253_17d01f18-b189-45de-be9c-f30db6d27116."

[PROJECT TYPE] Text-to-video. 30-second dialogue scene. Vulnerability and fear.\n\n[Global Setting] Modern bedroom at night. Soft warm lamplight. Both lying in bed. \nIntimate but vulnerable atmosphere. Quiet, introspective setting.\n\n[Character 1 The Woman] Mid-30s, dark long hair, wearing soft shirt. Calm but \nvisibly carrying emotional weight. Eyes reflect vulnerability and fear.\n\n[Character 2 The Man] Mid-30s, short dark hair, wearing casual shirt. Warm \nexpression, patient, listening presence.\n\n[Opening, 0–3s] Wide shot of bedroom. Both lying in bed facing each other, close \nbut not touching yet. Woman stares at him with fear in her eyes. Long pause. She \ntakes a breath and says quietly: "I'm scared you're going to leave me." Her voice \nis small, trembling, raw with vulnerability. Man doesn't respond immediately \njust holds her gaze.\n\n[Reaction Moment, 3–10s] Man's expression softens with tenderness. He reaches out \nslowly and touches her face with his hand, thumb brushing her cheek gently. He \nsays quietly but firmly: "I'm not going anywhere." Woman's eyes fill with tears. \nHe continues: "I promise." His voice is steady, certain, safe. Woman nods slightly, \ntears falling.\n\n[Connection, 10–20s] Man moves closer, pulls her into his chest. Woman wraps her \narms around him, holding tight. They hold each other in silence for a moment. \nThen she whispers into his chest: "Don't ever leave me." Man responds softly, \ninto her hair: "Never. I've got you." He holds her tighter. Both breathing slowly, \nfinding calm in each other's presence. Lamplight catches their faces vulnerable \nbut safe.\n\n[Final Moment, 20–30s] They pull back slightly to face each other again. Woman's \ntears are still visible but her breathing is calmer. Man wipes her tears gently \nwith his thumb. He looks directly into her eyes and says softly: "You're safe with \nme. Always." Woman nods, fear slowly releasing from her face, replaced by trust. \nShe leans forward and kisses him soft, tender, vulnerable. They hold the kiss \nfor a moment. Pull back slightly, foreheads touching, both eyes closed. Breathing \ntogether in shared safety and love.\n\n[Audio] Soft ambient sound gentle breathing, heartbeats subtle underneath, \nbarely perceptible, intimate. Maybe one sustained soft string note or piano pad. \nVery minimal. Mostly silence and breathing. The quiet is sacred. No dialogue \noverlap space between words.\n\n[Tone] Intimate, tender, vulnerable but safe. Fear acknowledged and met with love. \nDeep emotional connection. This is about being truly seen and accepted at your \nmost fragile.\n\n[Consistency]\n— Same couple throughout\n— Warm soft bedroom lighting\n— Intimate bed setting\n— Emotions: fear → acceptance → safety → trust\n— Physical connection: eye contact → touch → embrace → forehead touch\n— Both equally vulnerable and present

[Generation Goal] Cinematic realistic texture, pure ancient Chinese Xianxia aesthetic. A collision between the solemnity of traditional martial arts epics and a deadpan "sect chore" style comedy, featuring: Restrained deadpan performance Three-beat progressive structure Clear setup and payoff Arri Alexa cinematic texture Clear and stable facial micro-details Fine film grain Natural volumetric lighting The core comedy must be understandable at first glance: Expert challenges -> Two true experts pass the chore to each other -> Master announces a 10% monthly allowance increase -> Both instantly fight to take the challenge -> Enemy starts to feel afraid. All comedy stems only from the characters' own choices, attitudes, dialogues, and reactions. The environment must never create accidents, trigger laughs, or change the plot. [Reference Responsibilities] @Image 1: Strictly lock Character ID A | Sword Fairy Senior Sister. @Image 2: Strictly lock Character ID B | Junior Sister. The set of background and location reference images uploaded this round collectively determine the environment's DNA. Before formal composition, silently integrate: Real terrain Architectural language Spatial scale Material age Vegetation Water bodies Weather Movement of clouds and mountain mist Main light direction Reflective relationships Atmospheric depth Realistic character walking routes Replan into a unique, complete, and unified new space for this round, rather than mechanically copying a single reference image. Background elements like water, mist, vegetation, banners, clouds, distant ordinary disciples, and ambient sound continue to function naturally but remain narratively neutral. [Character Settings] Character ID A | Sword Fairy Senior Sister | @Image 1 25–30 year old East Asian female, oval face, dark almond eyes, fair natural skin tone, long black hair partially tied up with a white jade hairpin, tall and slender figure. Fixed Appearance: White embroidered silk Hanfu, translucent layered wide sleeves, silver waistband, jade pendant, white cloth boots, a unique silver longsword. Performance Core: Dignified, calm, deadpan. Truly does not want to take this "chore" in the first half; instantly enters a true master state after hearing about the allowance increase in the second half. Character ID B | Junior Sister | @Image 2 20–25 year old East Asian female, rounded and lively face shape, black hair braided, petite build. Fixed Appearance: Blue-green linen Hanfu, dark belt, wooden hairpin, black cloth shoes, a unique dark steel sword. Performance Core: Quick reflexes, sharp-tongued, highly synchronized with Senior Sister; state changes must be almost simultaneous with Senior Sister. Enemy Swordsman One male. Extremely imposing aura, seriously here to challenge. Initially thinks he is about to face a legendary duel, then gradually realizes these two treat him like an unwanted sect chore. Elderly Master One male. Always several steps behind the two women, sitting or standing naturally. Calm, authoritative, unfazed. Only responsible for changing the rules and delivering the final verbal blow. [Shot Design] Shot 1 | 0–5s | Full Shot / Long Shot Adaptive choice of the most logical courtyard, platform, or open area based on the real space of the reference images. The enemy swordsman strides into the frame with great momentum, shouting loudly: "Who is the strongest here? Come out and fight!" The elderly master, a few steps behind the two women, replies extremely calmly: "Either of them will do." The enemy looks back and forth between the Senior Sister and Junior Sister: "Who goes first?" A complete pause. The same white-clad sword fairy and the same green-clad junior sister simultaneously look away at the exact same moment, as if neither heard a thing. Action Points: Enemy entry must have a true master's challenge momentum. Master's reply must be extremely flat. The two women looking away simultaneously is the first visual punchline. No exaggerated facial expressions; they just need to naturally pretend they didn't hear. Shot 2 | 5–10s | Medium Shot / Cowboy Shot Maintain the same characters, clothing, swords, enemy, master, and identical geographic space. Senior Sister says calmly: "I'm off duty today." Junior Sister immediately follows: "I guarded the gate last night." The master slowly raises his eyes and asks seriously: "When did the sect have shifts?" A half-beat pause. Senior Sister says without changing expression: "Just now." Junior Sister nods extremely seriously in total agreement. The enemy's originally majestic expression finally collapses: "I am here to challenge you, not to watch you push chores around!" Action and Performance Points: Senior and Junior Sisters must act like they are seriously discussing work schedules. "Just now" must be delivered very naturally. Junior Sister's nod is the second layer of the punchline. The enemy's breakdown comes from his epic challenge being downgraded to a "chore nobody wants." Background continues natural movement but does not intervene in the plot. Shot 3 | 10–15s | Close-up → Extreme Close-up The master sighs softly, as if used to this, and says casually: "Whoever fights gets a 10% increase in this month's allowance." Absolute silence for a half-beat. In the next instant— The same Senior Sister and Junior Sister completely change their state: Simultaneously step forward Simultaneously place hands on their respective sword hilts Eyes sharpen simultaneously Body postures instantly enter true master combat mode They say in unison: "I'll do it." But they don't look at the enemy first; they immediately turn to look at each other. Junior Sister: "Me first." Senior Sister (deadpan): "Weren't you tired?" Junior Sister (without blinking): "Suddenly energized." Extreme Close-up: The previously aggressive enemy starts to retreat very slowly, tentatively raising a hand: "Maybe... I'll come back another day?" The Senior Sister and Junior Sister snap their gaze toward him at the exact same moment, shouting in unison: "Stay put." Focus shifts to the master in the back. He shows no surprise, just calmly takes a sip of tea, as if he knew all along that a 10% raise would have this effect. Precise cut to black. [Comedy Rhythm Constraints] Three phases must be very clear: Beat 1 | Pushing the chore: True masters don't want the challenge. Beat 2 | Rule change: Master only adds a "10% allowance." Beat 3 | Identity reversal: The two go from dodging to fighting over it; the enemy goes from challenger to the one who wants to leave most. The point isn't "greedy exaggeration," but rather: The change in attitude is too professional, too synchronized, and too matter-of-fact. [Consistency and Action Requirements] Stable character identities throughout (face, hair, clothing). Stable ownership of the silver sword and dark steel sword. Stable positions for the Master and Enemy. Synchronized actions of the two female leads must be clear and readable (looking away, stepping forward, gripping hilts, speaking in unison). Synchronized nodes must be precise but not mechanically stiff. [Environment Requirements] The environment is a functioning real world (flowing water, moving mist, swaying banners, moving clouds). The environment must NOT react to the punchlines (e.g., no sudden wind after the allowance reveal, no dramatic lighting changes at the punchline, no synchronized background crowd reactions). [Technical Specifications] Strict total duration: 15 seconds. 16:9 widescreen. Three continuous clear shots. Native synchronized Mandarin dialogue with precise lip-sync and clear comedy pauses. Arri Alexa cinematic texture. Clear facial details. Realistic silk fabric and hair movement. Natural parallax between foreground, midground, and background. Real spatial ambient sound. No subtitles. [Negative] blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, architecture mutation, lighting jump, glitching cuts, disappearing props, duplicated swords, changing sword ownership, extra foreground characters, exaggerated slapstick acting, exaggerated greed reaction, cartoon comedy, environment triggering joke, sudden wind at punchline, dramatic lighting change at salary reveal, synchronized background crowd reaction, unstable enemy position, unstable master position, broken eyelines, broken geography, random camera jump.

Seedance Prompt | Price Hiking I. Core Positioning Cinematic realistic texture, pure ancient style Chinese Xianxia aesthetics. Overall style: Sophisticated deadpan ensemble comedy Classical martial arts character blocking Gradually escalating negotiation rhythm Rule of three Precise reaction shots Arri Alexa cinematic texture Clear and stable facial micro-details Fine film grain Natural volumetric light The comedic core of this segment is a clear reversal of identity and stance: The enemy attempts to bribe the Junior Sister with silver to reveal the Senior Sister's weakness. In the first few seconds, make the audience briefly suspect if the Junior Sister is actually going to take the money. Then it is revealed that these two women are actually working in perfect tacit understanding to "hike the price" for the enemy. All supporting characters serve only as: Witnesses to the comedy Rhythm boosters (finishing blows) The true main subjects always belong to: Image 1: Senior Sister (Sword Immortal) Image 2: Junior Sister II. Reference Controls 1. Character Identity Anchors @Image 1 strictly locks Character ID A: Sword Immortal Senior Sister @Image 2 strictly locks Character ID B: Junior Sister 2. Environment Reference All background and location reference images uploaded in this round jointly determine the same environmental DNA. Before formal composition, silently integrate compatible: Real terrain, architectural language, spatial scale, textures, vegetation, water bodies, weather, mountain mist, clouds, main light direction, reflection relationships, atmospheric depth, and realistic walkable paths. Then re-plan into a unique, unified, physically logical, and spatially continuous new space. 3. Environmental Principles The environment must remain vivid from start to finish but be absolutely neutral in the narrative. Elements allowed for natural continuous movement include: Wind, water bodies, mountain mist, vegetation, reflections, distant disciples, and spatial ambient sound. But must adhere to the following principles: The environment cannot trigger the plot, solve problems, create comedic bits for characters, or interrupt negotiations. It only provides a sense of real-world presence. III. Character Settings Character ID A | Sword Immortal Senior Sister Same as @Image 1. 25–30 year old East Asian female, oval face, fair skin, dark almond eyes, long black hair half-up fixed with a white jade hairpin, tall and slender. Fixed Look: White embroidered silk Hanfu, translucent wide sleeves, silver waist belt, jade pendant, white cloth boots, a single silver long sword. Temperament: Calm, dignified, logically stable, deadpan, unblushing in absurd situations. Key Performance Requirements: Must turn head very slowly to look at the Junior Sister when she says "Too little." Must say "Eighty" like a reasonable market price. Must be completely serious when saying "My weakness cannot be sold cheaply." When pushing silver to the Junior Sister, the attitude should be like handling official business. Character ID B | Junior Sister Same as @Image 2. 20–25 year old East Asian female, round and lively face, black hair in braids, petite build. Fixed Look: Cyan-green linen Hanfu, dark belt, wooden hairpin, black cloth shoes, a single dark steel sword. Temperament: Smart, quick-reacting, emotionally restrained, natural tacit understanding with the Senior Sister. The comedy comes from being "absurdly serious." Key Performance Requirements: Must make the audience briefly doubt her when she first observes the silver. Say "Too little" calmly and seriously, not jokingly. Nod at "Eighty" as if approving a professional quote. Weighing the silver at the end and saying "Afraid I'll lose out" must be as natural as stating a fact. Supporting Character 1 | Enemy Swordsman One male. Responsible for initiating the bribe. Initially confident, thinking he controls the situation; starts to be confused in the middle; completely frozen at the end. Key Performance Requirements: "Thirty taels, tell me her weakness" should be direct and pragmatic. At "Fifty," he should clearly feel he is adding to the stake. "Eighty! What exactly is she afraid of?" should carry the pressure of losing patience. Finally, looking at his empty hands, he should look as if he has been totally bypassed. Supporting Character 2 | Elder Master One male. Responsible only for an authoritative finishing blow at the end. Must not steal the scene. Tone should be like handling sect accounts. Key Performance Requirements: "A lesson" should sound like an objective conclusion. "Sect takes thirty percent of the silver" must sound like announcing a normal rule. Walking cannot stop. Cannot turn into an obviously comedic performance. Distant Background Characters Small-scale disciples in the distance may be present. They only provide world presence. Cannot join the main plot, obviously watch the protagonists, or react excessively to the jokes. IV. Core Props Silver Ingots: The same batch of silver, a few ingots. Clearly placed on a naturally existing stone surface between the three. Continuous and stable across three shots. Must have real metallic weight and reflection. Long Swords: Senior Sister has a single silver sword; Junior Sister has a single dark steel sword. Present as identity props throughout. Focus is on negotiation and relationship reversal, not fighting. V. Comedy Structure 1st Beat | Mislead: Enemy offers thirty. Junior Sister doesn't refuse immediately but looks seriously at the silver and says "Too little." Audience suspects a betrayal. 2nd Beat | Progression: Enemy increases to fifty. Senior Sister suddenly steps in herself to hike the price: "Eighty." The situation turns from "bribery" to "joint bargaining." 3rd Beat | The Hook: The enemy finally deals and asks for the weakness. Junior Sister answers seriously: "Afraid I'll lose out." The transaction was a hustle from the start. Authoritative Finishing Blow: Master's final lines: "A lesson." "Sect takes thirty percent." Stamps the absurd negotiation as official sect business. VI. Storyboard Script 0–5s | Full Shot or Wide Shot: Character ID A (Senior Sister) and ID B (Junior Sister). Enemy swordsman puts silver on the stone surface, saying to Junior Sister: "Thirty taels, tell me her weakness." Junior Sister observes seriously and answers calmly: "Too little." Senior Sister turns head slowly to look at her. 5–10s | Medium Shot or Cowboy Shot: Enemy adds silver: "Fifty." Junior Sister is about to speak, but Senior Sister says calmly: "Eighty." Beat of silence. Enemy: "I'm buying your weakness, and you're helping her hike the price?" Senior Sister: "My weakness cannot be sold cheaply." Junior Sister nods seriously. 10–15s | Close-up or Extreme Close-up: Enemy pushes all silver forward: "Eighty! What is she afraid of?" Junior Sister picks up an ingot, weighs it, and points to Senior Sister: "Afraid I'll lose out." Senior Sister pushes all silver to Junior Sister: "So the money stays, the person can leave." Enemy freezes: "Then what did I buy?" Master walks past: "A lesson. Sect takes thirty percent." Protagonists turn: "Master?" Enemy looks at empty hands. Cut to black. VII. Performance Rhythm Requirements Senior Sister: Completely serious throughout. The more serious, the funnier. No smiling, no explanation. Senior Sister and Junior Sister form a conspiracy. Junior Sister handles the reversal. Enemy reactions escalate from confidence to confusion. Master is calm and lethal. VIII. Camera and Spatial Requirements 16:9 Landscape. Three continuous clear shots. Arri Alexa look. Stable facial details. Natural volumetric light. Fine film grain. Precise reaction shots. Stable spatial relationships. IX. Sound Requirements Native synchronized Mandarin dialogue. Precise lip-sync. Clear comedic pauses. Realistic ambient sound. No subtitles. X. Technical Specifications Total duration: 15s. Format: 16:9. Three clear shots. Native Mandarin sync. High physical realism for fabric and hair. Organized according to Seedance 2.0 15s multi-shot audio-video generation and multi-modal reference continuity capabilities. XI. Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, background shift, glitching cuts, disappearing props

Project: Frozen Shell (15-Second Ice Cream Campaign) Format: Text-to-Video with Image Anchor (Start Frame) --- [00:00 - 00:03] THE HOOK (The Shell Crack) • Visual & Camera: Starts on the image anchor. Extremely tight macro close-up on the frozen dark chocolate shell of the ice cream bar. • Action / Movement: The 26-year-old Latina woman bites into the bar in hyper-slow motion. The hard chocolate shell cracks in intricate spiderweb patterns before a piece breaks off cleanly. • Audio / Sound Design: A sharp, wooden "CRACK-CRUNCH" sound effect followed by a deep sub-bass pulse. • On-Camera / Voice: None. High-contrast texture shift. --- [00:03 - 00:08] THE CREAMY DIP & DIP-FLOW • Visual & Camera: Macro shot of the interior showing dense, rich vanilla bean ice cream contrasting against the dark chocolate shell fragments. • Action / Movement: Slow-motion shot of a fresh ice cream bar being dipped into a vat of liquid chocolate, pulling out as a smooth, glossy coat quickly freezes into a matte finish. • Audio / Sound Design: Smooth, airy "WHOOSH" sound layering into a bright, summer-themed indie-pop synth line. • On-Camera / Voice: None. Focus on temperature contrast and surface transformation. --- [00:08 - 00:12] THE REFRESHMENT • Visual & Camera: Medium shot, eye-level. Golden hour sunlight highlighting her face. • Action / Movement: The 26-year-old Latina woman laughs softly, savoring the bite as sunlight reflects off her face. She looks straight at the camera. • Audio / Sound Design: Music track dips under clear vocal audio. • Dialogue (Real-time on-camera): "Worth every single bite." --- [00:12 - 00:15] THE HERO SIGN-OFF • Visual & Camera: Crisp hero shot of the ice cream bar resting on a marble surface next to vanilla beans and crushed dark chocolate pieces. • Action / Movement: A single drop of melted chocolate falls from the tip onto the marble in slow motion. • Audio / Sound Design: Upbeat final acoustic chime fade-out. • On-Screen Text / Call to Action: "Break the Cold."

Create a highly realistic 15-second video showing a young woman casually cooking alone in her home kitchen while her favorite upbeat song is playing in the background. The scene should feel like a genuine moment captured on a smartphone, not a commercial or music video. 0–3 seconds: She is naturally chopping vegetables on a kitchen counter while the music plays from a small speaker nearby. She looks relaxed and focused on cooking. Natural daylight enters through the kitchen window. 3–6 seconds: She suddenly recognizes her favorite part of the song. Her expression changes into a spontaneous smile. While continuing to cook, she starts subtly moving her shoulders and head to the rhythm. 6–9 seconds: She starts sings along the song and enjoying the song. 9–12 seconds: She stirs the food while moving naturally to the beat, briefly sings along with the music, and smiles at herself. 12–15 seconds: She tastes the food, reacts with a satisfied smile, then looks at camera and says "yummm" while holding the cooking spoon. End on a candid moment where she is genuinely enjoying herself. Environment: ordinary modern home kitchen, realistic countertop clutter, cooking ingredients, utensils, pan, small Bluetooth speaker, subtle imperfections and lived-in details. Camera: handheld smartphone footage, natural framing, slight operator movement, occasional small reframing, realistic autofocus and exposure changes. Begin with a medium-wide shot, move naturally closer during the dancing, then finish with a slightly wider candid shot. No artificial camera spins or dramatic cinematic movements. Lighting: soft natural window light mixed with normal indoor kitchen lighting. Realistic shadows, natural highlights, authentic skin texture. Performance: spontaneous, playful, relaxed, believable. Natural facial expressions and body physics. The dancing should look improvised rather than professionally choreographed. Realism requirements: photorealistic human appearance, realistic hands and fingers, accurate cooking interactions, believable food and steam movement, natural hair movement, physically correct contact with objects, consistent identity and clothing throughout. Audio: upbeat feel-good music playing naturally from the small kitchen speaker, with subtle cooking sounds underneath. Her quiet singing and laughter can be heard naturally. The music should feel like the actual source of her spontaneous dancing. Avoid: commercial-advertisement aesthetics, studio lighting, excessive beauty retouching, perfect posing, unrealistic dancing, slow motion, dramatic transitions, excessive camera movement, artificial-looking skin, exaggerated expressions, text overlays, logos, watermarks.

Create a cinematic, photorealistic lifestyle video of a young East Asian woman in a cozy modern apartment by the sea. The video has a warm, natural morning atmosphere with soft daylight coming through large windows, realistic skin texture, subtle facial expressions, natural body movements, shallow depth of field, and smooth cinematic camera motion. Scene 1 — Bedroom: A young woman with long straight dark hair, wearing a light beige/pink satin pajama set, sits on the edge of her bed looking sleepy. She gently yawns and rubs her eyes. The bedroom is minimal and modern, with a neatly made bed, wooden furniture, soft curtains, and large windows letting in diffused natural light. Scene 2 — Bed: She pulls and adjusts the duvet, then sits and stretches slightly on the bed. Capture her natural sleepy morning routine with realistic movements and a calm atmosphere. Scene 3 — Leaving Bedroom: She stands up and slowly walks toward the bedroom doorway. The camera remains cinematic and slightly distant, showing the warm wooden interior and softly illuminated bedroom in the background. Scene 4 — Cooking: Cut to the kitchen. Close-up of a black electric sandwich/waffle-style press on the kitchen counter. The woman pours smooth light-brown batter into the heated mold. Use detailed macro shots of the batter flowing into the appliance. Scene 5 — Preparing Food: She operates the sandwich maker on the kitchen counter and carefully checks the food while cooking. Show realistic hand movements, steam/heat details, kitchen reflections, and natural daylight coming through the nearby window. Scene 6 — Eating: She opens the appliance and removes a freshly cooked golden-brown waffle/pastry. She holds it with both hands, takes a bite, then smiles naturally with a satisfied expression. Visual style: photorealistic, cinematic lifestyle commercial, natural morning lighting, warm neutral color palette, realistic Asian facial features, authentic skin texture, detailed hair strands, realistic fabric physics, soft shadows, subtle film grain, shallow depth of field, professional cinematography, smooth transitions, realistic handheld camera movement, 4K quality. Camera: combination of medium shots, close-ups, macro food shots, slow push-ins, gentle tracking shots, and shallow-depth-of-field portrait shots. Mood: cozy, peaceful, warm, relaxing morning routine, premium lifestyle advertisement. Aspect ratio: 16:9 Duration: approximately 20 seconds No text, no subtitles, no watermark, no logo, no distorted hands, no extra fingers, no unnatural facial movements, no cartoon/anime appearance.

A cinematic, realistic short video of a stylish young woman with long wavy brown hair and bangs, wearing a beige trench coat over a white top and light pants, carrying a brown leather shoulder bag. She steps out of an elegant blue door of a classic European building onto a sunlit cobblestone street lined with plants and flower boxes. She walks confidently down the charming Parisian-style alley. She approaches and enters a cozy cafe with wooden doors and glass windows labeled something like “CHANEL ALEX” or similar. Inside the warm, inviting cafe with exposed beams, pendant lights, and a marble counter, she orders an “Ice vanilla latte, please” from a bearded barista in an apron. A close-up of a golden, flaky croissant on a white plate with a fork. She sits by the window, happily sips the layered latte from a clear glass mug, closes her eyes in enjoyment, and says “That’s the good stuff.” She then takes a big bite of the croissant and says “Perfect start to the day.” Final shot of her walking down a bustling cobblestone street in golden morning light, holding the coffee and half-eaten croissant, smiling and looking at the camera. Soft natural lighting, warm tones, shallow depth of field, high detail, lifestyle aesthetic,

Cinematic slow-motion food commercial of a gourmet cheeseburger being prepared. Extreme close-up of a hand lighting a gas stove with a match, blue flames igniting. A thick raw beef patty seasoned with black pepper and salt drops onto a hot black cast-iron grill pan, sizzling with sparks and smoke. Flames erupt around the patty as it sears. Melted cheddar cheese slice is placed on the juicy browned patty and starts melting. Fresh green lettuce leaves being pulled apart by hands. Floating red onion rings in the air against black background. Juicy red tomato slices stacked with water droplets flying. Thick red sauce being stirred with a wooden spoon. Sauce being spread on a soft burger bun with fingers. Onion rings placed on the sauced bun. Layers of lettuce, tomato slices, pickle slices, and red onion rings stacked. The cheesy beef patty is placed on top, sauce dripping. Final shot of the complete tall cheeseburger with sesame seed bun, melted cheese, dripping orange sauce, fresh veggies, steam rising dramatically, sesame seeds floating in the air, dark moody background, ultra realistic, high detail, mouthwatering, commercial food photography style, 8k.

[00:00 - 00:03] THE HOOK (The Clean Snap) • Visual & Camera: Starts on the image anchor. Macro tight shot of the dark chocolate bar held by the 28-year-old East Asian woman. • Action / Movement: Hands snap the thick chocolate slab cleanly in half in extreme slow motion. A tiny puff of fine cocoa dust explodes outward at the breaking point. • Audio / Sound Design: Crisp, loud, reverberating "SNAP" sound effect that echoes sharply into silence. • On-Camera / Voice: None. Pure structural precision and acoustic clarity. --- [00:03 - 00:08] THE MELT & DRIP (Rich Texture) • Visual & Camera: Macro close-up tracking shot moving along a square of chocolate as liquid caramel or melted dark chocolate cascades smoothly over the edges. • Action / Movement: Thick, glossy molten chocolate flows slowly over the rigid edges, coating the surface in a seamless, velvet layer. • Audio / Sound Design: Deep, warm, low-frequency atmospheric synth pad swells smoothly. • On-Camera / Voice: None. Focus on fluidity and luxury. --- [00:08 - 00:12] THE PERSONAL EXPERIENCE • Visual & Camera: Medium shot, eye-level. Warm, ambient lighting casting a golden glow. • Action / Movement: The 28-year-old East Asian woman places a small square on her tongue, closes her eyes, and takes a moment to let it melt, exhibiting pure indulgence. • Audio / Sound Design: Music track softens to a gentle hum. • Dialogue (Real-time on-camera): "Pure, unadulterated dark." --- [01:12 - 00:15] THE HERO SIGN-OFF • Visual & Camera: Overhead static hero shot of the broken chocolate bar sitting alongside gold foil on a dark slate block. • Action / Movement: Light sweeps smoothly across the glossy surface of the chocolate. • Audio / Sound Design: Ambient music settles into a deep, elegant final chord. • On-Screen Text / Call to Action: "Indulge the Senses."

Create a bright, premium lifestyle product video featuring colorful OLLY gummy vitamins. Start with a close-up of the gummy bottle on a clean, aesthetically pleasing surface. Slowly reveal the colorful gummies, highlighting their texture, shape, and vibrant appearance. Show a hand picking up a gummy and bringing it toward the camera, followed by a natural wellness moment. Use soft daylight, warm tones, smooth camera movements, shallow depth of field, realistic product details, subtle reflections, and a clean commercial aesthetic. End with the OLLY bottle clearly visible in a polished hero shot. High-end beauty advertisement, cinematic lighting, realistic textures, smooth transitions, 4K quality.

Create a product video ad from [image1],[image2], script "9:16, 24fps, 15 seconds. 0-3s: Close-up shot. In a dim room lit by warm ambient lamps, a stylish girl with bangs and freckles calmly puts on her over-ear headphones. 3-5s: Close-up of a finger tapping the "Play" button on a smartphone. As the music starts, the girl closes her eyes and begins swaying rhythmically to a heavy beat, fully immersed. 5-12s: Sophisticated Documentary Montage. The girl remains the absolute visual center within real-world settings: 1. Sitting on a couch at a dim party, she sways in her own world while crowds flow through the foreground and background. 2. At a livehouse venue, she sways in the center with people in the background rendered in artistic motion blur. The camera vibrates slightly with the beat, capturing her internal focus amidst external chaos. 12-15s: The camera snaps back to the girl in her room mid-sway. The background swiftly dissolves and transitions into a solid brand-red color, with a white brand logo “Apple Music”fading into the center. Negative Prompt: Any text, subtitles, watermarks, light effects, black background."

[REFERENCE LAYER] Uploaded reference images, in upload order: @Image1 — CHARACTER REFERENCE, BUNGEE STATE. The man, wearing his black bungee harness with the shirt collar closed. Identity and wardrobe reference only. @Image2 — CHARACTER REFERENCE, VILLA STATE. The same man with no harness and the shirt collar open. Identity and wardrobe reference only. @Image3 — PRODUCT REFERENCE. The NAGI beer can. Design reference only. Follow @Image1 strictly for his face, hair, stubble, build and clothing in every shot from 0s to 13s, including the black bungee harness on his waist and thighs. Follow @Image2 strictly for the same man from 13s to 18s, with no harness and an open collar. Follow @Image3 strictly for the can in every shot where it appears: the matte pale sea-glass body, the bare silver rim, lid and tab, the thin white horizontal line, the white word "NAGI", the small white "JAPANESE LAGER" beneath it, and the slender white barley ear. The man in @Image1 and the man in @Image2 are the SAME person. Keep one single identity. None of these images is a starting frame. Do not open the film on any of them. Do not inherit their composition, their angle, their crop, their flat grey background or their flat studio lighting. Take only the identity, the wardrobe and the product design from them. [ONE-LINE SUMMARY] 18 seconds | 16:9 | 24fps | 720p. Live-action beer commercial. The man from @Image1 dives head-first off a catwalk on a jungle bridge, falls into a gorge toward a single point of white light, grabs the @Image3 can resting on a crystal at the bottom, rises holding it, and arrives in a luxury villa where he finally relaxes. Cinematic live action, shallow depth of field, natural light. Fast-paced advertising edit. [GLOBAL SETUP] Environment and texture: Remote Southeast Asian jungle at morning, a vast limestone gorge filled with mist. Humid air. Emphasize extremely realistic physical texture. Visual style: Cinematic live action. Shallow depth of field. Natural light and backlight. Film grain. Camera language: Only one camera movement per shot. Never mix movements. Character: exactly as in @Image1 — lean and wiry, long narrow face, warm tanned weathered skin, narrow hooded eyes, straight nose bridge, thin lips, angular jaw, short black hair with grey at the temples, light patchy stubble, faded olive-green jungle shirt with sleeves rolled below the elbow, dark khaki cargo trousers, scuffed brown hiking boots. Preserve real fine pores and authentic skin texture. From 0s to 11s he always wears the black bungee harness from @Image1: the waist belt with its steel buckle and both thigh loops joined by flat webbing. The harness is visible whenever his body in frame. Never remove it early. From 13s onward he matches @Image2 exactly: no harness of any kind, collar button open. Product: exactly as in @Image3. Never change the colour, proportions or printing of the can. Falling posture rule: He always goes over the edge leading with his head and chest, tipping his upper body forward into the drop. His feet leave the catwalk last. He never steps off feet-first, never drops feet-down, and never falls with his legs below him. He is head-down and upper-body-first from the instant he leaves the catwalk until he reaches the bottom. Can orientation rule (important): While it rests on the crystal, the can stands UPSIDE DOWN in world space — its lid resting flat on the crystal surface, its base pointing up toward the sky, and the word "NAGI" reading upside down to the world. This is intentional; do not correct it. Because the man arrives head-down, the can appears upright from his point of view, so he grips it in a completely natural way, thumb toward the lid. When his body then rotates head-up, the can rotates with his hand and ends up correctly oriented, lid up, held naturally. His grip never changes or re-adjusts during this rotation. Core of the performance: Falling = everyday stress. Rising = release. The film opens on a tense face and answers it with the same face released. VOICE-OVER RULE (read carefully): There are exactly two spoken lines in this entire film. Both are voice-over, spoken by an unseen narrator with a calm, low, warm male voice, unhurried and quiet. Nobody on screen speaks these lines. The man's mouth does not move for them and there is no lip sync anywhere in the film. The two lines are: at 8.5s — 「うるさい毎日に、」 at 15.5s — 「静けさを、一本。」 These two short lines are the complete and total amount of speech in the film. Add no other narration, no other dialogue, no improvised words, no extra sentences, and do not repeat these lines. The only other human sound in the film is his scream during 4.5-7.5s, which is not speech. [TRANSITION SETUP] Forbid hard cuts. Forbid objects appearing out of nothing. Maintain the breathing of the camera in harmony with the shot lengths, and leave enough time for each transition. This film has exactly eight shots. Render all eight. Do not merge or drop any of them, including the short opening shots. [TIMESTAMPS] 0-1s [THE FACE — EXTREME CLOSE-UP] Action: The film opens on this shot. Extreme close-up of the @Image1 man's face filling the frame. Sweat beads at his temple and runs down. His pupils are wide. His jaw is tight and a muscle flexes at the hinge. He blinks once, hard. His breathing is shallow and fast through slightly parted lips, and his eyes flick downward and away, unable to hold still. Loose strands of hair move in the wind at the edge of frame. Intent: Put the viewer inside his stress before showing them anything else. This is dread, not adventure. Camera: Extreme close-up, locked off. No camera movement at all. Color: Cold blue-green. Desaturated. Hard side light from a low sun. Audio: A deep slow human heartbeat, low and very close, as if heard from inside his own chest. This is the only prominent sound. Beneath it, only his shallow breathing and the faint sound of wind moving through the gorge. There are no city sounds of any kind: no traffic, no cars, no crowds, no station announcements, no phones, no machinery, no distant voices. No background music. No voice-over in this shot. On-screen text: none. -> Natural cut 1-2.5s [THE VOID — STRAIGHT DOWN] Action: Cut to a camera positioned high above, looking straight down. Directly below, a narrow open steel-grating catwalk juts out from a rusted steel truss bridge into empty air, a thin bright line across an enormous field of drifting mist. The @Image1 man is a small figure standing near its tip, wearing his harness, the bungee rope curling behind him along the grating. Below and around the catwalk there is nothing but the mist-filled void of the gorge. Intent: One pure image of how high and how alone he is. No face, no acting — the thin line of metal and the void do the work. Camera: Directly overhead, looking straight down, locked off. No camera movement at all. Color: Cold blue-green. Heavy mist. Almost abstract. Audio: Only the heartbeat, close and loud. Wind moving through the gorge. The faint creak of steel grating. No city sounds. No background music. No voice-over in this shot. On-screen text: none. -> Natural cut 2.5-4.5s [THE JUMP — FULL BODY, SIDE ON] Action: Cut to a side view of the catwalk from across the open air, at roughly the same height as the man. His full body is clearly visible, matching @Image1 exactly, with the black bungee harness plainly visible on his waist and thighs and the rope running from it back along the grating. So is the structure he st