30-second 16:9 cinematic. @image1 is the fighter — photocopy their face. Outfit: dark blue-black battle armour, spear that later unfolds into a long energy rifle. Do not invent a different face. The enemy is an ICE DRAGON: colossal wingless oriental serpentine body, pale-blue ice-crystal scales, frost mane, ice horns, no fire, no lava. Frozen black-stone ruin over an ice abyss. Storm of ice crystals, not embers. 0–6s: Low TRACK under a giant ice claw slamming down. The fighter slides through, cyan spear flashing, stabs the claw. 6–12s: They climb the frozen arm and spine. Ice shards fly. FPV past the jaws. Frost breath blast — they vault a stone pillar. 12–18s: Along the ice-ridge of the back, a heavy spear cut. Ice scales crack, pale-blue frost-blood mist. The dragon recoils and a close frost-breath knocks them away. 18–24s: They slam onto a cliff altar, spear-tail braking. The same spear unfolds into a long cyan energy rifle. The ice dragon coils in the storm, ice-core glowing in the chest. 24–30s: Kneeling, they track the core and fire a white-cyan beam that punches the ice-core. Hold. No slow-mo. No on-screen text. AUDIO: ice crack, roar, spear hits, one rifle blast.
Drag the clip sideways
ASSET MAP: @image_1 is the first frame of this film and the only reference. The person in @image_1 is the cook and the beach shack in @image_1 is the only location. Identity: the cook in every shot is exactly the person in @image_1 - face, skin tone, hairline, hair length and colour, build and gender presentation unchanged, never masculinized or feminized, never restyled, and long hair is never added to a short haircut. Only one cook exists in this film. The t-shirt, the denim apron, the counter, the produce, the sea and the boats come from @image_1 and stay unchanged. Style: warm documentary food film at golden hour - low sun, soft haze, shallow depth of field, real handheld weight, crisp texture on skin, wood, char and crumb. Hard cuts, no dissolves. 0-2 seconds: the film starts exactly on @image_1 and moves at once - the knife finishes the cut and a tomato slice falls onto the fan of slices, the cook glances up into the lens for a beat, then back to the board as the camera pushes in. 2-5 seconds: low across the bars of a hot griddle - a thick seasoned white fish fillet is laid down, sizzles hard, and steel tongs turn it to show the golden crust, smoke drifting out toward the sea. 5-8 seconds: high shot down onto the board as the hands of the cook build the burger - the hot fillet onto the lower bun over lettuce, two tomato slices, three red onion rings, a spoon of pale sauce, the top bun pressed down. 8-11 seconds: the finished burger sits on grease-proof paper in the foreground and the camera pulls focus up to the cook, who straightens up, wipes both hands on the apron and looks into the lens. 11-13 seconds: the camera swings to the front of the counter as the cook passes the wrapped burger across the timber to a waiting guest, the sea bright behind and a few guests at driftwood tables on the sand. 13-15 seconds: the camera cranes backwards and up over the roof of the shack in one move, opening onto the whole beach - the low sun on the water, a fishing boat heading in, the shack small and warm below. Continuity: the fillet that is grilled is the fillet that goes into the burger, and that burger is the one handed over. Five fingers on every hand, the knife held with the fingertips curled back. Every surface in frame stays free of lettering, signage and menu boards. Audio: knife on the board, a hard sizzle, tongs, paper, surf, gulls, quiet chatter and a soft acoustic guitar line. Nobody speaks and there is no voice-over. No subtitles.
10-second 16:9 ultra-photoreal cinematic architectural commercial. Follow the same mechanical assemble-to-hero beats. @image1 is the completed building and site — photocopy that architecture, terraces, green roofs, hedges, street, cars, and facade. Same materials as the photo. Do not invent a different development. Only gloved hands in matte-black gloves. Bright hard sunlight, clear deep-blue sky. 0.0–0.6s: extreme macro, a finger draws a straight line through sand, revealing a concrete foundation. 0.6–1.3s: a hand places one concrete core; columns and structural grid assemble outward. 1.3–2.0s: two fingers pull upward; floors, slabs and columns rise and lock. 2.0–2.7s: a hand slides the upper volume sideways, forming terraces, setbacks and deep balconies; short orbit. 2.7–3.4s: a hand sweeps the frame; stone, metal, panoramic glass and balcony details install. 3.4–4.1s: fingers click the final panel; macro glide across stone + metal + glass. 4.1–4.8s: a finger draws a path through sand; paving assembles into paths, plaza and driveway. 4.8–5.5s: a hand plants one mature tree; surrounding trees, lawn and shrubs appear to match @image1 landscaping. 5.5–6.2s: fast hand sequence places bench → architectural light → a dark luxury SUV like the one in @image1. 6.2–7.0s: a finger draws an arc; ground lowers into a shallow reflecting pool and fills. 7.0–8.2s: fingers adjust the final tree; camera pushes into leaves; scale shift from miniature to real-world; hands disappear. 8.2–10.0s: camera emerges under the real tree, flies low over paving, rises along the facade, orbits the corner, ends on a wide HERO SHOT matching @image1. Physical mechanical transformations only: sliding, unfolding, assembling, locking, realistic mass and shadows. No magic, glow, holograms, toys, cranes, workers. No extra on-screen captions. AUDIO: quiet construction clicks, wind, no voiceover.
Create a 22-second square 1:1 ultra-photorealistic live-action cooking video, closely following the pacing, composition, camera angles and cooking actions described below. ASSET MAP — every reference is bound here and cited by tag below: @image1 — The product. This is the raw food item; it defines shape, size, texture, color and identity. The attached photo is a pile of raw chicken wings (whole wing pieces with drumette and flat, pale pink skin, natural moisture) on a dark slate. OBJECT LOCK — HIGHEST PRIORITY: The food item shown in every shot is exactly the same physical piece shown in @image1. Preserve its recognizable shape, size, texture, color, proportions and distinctive surface details — identically — as it is seasoned, marinated, fried, lifted, tossed and plated. Never replace, duplicate, regenerate, morph or substitute it with a different piece. The final cooked, plated result must visibly originate from the exact item shown in @image1. ATTACHED IMAGE IS MANDATORY AND FINAL: @image1 is the real object — reproduce it exactly, in every single frame. Never invent a different item, shape or texture. The attached image outranks every word written below it: wherever the image and this text disagree, follow the image. EVERY FRAME MUST LOOK LIKE REAL LIVE-ACTION FOOD FOOTAGE: real ingredients, real hands, real utensils, real oil, real steam, real kitchen, physically accurate cooking. No CGI, 3D render, animation, artificial food, plastic textures or impossible physics. Natural imperfections, realistic moisture, oil bubbles, steam and browning. Hands have five fingers each, anatomically correct, natural grip. Do not add any lettering, logo, brand mark or on-screen text that is not present on the reference image. No subtitles. No watermark. 0-2s — RAW PRODUCT: Extreme macro close-up of the exact product from @image1 on a white ceramic plate. Show its raw texture, color and structure exactly as shown in the reference — natural moisture, real surface detail. Camera slowly pushes across it, clearly establishing its individual shape and identity. 2-4s — SEASONING: Cut to a clear glass mixing bowl containing the same product. Golden beaten egg is already present around it. Seasoning powder is poured from above, visibly landing on the same product. Show realistic powder particles and glossy egg coating. 4-6s — MIXING: Black-gloved hands vigorously mix and massage the same product inside the glass bowl. It becomes evenly coated in a glossy reddish-orange marinade. Close-up of fingers turning the same recognizable pieces. Preserve their original shapes exactly as in @image1. 6-9s — FIRST FRY: Wooden chopsticks place the same marinated product piece by piece into a black frying pan filled with shimmering hot oil. Vigorous bubbles erupt immediately around each piece. Camera stays close to the pan. The product gradually changes from raw to golden brown while retaining its original shape from @image1. 9-11s — LIFT FROM OIL: Use a metal spider strainer to lift the same freshly fried product from the oil. Oil drains naturally through the wire mesh. Show crisp golden-brown surfaces, bubbling oil droplets and realistic steam. No replacement pieces. 11-14s — AROMATICS + TOSS: The same fried product is placed back into the pan. Add fresh glossy green kaffir lime leaves, whole garlic cloves and bright red chilies. Chopsticks and the strainer gently toss everything together. The product becomes deeper golden and crisp while the leaves and garlic fry naturally. 14-17s — PLATING: The same cooked product is lifted directly from the pan with the metal strainer and transferred onto a clean white ceramic plate. Arrange it naturally into a small stack. Fried kaffir lime leaves, sliced red chilies and garlic are placed visibly around and on top of the same product. 17-20s — MACRO FOOD DETAIL: Extreme macro close-up of one specific piece from the original raw set shown in @image1. Show its crisp blistered golden crust, browned edges and cooked texture. The camera slowly pushes in. Tiny oil droplets and delicate steam are visible. It must retain recognizable characteristics from its raw appearance in @image1. 20-22s — FINAL HERO SHOT: Elegant close-up of the completed plate on a dark kitchen countertop. The exact same product is stacked naturally with green kaffir lime leaves, red chilies and fried garlic. Warm natural cinematic lighting, shallow depth of field, realistic steam and subtle camera movement. End on a sharp appetizing macro detail of the crispy crust and cooked texture. CONTINUITY RULE: RAW PRODUCT (@image1) → MARINATED → FRYING → FRIED → PLATED = THE EXACT SAME PHYSICAL ITEM. Never introduce a new piece. Never replace a piece during a cut. Never regenerate the cooked result as a different item. Cooking changes only its color, moisture, texture and crispiness naturally. VISUAL QUALITY: High-end professional food commercial, photorealistic detail, real camera optics, realistic depth of field, natural motion blur, cinematic warm lighting, authentic reflections and shadows, extremely detailed texture, fat, crispy crust, oil and steam. AUDIO: diegetic kitchen only — oil sizzle, powder pour, mixing, strainer lift, toss, soft plate set. No music, no narration, no lyrics, no on-screen text.
PAINTED ACTION SHORT. ONE PERFORMER AGAINST A PATROL. NO DIALOGUE. FIFTEEN SECONDS. ASSET MAP - every reference is bound here and cited by tag below: @avatar_1 - the tracker, and the only named performer in the film. Face, hair, skin, age, build and every physical characteristic come entirely from @avatar_1 and never from this text. IDENTITY LOCK - HIGHEST PRIORITY: the tracker is exactly the person in @avatar_1. Keep their face, hair, skin tone and build the same from the first frame to the last. Do not restyle them, do not age them up or down, do not slim or bulk the build, and never put a second or third performer in frame with a speaking part. One performer, the same one, for the whole film. DRAWN, NOT PASTED: @avatar_1 is drawn into the painted style of the film - brush texture, ink shadow, flat graphic light - and never composited in as a photograph. Their face still has to read as unmistakably theirs: the same features, the same proportions, the same hair and skin tone, translated into paint. ATTACHED REFERENCES ARE MANDATORY AND FINAL: if a reference is attached, that is the performer - reproduce them exactly, in every shot, and never invent a different one. The attached reference outranks every word written below it. FACE TIME: @avatar_1's face is clearly visible and clearly lit in the first shot and again in the last. Never bury it in shadow, never cover it, and never hold the camera behind them for a whole shot. STYLE: hand-painted animated action in the manner of a printed comic brought to motion. Thick visible brush texture on every surface, hard graphic shadows, saturated jungle greens against warm skin tones, shafts of hard sunlight cutting through canopy haze, dust and spray catching the light. Bold clean shapes, strong silhouettes, cel-animation timing with snappy holds on the impacts. Not photoreal at any point. WARDROBE LOCK: @avatar_1 wears a mud-streaked dark green oilskin coat open over a grey undershirt, a long strip of cloth wound around the left forearm from wrist to elbow, heavy laced boots, and a plain machete in a worn leather sheath at the hip. No hat, no headband, no bow, no dog tags, no insignia, no lettering on any garment. These garments are fixed for the whole film and only their drape adapts to the body wearing them. THE PATROL: four armed figures in plain olive field uniforms with soft caps, carrying rifles. Faces kept in shadow or turned away. No badges, no flags, no unit markings, no readable lettering anywhere on them. THE PLACE: deep rainforest. Buttressed roots, hanging vines, a floor of wet leaf litter, a shallow brown river running through the middle of it. Mist hangs between the trunks and the light comes down in hard vertical columns. 00:00-00:03 Close on @avatar_1 turned to camera, canopy light falling across the face, eyes up and steady, drawing the machete in one unhurried motion. The face is large in frame and fully readable. Far behind them down the corridor of trees, one of the patrol is a small shape in the haze. The camera holds. 00:03-00:06 Cut to boots at ground level, running hard through leaf litter, water and leaves kicking up past the lens. The camera tracks alongside at speed. 00:06-00:09 Cut to a medium shot in a clearing: @avatar_1 turns hard into two of the patrol at close quarters and strikes the nearest rifle aside. Fast, heavy, unglamorous movement, and the face catches the light through the turn. The camera whips with the turn. 00:09-00:12 Cut to a wide shot across the shallow river: two of the patrol stand in the water at the far bank with rifles raised, and @avatar_1 swings low and hard, sweeping the legs from under the nearest one so they go down into the brown water in a sheet of spray. The camera pushes in with the movement. 00:12-00:15 Cut to @avatar_1 alone in the shallows, coat heavy with water, the patrol down around them, a small fire burning in the wreck of a camp behind. The camera pushes in slowly until the face fills the frame, lit and clear, and holds there to the last frame. This is the first and only fire in the film - nothing burns before this shot. AUDIO: rainforest under everything - insects, distant birds, water running over stones. Boots in wet leaves, cloth and leather under strain, the dull heavy impacts of the fight, a body hitting water, fire crackling at the end. A spare percussive score built on low drums and one held low string note, kept well under the effects and never rising to a peak. No voices, no shouting, no dialogue, no lyrics. Do not generate subtitles, captions, on-screen text, a title card, a logo, a wordmark, a watermark or any readable lettering or numerals anywhere in frame. Do not generate spoken dialogue or narration.
CHARACTER LOCK — HIGHEST PRIORITY: Use @image1 for the MAIN FIGURE’s facial identity ONLY. Preserve their exact facial structure, eye shape and colour, eyebrows, nose, lips, cheekbones, skin tone, freckles and overall facial identity, identically across every shot. Preserve their gender presentation exactly as shown; do not masculinize, feminize or otherwise reinterpret it. Do NOT copy their reference-image clothing, hairstyle, jewellery, background or pose — their wardrobe and hairstyle are newly designed for this scene, described below. ATTACHED IMAGE IS MANDATORY AND FINAL: wherever @image1 and this text disagree on facial identity, follow the image. No face changes, no identity drift, no other people sharing their face. @image1 is face and identity only. It is not a green satin shirt, not a gym, not a brown handbag, not a long side braid, not beige loafers. Ignore wardrobe and set in the still. DECLARED VARIABLES (edit these to reskin the workflow — the body below refers to them by name and must not restate their values): CROWD_OUTFIT: identical matte charcoal-grey tactical suits HERO_PROP: a sleek pair of black sunglasses VOICE_STYLE: calm, low, effortlessly confident — no forced accent LINE_1: “Took you long enough.” LINE_2: “Let’s dance.” COLOR_GRADE: cool blue-gray cinematic grade, overcast daylight, subtle atmospheric haze SETTING: A large outdoor stone staircase completely filled with rows of people wearing CROWD_OUTFIT, standing in perfectly organised rows. The crowd remains visually consistent throughout — same outfit, same formation. COLOR_GRADE. Realistic depth, dramatic high-end music-video cinematography. THE MAIN FIGURE wears an elegant, sophisticated monochromatic black-and-white fashion outfit: a fitted black tailored blazer, crisp white shirt, elegant high-waisted black bottom, sheer black tights or tailored trousers, and polished black designer-style boots. Hair is a newly designed sleek elegant style for this scene — not the gym braid from @image1. Their styling is luxurious, refined and powerful. They are the ONLY person wearing this fashion-forward outfit, creating a strong visual contrast against the uniformed crowd. Their presence must feel like an elite, intimidating fashion icon — effortless, never rushed or casual. Hands have five fingers each, anatomically correct, natural grip throughout. No subtitles. No on-screen text or logos. No watermark. 16:9. TOTAL DURATION: 15 SECONDS. CUT SCENES ALLOWED. OPENING: already on the stone staircase in the black-and-white fashion look. Never a studio portrait, gym interior, or the attached still as frame 0. 0-3s — SHOT 1: Wide frontal shot, framed from mid-thigh up so face and outfit both read clearly. The MAIN FIGURE stands calmly in the centre foreground on the stone staircase. Behind them, the massive CROWD_OUTFIT crowd stands completely motionless in perfectly organised rows. The camera slowly pushes toward them. They look relaxed, confident and completely unbothered, HERO_PROP held loosely in one hand. 3-5s — SHOT 2: Cut to a sophisticated medium close-up. They slowly and deliberately put on HERO_PROP with effortless precision. Their expression remains composed and slightly amused. They look directly into the camera and say naturally, in VOICE_STYLE: “LINE_1” 5-7s — SHOT 3: Cut to a dramatic over-the-shoulder shot from behind them. They slowly turn toward the enormous staircase filled with the motionless crowd. Their playful expression instantly changes into a fierce, commanding stare. They shout, in VOICE_STYLE: “LINE_2" 7-10s — SHOT 4: Cut immediately to a dynamic wide action shot. The entire crowd suddenly charges down the staircase toward them. The camera rapidly tracks backward while keeping the MAIN FIGURE centred. Several attackers rush toward them simultaneously from different directions. They immediately launch into extremely fast hand-to-hand martial-arts combat — precise blocks, parries, rapid punches, spinning kicks, low sweeps and evasive movements. Their blazer, outfit and hair react naturally to every movement. Aggressive speed ramping, realistic motion blur, dynamic tracking, physically believable impacts. 10-12s — SHOT 5 — TIME BULLET: At the peak of the fight, they launch into a spectacular spinning aerial kick. TIME BULLET. Everything suddenly FREEZES in mid-air. The MAIN FIGURE is suspended dramatically at the centre of the frame. Multiple attackers are frozen around them at different distances and angles, some airborne, some falling backward, some reaching toward them. Their hair and clothing are frozen mid-motion. Dust particles, tiny debris and fabric fibres are suspended perfectly in space. The camera performs a smooth cinematic 360-degree orbit around the frozen action, revealing the enormous staircase and frozen crowd from multiple angles. Ultra-detailed frozen motion, realistic anatomy, cinematic depth of field. 12-15s — SHOT 6: TIME RESUMES suddenly at extreme speed. Their spinning kick completes with a powerful impact. Several attackers are thrown backward and tumble down the staircase. They land perfectly and calmly. The chaos abruptly settles. They casually adjust their blazer cuff, and look directly into the camera. Final expression: completely calm, confident and slightly amused, as if absolutely nothing happened. CINEMATIC STYLE: High-end realistic live-action music video. COLOR_GRADE. Realistic skin texture. Realistic physics. Dynamic handheld camera energy. Controlled motion blur. Aggressive speed ramping during combat. Highly cinematic depth of field. Strong visual contrast between the fashion-forward MAIN FIGURE and the uniform crowd. AUDIO: spoken LINE_1 and LINE_2 in VOICE_STYLE, diegetic footsteps and combat impacts, no lyrics, no extra on-screen captions.
A cinematic fantasy fight. Epic, brutal and relentless — every blow lands with devastating, ground-breaking weight. The hunter wins. VISUAL STYLE — HIGHEST PRIORITY Match the reference images exactly: animated painterly concept art, visible brush strokes, hand-painted textures, saturated but controlled palette, anime and game splash-art aesthetic. This illustrated rendering must hold through every movement, power effect and camera angle — characters and environment share the same painted look, no style mismatch. Absolutely no photorealism, no live-action textures, no glossy 3D-CGI or plastic rendering, no outlines or ink line-art, no cel shading, no flat banded shadows, no vector fills, no smooth airbrushed gradients. REFERENCE ASSIGNMENTS @image1 is an absolute character lock for the hunter: face, body, hair, costume, accessories, materials, colours and the sword are taken from it exactly, with no drift or reinterpretation in any shot. @image2 is an absolute character lock for the oni: face, horns, body, costume, materials, colours and the club are taken from it exactly, with no drift or reinterpretation in any shot. The height relationship between the two characters established in the references holds in every shot. @image3 is the storyboard and is followed exactly — shot order, framing, camera angle, staging, pose and the location's architecture, perspective, lighting and colour palette. Stage the entire battle in that one location, its facades bordering the action, the tall structure and the blossom trees clearly visible in wide compositions. CHOREOGRAPHY Begin in motion with the hunter charging. Low lateral tracking shot as they sprint down the street straight toward the oni, the great sword trailing sparks along the stone, wisps of dark energy already bleeding off the blade. The oni plants a wide stance and swings the club down horizontally to meet the charge. The hunter slides beneath it; the club smashes a facade apart, beams exploding outward. They spring up into an ascending strike that rakes the oni's forearm — first hit, the oni recoils. The hunter leaps onto a facade, two fast wall-running steps, and launches back at the oni. Whip pan catches the airborne slash, releasing a sharp crescent of dark energy that carves across the oni's chest. The oni roars, raises the club with both hands, then hammers it straight down into the street — the ground splits open in a jagged fissure, slabs heaving upward, dust and petals blasting out in a shockwave. Quick close-up in slight slow motion as the hunter reads the impact and kicks off a rising slab to clear the widening crack. The hunter lands and attacks at superhuman speed, darting around the oni's flanks with painted motion trails, dark energy trailing off the sword. Three readable, connected strikes — each one visibly biting into the oni, staggering them back a full step. The oni pivots, catches the third on the club, and backhands the hunter across the street; they smash through a railing, skid, and rise without missing a beat. Tight camera orbit through the exchange. Red flames coil around the oni's club and a dark crimson demonic aura erupts around the oni's body. The oni raises the club overhead and brings it down like a meteor — a massive flaming ground-slam that tears the street apart, a wall of fire and debris racing outward. Slow-motion dodge beat: track tight beside the hunter as they spring off a collapsing facade and vault over the erupting shockwave in slow motion, red flames licking inches away, petals suspended, eyes locked on the oni. Snap back to real time as they thrust their free hand forward and hurl a bolt of dark energy into the oni's chest, staggering them. Cut on impact into a wider view that re-establishes the tall structure behind. The finisher builds. The hunter sprints and leaps high into the air above the oni, rising against the tall structure and the blossoms, the great sword raised overhead in both hands, channelling a massive surge of dark energy along the entire blade — black tendrils spiralling out, the light draining from the street. The oni looks up and raises the club to guard. At the apex they swing down and release a colossal black energy slash. Drop into dramatic slow motion as the black crescent splits the flaming club and cleaves straight through the oni — a black shockwave ring detonating outward, swallowing the red flames, petals frozen mid-air, the ground buckling beneath the blow. Real time returns. The oni is cut clean in two along the energy line; the halves separate and fall away, dissolving into red embers and black smoke before they hit the ground — no gore, a clean dispersal of light and ash. The hunter lands lightly in a crouch amid the wreckage, dark energy fading from the sword, and rises. End on a strong wide composition: the hunter standing alone in the shattered, split street, embers and black smoke dispersing where the oni fell, the tall structure behind them, blossoms drifting down through the settling dust. CAMERA AND CONTINUITY Frantic but readable action. Tracking, motivated orbits and controlled whip pans, clear spatial continuity, cuts driven by movement or impact. Reserve full slow motion for the mid-fight dodge and the final cleaving strike, plus the brief close-up in the second beat. Maintain consistent anatomy, weapon geometry, grip, clothing and height difference throughout. Every attack has anticipation, contact or evasion, and follow-through. Environmental damage persists and accumulates — cracks, fire and debris carry across the fight. Give every beat comparable screen time; no beat is rushed and none is held long. Exactly two characters. No duplicates, additional characters, changing faces, extra limbs, extra weapons, text, subtitles or logos. Stylized clean defeat, no gore.
12 seconds, 1:1 split screen. @image_1 is the person in the LEFT panel, @image_2 is the person in the RIGHT panel. Exactly two people, one per panel. COMPOSITION: the frame is split down the exact middle by a hard vertical edge into two equal panels with a thin black seam. Each panel is its own separate vertical phone video, filmed by a different person, on a different phone, in a different room, at a different time. Nothing ever crosses the seam and the two halves never merge into one room. ATTACHED IMAGES ARE MANDATORY AND FINAL: reproduce each person exactly as in their own attached image, in every frame — face, hairline, hairstyle, hair colour, skin tone, build and apparent age. Preserve each one's gender presentation exactly as shown. Do not beautify, slim or restyle either of them. They are two different people with different faces and different hair: never draw one with the other's face, and never swap the panels. Each voice belongs to its own person. Both wear a plain fitted black cotton t-shirt with a plain crew neck, a thin gold chain and small gold hoop earrings. SETTINGS, one per panel and deliberately different: left is a lived-in bedroom with an unmade bed and a cluttered shelf, daylight from the left; right is a kitchen worktop with mugs and a kettle, daylight from the right. The two rooms must never look like the same place. REALISM — the point of the shot: both panels are raw phone front-camera footage held at arm's length, natural daylight only, handheld drift, slight autofocus breathing, visible sensor grain, a little motion blur. Both faces read as real skin — visible pores, peach fuzz along the jaw, a faint sheen on the nose, soft shadows under the eyes, a few hairs out of place. No beauty filter, no smoothing, no retouching, no studio lighting, no colour grading, no gimbal steadiness. [0-4s] @image_1 in the left panel says: "everyone keeps asking what filter i use." @image_2 listens and half-smiles at their own lens. [4-8s] @image_2 in the right panel leans in, shrugs: "i don't use one either." @image_1 laughs quietly and looks off to the side. [8-12s] @image_1: "this is just my face." @image_2: "same." Both laugh once and both clips end mid-laugh. Only one person speaks at a time. Each panel is one unbroken take. The lines are in English, each in that person's own natural voice, close to the phone mic. The first frame is already lit and in motion in both panels; no fade from black. Do not add subtitles. No on-screen text. No music.
ELARA is the only person. @image1 is her FACE AND OUTFIT — photocopy exactly: silver-grey hair in a high ponytail, brown eyes, pink satin bomber over white crop tube, pale-pink pleated mini with silver buckle, white mid-calf socks, white chunky lace-up boots, silver mini bag, silver choker, hoop earrings, rings. Same nails as the photo. Do not restyle her. Do not recast. ONE GIRL ONLY. No duplicate Elara. No second idol. 15 SECONDS, 9:16, 24fps. Vertical phone fancam. High-energy K-pop girl-group fashion MV. Seoul Y2K street: bright sun, pink storefronts, cafés, photo booths, chrome, blue sky. Distant crowd only as unreadable bokeh — no second face. AUDIO — REQUIRED: generate an original English K-pop song for the full 15s. Bubblegum K-pop × Y2K pop × light hyperpop × dance pop, ~150 BPM, punchy kick, bouncy bass, bright synths, claps, digital bells. Sweet clear idol-style female vocal. Elara SINGS this song the whole time — accurate lip-sync on every shot. English lyrics. No spoken dialogue. No meaningless ah/woo. No on-screen lyric text. 0–1.7s extreme close-up fisheye, eye contact, one hand on collar, other near chin, cool → eyebrow lift → smile, micro push-in, tiny pink sparkle near eye. 1.7–3.3s point one finger at lens, wide-angle fingertip, forward lean, shoulder on the beat, pink fingertip heart trail. 3.3–5s hands near face, fingers trace out, wink → smile, hair moves, brief pink sparkle/chrome heart stickers. 5–6.7s side turn, S-curve, look back while singing, hair flip, hand on waist, fast side orbit, short glitter trail. 6.7–8.3s two cat-paw gestures left then right, nose scrunch → wink, brief pink doodle cat ears, camera dips then returns. 8.3–10s finger heart beside face, other hand points to self then camera, strong lip-sync, lock lens, slight pull-back, one pink heart pulse. 10–11.7s big heart above head, sway once, low-angle fisheye, chrome hearts rise. 11.7–13.3s close to lens, one hand near mouth as if singing a secret, glance aside then back, camera rotates 5–10°, brief pink stars/hearts. 13.3–15s step back, hands open near face, then one finger heart near lens, other hand on chest, final lyric, laugh → wink → head-tilt freeze, burst of hot-pink hearts, chrome stars, glitter; settle. End. Camera: 12–14mm fisheye fancam, close, beat-synced. New gesture, expression, pose, camera each shot. Effects after the action, gone in 0.3–0.8s. No subtitles, no logos, no extra limbs, no frozen mouth while singing.
Ultra-cinematic high-energy fashion transformation, dramatic and punchy: a hard cinematic key light cuts through near-total darkness, the subject full-length and centered in a pool of light against a deep black void, fine film grain, high contrast, snappy dynamic camera. The subject is @image1 — preserve @image1's exact face, hair and identity throughout, real skin texture, natural anatomy; a real person, never a beauty-filter or AI look; @image1 starts in a plain casual base outfit, standing still facing the camera. THE MAGIC HANDS: a single pair of elegant long black leather opera gloves on hands snaps up fast into the foreground from the bottom of the frame, very close to the lens and far larger than the subject, moving with sharp confident energy. 0-1.5s: darkness, then the key light SNAPS on revealing @image1 in the plain base outfit; the long black-gloved hands whip up dramatically into the foreground. 1.5-3s: the gloved hands sweep across in ONE fast snappy whip; exactly on a hard musical BEAT-DROP the clothing transforms in an INSTANT into exactly the outfit shown in @image2 — same garments, colours, styling and details from @image2, worn naturally and fitted to the body, matching the reference exactly, no garment described in words; a quick burst of light and a punchy whoosh sell the instant change. 3-6s: fast confident reveal, the camera pushes in quickly then settles, @image1 wearing the full @image2 look, dramatic rim light tracing the silhouette. 6-8s: @image1 holds a powerful confident final pose as a last light flare sweeps across, the gloved hands already gone from frame. Keep exactly one pair of long-gloved hands in the foreground and @image1's own two hands natural at the sides — never a third hand or extra arm on the subject. The transformation is FAST, snappy and dramatic — one punchy instant change, not a slow morph. ONE continuous coherent take; let the beat play out, no abrupt cut to an unrelated scene. Realistic fabric and leather physics, believable weight and drape, energetic but smooth camera, high-contrast dramatic lighting, premium cinematic colour grade. SOUND: a tense cinematic build with a hard beat-drop hitting exactly on the instant transformation, a punchy whoosh, subtle leather-glove creak; no spoken dialogue, no invented language. Negative: a slow or gradual morph instead of a fast snappy instant change; the subject sweeping their own hands over themselves instead of the separate long-gloved magic hands doing it; more than one pair of gloved hands; a third hand or extra arm on @image1; describing any garment, colour or fabric in words instead of following @image2; an outfit not matching @image2; @image1's identity, face or hair changing; the clothing flickering, morphing between random outfits or clipping through the body; a scene cutting off mid-action or jumping to a random unrelated scene; skipped or truncated beats; @image1 or the outfit vanishing, dissolving, fading or teleporting; on-screen text, captions, numbers or watermark; plastic or airbrushed skin.
Paddle OutRecreate workflow