Music Video

Use the character from the reference image as the protagonist and maintain them as the same person throughout. The reference image is used only as design material to fix the face, eye shape, iris color, highlights, lashes, eyebrows, nose, mouth, cheeks, jaw, age, hairstyle, color, clothing, ornaments, build, and character color. Do not reproduce the background, room, furniture, or layout from the reference image. [Worldview] A fantastical high-density high-speed magic battle set in floating ruins submerged in moonlight. A high-end 2D hand-drawn anime MV style. Fixed 9 cuts with no additions. High-output magic light explodes from frame 1; no long waits or fades. Each cut clearly shows one main action, connecting magic media, circles, teleportation, defense, impact, and character faces at high speed. [Character Rules] Maintain character consistency throughout. No switching clothes or hairstyles. No cloning or extra people. Teleportation must show the character disappearing before reappearing as the same individual. [Equipment] Use a slender wand, sword, or rod suitable for flashing, rotating, and firing light beams. If the reference lacks one, add a single magic wand matching the character's style. No second weapons or召喚 tools. [Art Style] High-end 2D hand-drawn anime with the density and 발광 of the reference. Cinematic key animation, high-quality compositing, and transparent lighting. Fine, delicate lines—avoid heavy outlines. Shadows are multi-layered and soft. No 3DCG or plastic looks. [Scenes] Cut 1: Extreme close-up of magic media tip. Immediate explosion of character-colored light. Protagonist's eye reflected in the flash. Camera snaps back to show action pose. Cut 2: Protagonist slashes magic media. A sharp glowing arc forms a magic circle that wipes the screen to a new location in the ruins. Cut 3: Protagonist teleports to the other side. Absorbs a shadow projectile. High-speed camera angle shift. Cut 4: Rotates magic media. Three magic circles deflect projectiles and open a path of light. Cut 5: Camera enters light path. Protagonist teleports to a pillar top and aims forward. Cut 6: Nucleus fires shadow projectiles. Protagonist draws a circle to deploy multiple slowing arrays. Projectiles decompose into light shards. Cut 7: Eyes fixed on target. Magic media aligns with multiple circles and the nucleus. Camera follows the axis. Cut 8: High-density beam fires through circles, hitting the nucleus. No explosion; magic patterns spread and cracks run through the core. Cut 9: Cracks dissolve into motifs like petals or feathers. Final teleport to a hero pose near the camera. Hair and clothing flow as the scene ends.

Close-up of the face from the current image. The singing voice is bright and light. Sing "Thank you God, everyone" to the notes Do-Re-Mi-Fa-So-La-Ti-Do. Accompanied by synthesizers and cymbals. Red, white, and yellow petals, light, diamonds, stars, and rainbows fly up. They spread into a heart shape, form a circle, and scatter. Gold dust dunes in the background.

Tall blue alien and beautiful lady with long wavy blonde hair in braids with diamond earrings a diamond necklace wearing a black catsuit are in the mosh pit at a loud concert with fireworks on the main deck of the spaceship

Cinematic hip-hop/rap music video, photo-realistic quality, high-end, beach setting. Build the scene based on @Image1: a band performs on a golden beach with waves crashing. A lead singer passionately sings with a microphone stand on the wet sand, a guitarist stands on the left, another on the right, and a drummer sits behind the kit. The vast coastline stretches behind them with rolling waves. A giant, warm golden-hour sun sweeps across the beach, sparkling on the water with sea mist and salty spray in the air. The lead singer in a red tracksuit RAPs toward the camera with rhythm—mouth movements and jaw precisely aligned with every word, head nodding to the beat, driving the flow. Musicians sway to the rhythm as waves crash behind them. This is a bright, energetic rap song—fast-paced, confident, and strong beat. Hard cuts on the beat, with double contrast (shot scale and camera type) at each switch. Lyrics (the lead singer sings 'Hello' in these languages with precise lip-sync): English: 'Hello', Chinese: '你好', Japanese: 'こんにちは', Korean: '안녕하세요', Portuguese: 'Ola', Thai: 'สวัสดี', Spanish: 'Hola', Arabic: 'مرحب'. Shot 1 [0:00-0:03]: Low-angle wide establishing shot, Steadicam advances through golden sunset and sea mist as waves surge behind the band. Lyric line 1 (English 'Hello'). Hard cut. Shot 2 [0:03-0:05]: Close-up of red tracksuit singer rapping to camera, handheld whip-pan in, blurred shimmering water behind. Lyric line 2 (Chinese '你好'). Hard cut. Shot 3 [0:05-0:08]: Macro insert shot, fixed camera, guitarist's fingers moving fast on strings, sand and mist passing the foreground. Lyric line 3 (Japanese 'こんにちは'). Hard cut. Shot 4 [0:08-0:10]: 3/4 mid-shot of a musician, slow orbiting creep, instrument metal and wet highlights reflecting the sunset. Lyric line 4 (Korean '안녕하세요'). Hard cut. Shot 5 [0:10-0:13]: A musician by the water, fast tracking shot passing him, he turns to camera as a wave breaks behind. Lyric line 5 (Portuguese 'Ola'). Hard cut. Shot 6 [0:13-0:15]: Drummer by the water, handheld quick tilt up, sea breeze and spray blowing hair as he drums to beat. Lyric line 6 (Thai 'สวัสดี'). Hard cut. Shot 7 [0:15-0:18]: Tight push on red tracksuit singer in full flow, aggressive handheld, twilight sea behind band silhouettes. Lyric line 7 (Spanish 'Hola'). Hard cut. Shot 8 [0:18-0:20]: Heroic wide shot of the full band, aggressive handheld advance, band steps toward camera on beat, waves crash and golden sunset explodes into a halo behind them. Lyric line 8 (Arabic 'مرحبا'). White balance 4000K, teal-and-amber grading, 35mm, shallow depth of field, film grain, diffused sea mist, golden hour glow. Solid, high-end texture. Rhythmic rap performance, precise lip-syncing, head nodding to beat. No subtitles, no text overlays, no dissolves, no duplicate characters, hard cuts only. Total duration 20s.

They rhyme a rap. baby bop who is colored white falls down some 6767 stairs and blames blue barney the pee dino with a sausage nose. He is colored blue (). The things go 10x crazy up and down the stairs.

Genre / worldview: dreamlike fantasy MV, cinematic anime botanical garden short film. A high-density prismatic anime MV set in a fantastical wisteria garden. Color/light design features wisteria, cherry blossom, and milky white transparent backlighting. Camera language includes butterfly guide tracking and gentle orbit. The final payoff is a high-altitude reveal of the entire garden. High-quality anime. Hand-drawn drawing density of theatrical film class, delicate key animation, high-quality compositing, transparent lighting, and dense background art. Use the reference image <<<Image1>>> as the sole source. This time it's fine, as it's a faithful animation of a single illustration. Treat the girl and the butterfly in <<<Image1>>> as the same person and individual, and do not introduce additional people. [Fixed: Original image style of <<<Image1>>>] Thin, delicate light pink to light purple colored lines, transparent multi-layered thin shadows, soft backlighting on the verge of blooming, a pastel color scheme of wisteria, cherry blossom, milky white, and light mint. Prism-like glowing eyes and butterfly wings, and high-density drawing up to the hair, flower crown, lace, and wisteria trellis. Not a matte TV anime style, but somewhere between semi-transparent watercolor glaze and fine-detail illustration. Textural feel includes hair like silk thread, flowers as thin petals, and a dress of sheer lace with pearl decorations. Do not make the background low-density. [Fixed: Character identity of <<<Image1>>>] The same girl as in <<<Image1>>>. Maintain the small nose and soft mouth of the profile, transparent purple eyes, long light pink to lavender hair, flower crown, white lace dress, and slim, quiet silhouette. Do not change the hairstyle, hair color, eye color, facial features, clothes, or flower crown. Only expressions, gaze, breathing, and natural swaying of hair and dress can change. Prohibit averaging of the face, turning into another person, infantalization, photorealism, costume exchange, extra people, or clones. [Direction] Make the butterfly the guide. Do not make the butterfly giant; keep it small to medium-sized. The camera orbits the girl, following the butterfly's flight. The butterfly does not block the front of the face or overlap with the nose, mouth, or cheeks. The trail is not water drops or tears, but a flow of thin wisteria-colored light powder and thin petals. Cut 1: Profile close to <<<Image1>>>. The girl is quietly looking up in the backlighting of the wisteria trellis. The butterfly on the left flaps slowly and glows as if looking back for a moment. The camera moves forward minutely to match the butterfly's height. Cut 2: The butterfly does not cross in front of the girl's face but guides her through the outside of the flower crown and the clusters of wisteria. The camera follows the butterfly, sliding from the girl's front left to the outside of her shoulder. Details of the hair, lace, pearl decorations, and flower crown flow with foreground, mid-ground, and background parallax. Cut 3: The butterfly circles behind the girl and passes through the wisteria clusters and petals. The girl does not move her body much, only following the butterfly with her eyes. The tips of her hair and the lace of her dress sway gently in the wind, and the butterfly's light powder makes only the surrounding flowers glow faintly. Cut 4: The butterfly finishes its circle and returns diagonally above the girl. The butterfly's light is reflected small in the girl's purple eyes, making for a short eye-contact 'reward' cut. The face is clearly beautiful, and foreground flowers or light do not obscure the face. Cut 5: The butterfly soars high, and the camera rises with it through the wisteria trellis. The wisteria clusters, flower crown, white dress, and light-colored hair recede below, finally showing the entire garden from above. The girl remains small in the center of the garden, with wisteria and cherry blossom paths forming a circle around her. The butterfly shines small at the top of the screen, and the entire garden spreads out brightly as if blessed. [Prohibited] Thick black outlines, low-density backgrounds, simple zooms, TV anime-style flattening, 3D, live-action, semi-real, CG doll skin, facial distortion, collapse of limbs, turning into another person via character rotation, giant butterfly close-ups, compositions where the butterfly covers the face, water drops/dew/tear-shaped glass/transparent spheres in front of the face, expressions reflecting the face inside drops, transparent objects in positions that look like runny noses or drool, sinking into dark blue-purple, dullness of colors, excessive bloom, extra people.

The exact same character from the reference image @画像1 — 100% identical face, hairstyle, outfit, body proportions, no changes whatsoever, clean high-quality skin with subtle glow. She performs a fierce ballroom Vogue Fem dance break in an ultra-stylish environment that perfectly matches her specific aesthetic and original vibe. She executes rapid hand performance, complex geometric arm framing, striking precise poses, and sudden dramatic death drops, projecting explosive diva energy. Mind-blowing music video multi-cut camera work: starts with a rapid 360-degree orbit tracking shot around her, transitions with a fast whip-pan to an extreme low Dutch-angle looking up at her poses, followed by dynamic push-and-pull dolly movements tightly synced to her intricate hand frames. High-energy rapid editing. Cinematic color grading: dramatic high-contrast lighting designed to make her exact outfit pop, sleek environmental reflections, cinematic anamorphic lens flares, subtle film grain, high-end music video style. 4K, photorealistic, ultra-sharp skin and fabric details, intense rhythmic speed ramps synced to an imaginary heavy bass beat, hyper-fast whip transitions mixed with buttery slow-motion on her sharpest voguing poses, "ultra-dynamic high-end fashion music video masterpiece" optional, 24fps

Ultra cinematic chaotic hospital music video inspired by emo punk aesthetics. A beautiful curvy young woman with long black hair and soft curtain bangs sits inside a strange futuristic hospital room. She wears an oversized white patient shirt, messy stylish hair, glossy smudged makeup, pearl earrings, and small bandaids on her face and knees. The atmosphere feels emotionally unstable, surreal, and darkly comedic. Doctors and nurses move strangely in the background while fluorescent hospital lights flicker dramatically. She looks emotionally exhausted, numb, and mentally overwhelmed while trying to stay calm. Scenes: - sitting alone on hospital bed staring blankly - walking through long hospital corridor in slow motion - chaotic nurses rushing past camera - fisheye lens close-up with distorted perspective - lying on floor laughing and crying - pills scattered on table - dramatic flashing lights - emotional eye close-up - CCTV style security camera shots - dancing alone in hospital hallway Visual style: hyper realistic cinematic music video, emo punk energy, chaotic editing, handheld camera, motion blur, film grain, glossy skin texture, neon hospital lights, surreal atmosphere, emotional storytelling. Color grading: cold white lighting, green hospital tint, subtle neon blue and red accents. Aspect ratio 16:9.

High-energy Japanese anime sports style video, 8k resolution, cinematic lighting. A tall blonde Norwegian soccer striker in a red Norway jersey standing on a pitch in a stadium filled with roaring fans waving red, white, and blue Norwegian flags. He vigorously beats a large red war drum with mallets in sync with rowing team members and cheering crowds. Magical golden runic energy emanates from the drum in glowing rings. Suddenly, the scene transitions into a mythical Viking stadium under dark stormy clouds. The players transform into fierce Viking warriors clad in fur armor and holding carved wooden shields. Above the drum, a glowing, ornate golden soccer ball emerges, floating with intense light, surrounded by a roaring, epic crowd. Dramatic dynamic camera cuts, hyper-detailed, vibrant red atmosphere. Step-by-Step Scene Prompts 1. Modern Stadium & Viking Clap Chant Prompt: Dynamic anime scene, a blonde Norwegian football star standing on a stadium pitch, playing a massive red war drum. Sitting on the grass behind him, his teammates in red Norway jerseys perform a synchronized Viking rowing chant. Stadium crowds wave Norwegian flags under smoke flares, highly detailed 2D anime style, dramatic camera angles. 2. Magical Rune Transformation Prompt: Close-up anime shot, glowing golden Nordic runes and magical light rings bursting from a large red drum as mallets strike it. Dramatic smoke and flare lighting in a packed stadium, glowing particle effects, intense action scene, dark contrast.

Style: Moody, high contrast K pop, single color saturated set (deep red/crimson with black), theatrical spotlighting, slower beat drop builds Subject: Female idol, structured red and black outfit with sheer/metallic panel details [0:00-0:03] Close up, performer's hand emerges from darkness into a single spotlight, slow reveal of face on the beat, camera pulls back to a full silhouette against a red cyclorama. [0:03-0:07] Wide shot, sharp choreography with strong shadow-play against the red backdrop, camera does a slow lateral track, dramatic spotlight chases her movement. [0:07-0:11] Medium shot, performer delivers hook line seated on a minimalist throne like set piece, quick cut to a dramatic red fog stage backdrop. [0:11-0:15] Triple cut power poses each lit from a different angle (side, top, back), freeze frame hero pose with red haze burst, punch-in, fade to logo.

Use 2 reference images. Treat image1 as Person A and image2 as Person B. Represent Person A and Person B as characters appearing in a live-action music video (MV) while maintaining the facial features, hairstyle, body type, expression, costume impression, atmosphere, and character identity from the reference images. Do not mix the two people. Do not swap faces, hairstyles, costumes, physiques, vibes, VFX, or character traits. Maintain the individual personality of each reference image and depict them clearly as separate individuals. Integrate characteristic decorations or silhouettes from the reference images naturally as fantasy costumes for a live-action MV. Show realistic textures, hair swaying, fabric movement, accessory reflections, skin texture, and light and shadow. The overall style should be a cute, elegant, and transparent live-action fantasy MV. The background is not a real-world location, but a 'Pastel Ink Dream Stage' for a live-action MV. Arrange a soft white studio space with pale ink, watercolor light, translucent glass, swaying silk fabric, transparent rings, a faintly reflective floor, floating paper fragments, soap-bubble-like circles, and stardust particles. Focus on soft pastel colors such as milky white, baby blue, lavender, pale pink, cream yellow, mint green, and thin cyan. Colors should not emit strong light but appear to bleed faintly. The background is not static from the start; synchronized with 118 BPM, pale ink clouds, watercolor light, transparent rings, glass reflections, fabric, and light particles slowly expand, flow, sway, and melt. Add both cute ink-style VFX and delicate hand-drawn effects. Ink is presented as transparent watercolor, pale brush marks, milky splashes, small ink grains, and blurs spreading in the air. Hand-drawn effects appear momentarily as thin white lines, pale pastel outlines, small stars, thin lightning, hand-drawn speed lines, wavy lines, and sparkling trails. Shake the lines slightly to retain a hand-drawn warmth. Person A's VFX should be sharp and transparent, with thin lines, transparent rings, liquid glass ripples, and sharp hand-drawn trails reacting to the shoulders, wrists, feet, and turns. Person B's VFX should be soft and sweet, with pale ink blurs, stardust, small light flowers, soft ribbon-like brushes, and hand-drawn sparkles reacting to fingertips, hair, costumes, and turns. At the moment they approach, Person A's ring VFX and Person B's stardust VFX momentarily overlap as pale ripples of light before separating immediately. VFX must not hide faces, eyes, body lines, limbs, costume details, or dancing. Costumes are glamorously redesigned for a live-action MV while keeping the reference image worldview, adding glossy materials, layers, short jackets, swaying fabric, chains, belts, ribbons, accessories, boots, gloves, metallic parts, transparency, and accent colors. Person A and Person B show a sense of being a couple, but their costume and VFX personalities are clearly distinguished. Hair, fabric, ribbons, accessories, and characteristic decorations sway naturally with the dance. Expressions are cute and poised, showing a couple-like distance. When one looks at the camera, the other looks at the partner or slightly away; they do not make the same expression simultaneously. The overall impression is a live-action characterization of image1 and image2 in a pale pastel ink world couple dance MV. Person A is cool and sharp; Person B is soft and glamorous. The background is cute but not cheap-looking, moving beautifully with transparent watercolor ink and hand-drawn VFX. Prioritize faces, hairstyles, body types, expressions, and silhouettes. No text overlays, subtitles, logos, or readable characters. No graffiti, spray art, strong neon, cyber-style, or normal landscape backgrounds. No duplication of people, no new people, no perfect synchronization, no symmetrical copying, no standing still, and no just walking. Use one or more reference images. Treat all persons/characters in the reference images as protagonists. If there is 1 image, feature 1 person; if 2 images, 2 people; if 3 images, 3 people, making them all appear simultaneously in the same video. 15 seconds, 60fps. A high-quality dance video in the style of an idol concept MV. Ensure a tempo suitable for a short MV, cute choreography, and synchronized hand-drawn 2D effects. Describe smooth movements for hair, clothes, accessories, hands, steps, turns, gaze, and effects. Place everyone in the same background, shooting space, and MV scene at once. Do not change the background for each reference image. Do not split into different scenes per person. Do not swap characters mid-way. Do not introduce them one by one. For multiple reference images, treat each as an independent person. Do not mix faces, hairstyles, hair colors, builds, costumes, accessories, genders, age ranges, or vibes. Do not make them a pair with the same face, hair, or outfit. If using male and female references, maintain them clearly as male and female. Prioritize retaining the characterization from the reference images, including texture, art style, atmosphere, and personality. Even if textures differ between references, keep them individual for each character. Do not force a uniform art style on everyone. If full bodies or costumes are not visible, naturally complement them based on colors, materials, silhouettes, and character traits from each image. Complementary parts must follow the same direction as their respective reference image. Do not borrow features from other characters. Fix the background as a single common MV shooting space: either an indoor studio, a room by a window, or a simple shooting space. Use soft natural light and an organized background where everyone looks like a lead. Do not change the location per shot; only change camera angles within the same space. Analyze commonalities across the reference images to create a single idol concept MV space, adjusting for atmosphere (bright, pale, calm, humid, cool, cute, etc.). The music is a Japanese song; do not use real existing songs. The video concept is a short concept MV where all characters dance together cutely and impressively in the same space. Priority: everyone exists simultaneously, everyone looks cute, the MV is memorable, and dance integrates with hand-drawn 2D effects. Use side-by-side, triangular, diagonal, or front-to-back arrangements so everyone is visible. Everyone dances at once; no solo-only parts, standing still, or separate-location direction. Do not use template 'cuteness'; analyze the references to create a cohesive idol concept cuteness. Bright characters: bouncy cuteness. Calm characters: elegant cuteness. Cool characters: shy or poised cuteness. Ephemeral characters: soft cuteness. Street-style characters: rough and playful cuteness. Choreography: solo dance if 1 person, group dance if 2+. Include synchronized moments and moments showing individuality. Use hands, shoulders, neck, waist, feet, gaze, turns, and poses naturally. Movements should be cute, easy to mimic, and memorable. No overly childish, overly sexy, excessively intense, unnatural jumping, or random movements. Hand-drawn 2D effects are auto-generated based on character analysis. Ensure shapes, lines, textures, colors, density, and timing suit everyone. Unify the screen with one MV design while giving characters slight variations in color/line style as needed. Effects are hand-drawn motion graphics composited over the video, resembling rough, cute 2D animations like crayon, pencil, oil pastel, or pale ink. Sync them with hands, faces, feet, body movements, turns, gaze, and poses. Balance them across the screen without favoring one person. Do not hide faces or bodies too much. Configure as if shot with 5 cameras, switching every 2 seconds: Front, Top, Left, Right, Back. Maintain the same background space, only changing angles. Prioritize compositions where everyone is visible, including full-body or knee-up group shots. No close-ups that cut people out. Do not fix the camera completely; shake it lightly to the beat (small bounces, punch-ins, slight sways). Do not use messy handheld shaking. 0.0-2.0s: Front camera. Everyone appears in the same background. Full-body or knee-up. Arrange side-by-side or in formation. Everyone rhythmically performs a cute intro pose and small steps fitting their personality. Hand-drawn 2D effects matching the atmosphere appear near hands and faces. 2.0-4.0s: Top camera (overhead). Everyone enters the frame and acknowledges the camera. Coordinate facial gestures and hand accents while maintaining individual character traits. 4.0-6.0s: Left camera. Short choreography using body orientation, waist, shoulders, and hands. Keep knee-up to full-body distance. Hand-drawn effects flow across everyone following movement trails. 6.0-8.0s: Right camera. Everyone faces the opposite way while acknowledging the camera. Cute beat-synced choreography with shoulders, hands, and feet. Show clear steps with a natural composition. Effects derived from reference colors/shapes appear at shoulders and feet. 8.0-10.0s: Back camera. Hair and costumes sway softly. Natural silhouette composition. Everyone looks back over their shoulder at the camera, synchronized or slightly staggered. Includes turns and hand poses with effect trails. 10.0-12.0s: Front camera (Climax). Knee-up to full-body. Memorable group choreography using full body. Increase effect density. Spread shapes/lines across the screen. Add short camera shakes on 'hits'. 12.0-14.0s: Top camera. Everyone looks up with cute facial/hand gestures. Hand-drawn effects spread softly from hands. Maintain all facial/hair consistency. 14.0-15.0s: Ending scene. Front-facing or slight diagonal. Final group pose flowing from the dance. Not necessarily a face close-up; choose the most attractive group composition (full-body to bust-up). No sudden zooms, jump cuts, or abrupt face close-ups at the end. Use a smooth camera approach or slight stop. Emphasize the pose through body stillness, hand shapes, gaze, and lingering movement of hair/fabric and a short burst of 2D effects. Constraints: No characters other than the protagonists. Do not merge multiple references into one person. Do not mix male and female faces. Do not swap hairstyles/costumes. No twin-ification. Do not share character traits just because they share a background. No costumes contradicting references. No poses/expressions/icons/colors inconsistent with characters. Do not use the same template for every effect/gesture. Ensure dance and effects are perfectly synced. Effects should be part of the choreography, not too weak. No explosions, fire, smoke, excessive glitch, or flashy magic. No text, subtitles, logos, or watermarks. Render the entire 15 seconds smoothly at 60fps.

Use the person from the reference image (attached blonde Korean idol character sheet) as the protagonist. 24fps. ★Live-action — Photorealistic video of an actual Korean woman. Animation or illustration is strictly prohibited. High-quality music video level, lively filming. A short dance video with high cuteness, suitable for going viral. The protagonist is the person from the reference image itself — maintain the long blonde hair, eye shape, facial features, body type, age, clothing impression (Hongdae hip street style), and overall atmosphere exactly. No transformation into other people or characters. The background is a bright, clean, and cutely arranged studio or classroom-style space (trendy Seoul vibe). The background should be simple yet pop and refreshing so the character stands out as the main focus. [Music] A bright, bouncy K-pop style song with English lyrics. Do not use existing songs. [Video Concept] The protagonist performs a cute dance. Prioritize cuteness. Slightly aegyo and friendly, with a strong feeling of dancing towards the camera. 2D flat animations, as if made with After Effects (AE), appear in sync with the dance moves. The effects are cute, hand-drawn style illustrations like crayons or oil pastels — circles, stars, hearts, flowers, notes, arrows, ribbons, crowns, sparkles, wavy lines, swirls, speech bubble borders, checkers, diagonal lines, halftone dots, etc. All are flat 2D. The color palette centers on pink, sky blue, yellow, white, lavender, and mint green, fitting for the character. [Camera Configuration] A setup that feels like it's filmed with 5 cameras. Not fixed, but switching every 2 seconds. Cameras used: Front, Top, Left, Right, Back. Even when the camera changes, the protagonist always stays aware of the lens and maintains eye contact. Even from non-frontal cameras, they must look at the lens by adjusting head direction, neck tilt, or gaze. When filming from the back, they look back over the shoulder at the camera. Dynamic camerawork with focal length changes and good tempo. Each camera is not completely static but shakes lightly to the beat — small vertical bounces, short punch-ins, slight left-right sways, etc., matching the music for a cute camera shake. Not messy shaking, but intentional and aesthetically pleasing. [Dance Content] Cute and easy-to-follow choreography. Light steps side to side, two-handed heart in front of the chest, one-handed heart, V-sign next to the cheek, gently opening both hands next to the cheeks, finger pointing, small shoulder pops, light waist sways, stepping one foot out and back, a small turn, and a wink at the end. Plenty of upper body highlights, short but memorable choreography. Always maintain eye contact while dancing. [15-Second Timeline] 0.0~2.0s: Front camera. Standing in the center, riding the rhythm with a two-handed heart in front of the chest, small side steps. Light vertical camera bounce. Crayon hearts, stars, circles, and short sparkle lines pop up around hands and face. Always facing front. 2.0~4.0s: Top camera. A cute high-angle shot. Looking up at the camera with a V-sign, one-handed heart, and finger pointing with good tempo. Crowns, flowers, hearts, and speech bubble borders appear around the face. 4.0~6.0s: Left camera. Body slightly to the left, face and gaze towards the lens. Light waist sways, one hand on the cheek, waving slightly with the other. Crayon ribbon lines, notes, wavy lines, and stars appear following hand movements. Small left-right camera shake to the beat. 6.0~8.0s: Right camera. Turning the body to the other side while maintaining eye contact. Rhythmic shoulder pops, raising and lowering both hands, and a light step out and back. Flowers, circles, checkers, halftone dots, and sparkles pop around feet and shoulders. Short punch-in. 8.0~10.0s: Back camera. Starting from the back but turning quickly over the shoulder to clearly eye the camera. Finger pointing, hearts, and a light blowing kiss gesture while turning. Swirls, stars, ribbons, and hearts appear softly matching hair and clothing movement. 10.0~12.0s: Return to front (climax). Opening hands wide then bringing them to the chest, with small bouncy side jumps. Large hearts, stars, flowers, crowns, sparkles, and diagonal lines appear on the sides and behind to maximize cuteness. Short camera shakes at key moments. 12.0~14.0s: Re-switch to top camera. Looking up, opening hands gently in front of the face, one-handed heart, the other hand next to the cheek. Crayon borders, hearts, flowers, checkers, and rough circular lines spread in the background. Always eyeing the camera. 14.0~15.0s: Front hero shot. Bust-up, leaning slightly forward. Smiling brightly and winking at the end. At the moment of the wink, crayon square stars, hearts, radial lines, and ribbon lines pop near the eyes. Very short camera punch-in then freeze, ending on a viral-worthy key cut. [AE-style 2D Effect Direction] All effects are 2D flat animations as if made in AE. However, the texture is a rough, cute hand-drawn style like crayons or oil pastels. Motions include appearance, bounce, scale, rotation, parallax, and smooth fade-out. Effects appear synchronized with hand, face, feet, and body movements. Do not excessively cover the character's face or body; use them as aids to enhance cuteness and tempo. [Constraints] No people other than the protagonist. Maintain the atmosphere, face, hair (long blonde), eyes, body type, age, and clothing of the reference character. No transformation into other characters. No arbitrary clothing changes. Clear 5-camera configuration switching every 2 seconds. Front, top, left, right, and back must be used. The protagonist always looks at the camera regardless of the angle. From the back, they must look back over the shoulder. Effects must be cute 2D crayon drawings. Prohibit realistic explosions, fire, smoke, or excessive glitches. Do not cover the face excessively with effects. ★Maintain a real live-action person — animation or illustration is prohibited. Do not display text, subtitles, logos, or watermarks.

Shaky Snorricam footage with several dynamic cuts, heavy body-mounted camera vibration and natural shake, raw chaotic real footage. Underground rave under concrete overpass at night, thick smoke, pink and purple lasers. The protagonist starts doing the "pajaritos" dance and it spreads to the people around him. Use the provided image as loose inspiration for the main chaotic dancer. Keep visual consistency on the central guy but let the model drive natural appearance and energy. 0-3s: [Shaky Snorricam Start] He is in the middle of the crowd doing the pajaritos dance hard, flapping his arms like wings energetically and rhythmically, full commitment, camera shaking violently with his movements. 3-6s: [Dynamic Cut] A few people around him notice and start joining, copying the wing-flapping gesture. Small group begins doing the pajaritos dance with him, camera still very shaky on the protagonist as more arms start flapping around him. 6-9s: [Dynamic Cut] More people join in, now a bigger group is doing the silly chicken dance around him. Everyone flapping arms like wings in chaotic unison, sweaty bodies moving together, snorricam shake remains intense on the main guy. 9-12s: [Dynamic Cut] The chicken dance spreads even more, bigger circle forming with him in the center. Lots of people wildly doing the pajaritos move, arms up flapping, laughing and going crazy together, camera shaking hard with the group energy. 12-15s: [Final Shaky Cut] Peak moment, decent crowd around him all doing the pajaritos dance like pollitos, arms flapping everywhere in chaotic sync. He stays in the middle leading the silly energy, maximum shaky snorricam, pure fun chaotic rave moment. heavy film grain, detailed sweat and skin, realistic fabric and arm movement, perfect chaotic motion blur, coherent group physics and interactions, stable central character, cinematic neon lighting, raw and fun, no artifacts.

The character from the attached reference image dances playfully, rendered in exactly the same art style as the reference — whether photorealistic, anime, or 3D, match it faithfully. Keep the character's face, hairstyle, outfit and overall design perfectly consistent with the reference throughout the entire video. Background: a colorful pop-art style room — vivid candy-pink and sky-blue walls, oversized polka dots and star shapes, confetti accents, bright cheerful studio lighting. The character tilts their head side to side twice with fists near their cheeks, then crosses their arms and flicks them open with fluttering fingers while bouncing their shoulders, side-steps left and right poking their cheek with one finger, freezes completely for half a second as the music breaks to silence, then does a quick two-step making a small heart with their hands, spins half-turn and looks back over their shoulder with a wink and finger heart. Rhythmic, bouncy, K-pop idol energy but innocent and charming. Fixed camera, medium shot from waist up, slight low angle, clean composition, subtle film grain, 4K quality, 10 seconds. Audio: cute kawaii Japanese pop jingle at 120 BPM, ukulele and glockenspiel melody, handclaps on the beat, playful female scat vocals like "pa pa pa", bright and bouncy TV commercial style, the music cuts to one beat of complete silence when the character freezes, then kicks back in with energy for the finale.

A Japanese full-color anime with no captions, no background music, rapid-fire editing, a high frame count, and 24 FPS. Use @image1 as the main reference for the overall pastel beach-pop mood, shark-hoodie styling, cute character proportions, and playful summer energy. Use @image2 as the reference for calm-cute facial rendering, pastel eye design, and soft fashion balance. Use @image3 as the reference for the brighter pastel shark motif, simplified cute silhouette, and pop color blocking. Use @image4 as the reference for expressive face variety, comic-style split-panel layouts, and energetic reaction diversity. Use @image5 as the reference for exaggerated cute-chaotic expressions, candy-like graphic decoration, and playful visual intensity. Reference @audio1 only for beat timing and rhythm if an audio reference is later provided. Do not generate music. Goal: Create a 15-second 720p anime MV-style sequence showing a small group of cute pastel mob dancers joyfully dancing together. The scene should feel playful, bright, social, and energetic, like a fun crowd dance or kawaii flash-mob moment. Do not lock the performance into one exact choreography. Let Seedance infer the dance naturally from the references: group swaying, bouncing, step-touch rhythms, simple synchronized moves, playful arm gestures, turns, reactions, spacing changes, and cheerful group interaction. Focus on lively group motion, camera play, split-screen rhythm, and a cute party atmosphere. Use camera effects and panel composition actively, but keep the characters readable and fun. 0-3s: Open with a lively group reveal. Show multiple cute mob dancers already in motion, using a wide shot, medium group shot, or fisheye-led opening chosen naturally by the generation. Let the group feel immediately active and upbeat. Use playful camera motion, light fisheye distortion, and buoyant rhythm. The dancers should not hold one fixed pose; they should already be moving together in a casual but coordinated way. Floating bubbles, stars, candy-like shapes, clouds, pastel particles, and beachy pop motifs may move through the frame. 3-6s: Move into a more dance-focused section. Let Seedance infer a variety of cute mob-dance actions: small synchronized steps, side-to-side movement, hand waves, shoulder bounces, light turns, call-and-response gestures, and little formation changes. Use alternating camera sizes: medium group shots, brief close-ups, and occasional fisheye push-ins. Show the group’s fun and shared rhythm rather than one leader doing all the work. The dancers should feel like a cheerful crowd moving together. 6-10s: Introduce split-panel and collage rhythm. Break the frame into multiple angled panels that show different dancers, different moments, or different fragments of the same group dance. Some panels may show close-up expressions, some body movement, some group spacing, some hands or accessories. Do not repeat the same pose across all panels. Let the panels behave like a playful remix of the dance. Use sliding panel borders, snapping cuts, slight rotation, and rhythmic reassembly. Allow occasional prism-like edge duplication, cute glitch fragments, or macro inserts of accessories and eyes, but keep the dance energy central. 10-13s: Escalate the group energy. Let the mob dance become more animated and varied without becoming chaotic noise. Seedance may infer little jumps, spins, bounce accents, quick turns, mirrored motions, or playful interaction between neighboring dancers. Use camera effects more actively here: fisheye close-ups, brief macro flashes, snap zooms, whip-like transitions, and layered panel bursts. The group should feel increasingly joyful and lively, as if the dance is peaking. 13-15s: Do not force a standard final hero pose. Let the ending emerge naturally from the dance. Possible endings may include: the group clustering together while still moving, a playful freeze during motion, a split-panel collapse into one group image, a joyful reaction burst, a cute jump or bounce accent, or a clean energetic cut while the dance is still alive. The final beat should feel fun, spontaneous, and satisfying, without looking mechanically predetermined. Keep: - Keep the overall design language consistent with the references: pastel pink, cyan, mint, yellow, lavender, sky blue, and candy-pop accents. - Keep the characters chibi-cute, stylized, expressive, and visually unified, while allowing some variety between mob dancers. - Keep the shark-hoodie / beach-pop / candy-kawaii styling language from the references. - Keep the mood fun, light, social, and energetic. - Keep the animation group-oriented, with multiple dancers visible across the sequence. - Keep the visual emphasis on cheerful dancing, panel rhythm, cute expressions, and playful camera effects. - Keep the rendering flat, graphic, clean, colorful, and slightly sketchy with lively line energy. - Keep the sequence readable even when the edits become fast. Avoid: - No generated text, no subtitles, no logos, no readable signs. - No dark or horror tone. - No gore, no violent imagery. - No realistic live-action look. - No lonely solo performance for the whole clip; it should clearly feel like a mob dance or group dance. - No rigid repeated loop of the exact same move. - No overcomplicated choreography that looks stiff or mechanical. - No excessive effects that hide the dancers for long periods. - No muddy colors or desaturated lighting. - No off-model redesigns that break the reference style.

A digital anime-style illustration of Android Deltamon features as a DJ-loop, against a glowing circular structure in a very starry night sky fantazi, no zoom, 12 FPS No dialogue

Main Character: A young East Asian woman with light brownish-gray long straight hair and air bangs, fresh no-makeup makeup look, wearing a loose white long-sleeved shirt with a slightly open collar. Her skin is delicate and translucent, with a cool, relaxed, and casual tension. Maintain consistent identity, clothing, hairstyle, and appearance throughout the video. Location: By an indoor window with ample afternoon sunlight, white blinds, and off-white walls. Sunlight filters through the blinds, casting parallel stripes of light and shadow. The space is clean, minimalist, with a lazy and quiet home atmosphere, free of clutter. Visual Style: Fresh Japanese realistic texture. Rap movements are relaxed and on-beat without appearing forced. Body movements and facial expressions match the rhythm. Strong play of light and shadow, translucent glowing skin, and soft, clean overall tones. Camera Style: Lightweight handheld camera texture with subtle natural breathing jitter. Small-scale tracking shots following the rap rhythm. Real light and dark layers under natural light, soft focus, preserving original light and shadow texture without over-grading or excessive skin smoothing. 00:00–00:02 Standing against the wall by the blinds, she lightly pinches her shirt collar, head tilted back slightly, looking lazily at the camera, nodding to the rap intro beat. Striped light shadows move slowly across her face, creating a soft and quiet scene. 00:02–00:04 Medium shot at eye level, her body sways gently to the beat, one hand naturally raised, fingers tapping the air. Her mouth opens slightly to follow the rhythm, shirt hem swaying with movement, light flowing slowly over her shoulders. 00:04–00:06 Close-up shot pushing in, she tilts her head slightly to the beat, making natural rap hand gestures. Her eyes have a relaxed 'cool' look, lips sync with lyrics. Light and shadow from the blinds alternate on her neck, full of atmosphere. 00:06–00:09 Slightly wider shot, her body pops gently on heavy beats, one hand resting at her side, the other waving with the rhythm. Movements are sharp but not stiff. Sunlight prints clear stripes on her white shirt. 00:09–00:11 Profile close-up, she turns toward the blinds and sings softly. The jawline is sharp under the light, hair brushing against her cheek. The camera has a natural slight breathing jitter. 00:11–00:13 Switching back to a front close-up, she tilts an eyebrow at the camera, fingers sliding down the shirt collar, precisely hitting a rap pause. Her gaze is lazy yet powerful, light and shadow intersecting at the collarbone. 00:13–00:15 Ending still, she returns to leaning against the wall with a faint smile. As the final heavy beat drops, she looks up at the camera. Blind shadows rest quietly on her body as the screen fades out slowly. Audio: Natural sound field rap track—clear close-up rap vocals, crisp and fast drum beats, paired with faint ambient white noise and slight wind sounds outside. No heavy reverb, overall relaxed yet rhythmic. Goal: Create a lazy and relaxed window-side impromptu rap clip, using blind shadows to set the mood. Movements are natural and perfectly timed, clean, healing, yet possessing a cool tension, like a casual daily recording.

Style: High budget Kpop music video, glossy commercial lighting, vibrant color grading (magenta/cyan/electric blue), 4K cinematic quality, sharp focus, dynamic camera movement, idol-group choreography, fashion forward streetwear meets futuristic outfits, confident charismatic energy, beat synced cuts\nSubject: One female idol performer, toned dance-ready physique, glossy makeup, holographic/metallic outfit accents\nNegative prompt: no blurry motion, no distorted hands, no extra limbs, no text artifacts, no watermark, no low resolution, no static camera\n\n[0:00-0:03] Close up shot, performer's eyes snap open on the beat drop, quick zoom out reveal to full body pose under strobing neon backlight, hair flip, confident smirk, camera whip pans left.\n\n[0:03-0:07] Wide shot, sharp synchronized choreography (point formation), sharp isolations on snare hits, camera does a low-angle orbit around the performer, neon light trails following hand movements, fog machine haze in background.\n\n[0:07-0:11] Medium shot, performer lip-syncs main hook line with emotive expression, hand near face/mic gesture, quick cut to a rooftop-city backdrop with lens flare, wind machine blowing hair.\n\n[0:11-0:15] Fast triple cut montage (0.5s each) of three different power poses, final freeze frame on a hero pose with sparkler/confetti burst, camera punches in, fade to logo card.