Открыть промт целиком · 20 980 символов
{
"title": "Street Collision — metropolis golden hour",
"duration_seconds": 12,
"rendering_rule": "For every element backed by a reference image (@Image1 face, @Image2 outfit, @Image3 eyewear, @Image4 cup), the reference image is the sole source of that element's appearance. This prompt describes where those elements are, how they move and how they are lit; their look comes from the images alone. Read their appearance off the images.",
"references": {
"@Image1": {
"role": "face reference",
"use": "identity only",
"reproduce": "Her face is the face in @Image1 — the same person. @Image1 is the complete and only source of her facial identity and hair; copy it exactly as that image shows. Keep natural facial proportions, head sized about 1/7-1/8 of total body height, in natural scale to her shoulders and torso."
},
"@Image2": {
"role": "wardrobe reference",
"reproduce": "Her clothing is the clothing in @Image2, worn as shown there. Copy it exactly as that image shows it, garment for garment. @Image2 is the complete and only source of how she is dressed — read every detail off the image itself."
},
"@Image3": {
"role": "eyewear reference — HERO OBJECT of the video",
"reproduce": "The eyewear in the video is the same object as the eyewear in @Image3. Copy it exactly as that image shows it, in full and in every element. @Image3 is the complete and only source of this eyewear's appearance — read every detail off the image itself, and let the image alone decide what is on it and what it is made of."
},
"@Image4": {
"role": "coffee cup reference",
"reproduce": "The cup in the video is the same object as the cup in @Image4. Copy it exactly as that image shows it, in full and in every element. @Image4 is the complete and only source of this cup's appearance — read every detail off the image itself, and let the image alone decide its form, its surface and what sits on top of it."
},
"location_source": "The street comes from the written 'location' block below."
},
"reference_binding": {
"purpose": "Each reference is bound to one visible element. Confirm all four are correct in every shot where they appear.",
"@Image1 → her face": "Visible and matching from 0:04 onward, fully visible and bare from 0:05.5 to the end.",
"@Image2 → her clothing": "Visible and matching from 0:02 onward, in every shot she appears in.",
"@Image3 → the eyewear on her face": "On her face and matching from frame one through 0:05.5, then matching in close-up from 0:06 to 0:09.2 as it falls.",
"@Image4 → the cup in her left hand": "In her left hand and matching from frame one, clearly visible and matching at the 0:02-0:04 reveal, matching at 0:05.5 as it is knocked loose, and matching in close-up from 0:09.2 to 0:10.1 as it falls and spills."
},
"core_scene": "A young woman walks down a wide downtown avenue toward the camera at golden hour, wearing the eyewear from @Image3 on her face, the cup from @Image4 in her left hand and an old flip phone in her right hand. She is looking down at the phone instead of the road. A man walking the other way clips her shoulder with his as he passes. The impact spins her sideways, she stumbles on her heel, and everything is torn loose at once — the eyewear is knocked off her face, the cup is knocked out of her left hand and the flip phone is jolted out of her right. Time slows and the camera drifts from one falling object to the next in one unbroken move: the eyewear from @Image3 tumbles and rotates in a long close pass past the lens, hands the frame to the spilling cup from @Image4, which hands it to the tumbling flip phone. Time returns to normal as everything hits the asphalt and the coffee splashes across the road. She steadies herself and stands over the mess, arms raised, at a loss.",
"character": {
"who": "A young woman",
"face": "The face from @Image1.",
"hair": "The hair from @Image1, worn straight and loose.",
"build": "Slim tall build with slightly elongated legs, editorial-fashion proportions, realistic waist. Stylization applies to body build; the face stays as @Image1."
},
"wardrobe": {
"outfit": "The clothing from @Image2, worn as shown in that image.",
"eyewear": "She wears the eyewear from @Image3 on her face in the very first frame and in every frame through 0:05.5. It sits on the bridge of her nose with the temple arms over her ears and the lenses covering her eyes. It is one single pair and it matches @Image3 in every visual detail. It stays on her face until the collision knocks it off at 0:05.5.",
"head_and_face": "The @Image3 eyewear is the only thing on her face or head."
},
"hands": {
"rule": "Her hands are loaded for exactly one stretch of the video: 0:00 to 0:05.5. They are empty for the rest.",
"left_hand_0to5_5": "The cup from @Image4, gripped around its body, turned so the camera sees the same side of it that faces the viewer in @Image4.",
"right_hand_0to5_5": "The old flip phone, held open, screen facing her, thumb resting near the keypad.",
"release_at_5_5": "The impact breaks both grips at the same instant. Her fingers spring open and both objects leave her hands and travel away from her body into the air.",
"after_5_5": "From 0:05.5 to 0:12 both of her hands are bare and empty in every single frame. Her fingers stay open and her palms stay clear. The cup and the phone are airborne, then on the asphalt — they are away from her body from the moment of impact onward."
},
"props": {
"coffee_cup": {
"identity": "The cup from @Image4.",
"appearance_source": "@Image4 only. Every visual detail of the cup is read off the reference image.",
"presence": "Held in her left hand from frame one, turned so the camera sees the same side of it that faces the viewer in @Image4. It is one single cup and it looks identical to @Image4 at the 0:02-0:04 reveal, at the moment it is knocked loose at 0:05.5, and through its close-up at 0:09.2-0:10.1."
},
"flip_phone": {
"identity": "An old early-2000s silver-grey Motorola-style clamshell flip phone.",
"description": "Chunky rounded plastic-and-metal body, brushed silver finish, thick visible hinge, a small secondary screen on the outer lid, a small bright colour screen above a physical rubber keypad, scuffed and worn from age. Its surfaces are blank — plain brushed metal and plastic.",
"presence": "Held open and low in her right hand from frame one; she glances down at its little screen as she walks. It leaves her hand at 0:05.5, torn loose by the impact."
}
},
"location": {
"setting": "A wide multi-lane downtown avenue in a large modern metropolis. Tall glass-and-steel towers and high-rise office blocks rise on both sides and recede in a deep perspective toward the horizon, their facades catching the low sun in long vertical reflections.",
"street_level": "Painted crosswalk stripes and lane markings on dark asphalt, traffic lights, tall street lamps, illuminated shopfronts and glowing signage panels rendered as abstract blocks of color thrown far out of focus, a few parked cars along the kerb, a bus shelter, faint steam drifting up from a manhole grate.",
"traffic": "The lane she walks down is clear and open. The avenue reads as a big city caught in a quiet moment.",
"time_and_light": "Golden hour just after sunset. Warm amber sky low on the horizon burning down the canyon between the towers, cool blue at the top of the frame, warm practical lights already on in the windows and along the street. Long shadows stretching across the asphalt.",
"other_people": "Exactly one passer-by: a man in an olive jacket walking the opposite way toward the camera's back, joined near the end by a small black dog on a leash.",
"anonymity": "The architecture, storefronts and signage are invented and anonymous — generic modern city, abstract unreadable signage."
},
"style": {
"look": "Cinematic full-frame look, 35mm lens character, shallow depth of field, background softly out of focus throughout.",
"grade": "Warm golden grade with deep contrast, subtle film grain, natural lens flare from the low sun. Deep saturated blacks, creamy neutrals.",
"camera_behavior": "Handheld-smooth, stabilized camera with steady deliberate moves. Realistic motion blur on the falling objects.",
"handoffs": "The falling objects hand the frame to each other seamlessly. At every handoff both objects share the frame: the outgoing object drifts toward the edge and softens while the incoming one settles into its place and the focus racks onto it. The camera stays in one continuous drifting move through the whole falling section.",
"mood": "High-end fashion campaign shot on a real city avenue at golden hour. Effortless elegance interrupted — an expensive look undone in one second by a stranger's shoulder. Editorial, cinematic, unhurried."
},
"timeline": [
{
"t": "0:00-0:02",
"speed": "normal",
"camera": "Almost at ground level, tilted up, framed on her footwear and lower legs only. The camera retreats backward at her walking speed so the distance between them stays constant.",
"action": "She walks toward the lens. Heels strike the asphalt, the leather catching the warm light.",
"state": "Already established above the frame line: the @Image3 eyewear on her face, the @Image4 cup in her left hand, the open flip phone in her right hand. She carries these three things through the whole walk."
},
{
"t": "0:02-0:04",
"speed": "normal",
"camera": "Rises smoothly from ground level to hip height, then to chest height, still retreating. The framing change comes from the camera rising and retreating while she keeps a steady walk.",
"action": "The rise reveals in order: the lower half of the @Image2 outfit, then the @Image4 cup in her left hand, then her torso, then the open flip phone held low in her right hand with brushed silver catching the warm light.",
"reference_check": "@Image4 lands here. The cup entering the frame is the @Image4 cup, matching that image in every element the instant it appears.",
"background": "The avenue and its towers stretch away behind her in deep perspective, out of focus."
},
{
"t": "0:04-0:05",
"speed": "normal",
"camera": "Medium shot, slightly below eye level, her full upper body and face in frame.",
"action": "Her face is the @Image1 face and the @Image3 eyewear covers her eyes, frames and lenses matching the reference, low sun catching the lens surface and the edge of the frames. She is absorbed by the phone: head tipped down toward the flip phone in her right hand, she glances at its little screen, then looks up and off to the side at something past the frame, wearing a light easy smile, relaxed and unbothered, her eyes away from the road ahead.",
"background": "Behind her right shoulder the man in the olive jacket walks toward her on a collision course, entering the frame from the right edge and closing the distance fast."
},
{
"t": "0:05-0:05.5",
"speed": "normal",
"camera": "Holds the medium shot.",
"action": "The man's shoulder strikes hers hard as he passes. Her upper body is knocked sideways and twists, her head snaps with the impact, her front heel catches the asphalt and skids, and she staggers a step while staying on her feet. He walks on at a constant pace and exits the frame."
},
{
"t": "0:05.5-0:06",
"speed": "ramping into slow motion",
"camera": "Begins to drift in toward the falling objects.",
"action": "The jolt tears all three loose at once: the whip of her head knocks the @Image3 eyewear clean off her face and it flies away from her, the @Image4 cup is knocked out of her left hand, still matching @Image4, and the flip phone is jolted out of her right hand. Both objects leave her grip completely at that instant and fly away from her, and her hands are bare from this frame on. Her bare eyes are revealed for the first time. Her lips part in a startled gasp, both arms fly out to catch her balance, fingers spread and hands empty."
},
{
"t": "0:06-0:07.3",
"speed": "slow motion",
"camera": "Close-up, drifting with the eyewear as it comes toward the lens.",
"action": "The @Image3 eyewear tumbles toward the lens while rotating slowly on its own axis, both temple arms swung fully open and trailing behind the frames, sweeping through the light as it turns. It grows in the frame, its surfaces sharp and in focus, matching @Image3 in every detail.",
"hero_rotation": "The rotation is deliberate and readable: the eyewear turns from a straight-on view of the front of the frames, through three-quarter, into full side profile, presenting the entire length of one temple arm flat to the camera, exactly as @Image3 shows that temple.",
"background": "She is behind it, defocused, still recovering her footing, her eyes bare and both of her hands empty and open."
},
{
"t": "0:07.3-0:08.4",
"speed": "slow motion",
"camera": "Holds and lets the eyewear drift past at its closest point, filling most of the frame.",
"action": "The eyewear continues its slow rotation through a further turn as it passes: the open temple arms sweep across the frame one after the other, the hinge detail comes into view, warm sunlight rakes across the lenses and catches the edges of the frames and the full length of the temple.",
"hero_rotation": "Keep the spin slow and continuous so the frames, the lenses, both open temple arms and their surfaces stay individually readable throughout the pass, each one matching @Image3.",
"background": "Soft warm bokeh of street lights and lit tower windows."
},
{
"t": "0:08.4-0:09.2",
"speed": "slow motion",
"camera": "Tilts and follows the eyewear down as it keeps falling, staying in close-up on it.",
"action": "The eyewear keeps rotating slowly as it falls, temple arms held open and flexing.",
"background": "Behind it, deep out of focus, the coffee cup enters the frame from above and starts falling into the same space. She is far behind, out of focus, her hands empty."
},
{
"t": "0:09.2-0:10.1",
"speed": "slow motion",
"camera": "One continuous drifting move, focus racking from the eyewear to the cup.",
"action": "The @Image3 eyewear drifts toward the bottom edge and softens as the focus racks onto the @Image4 cup, which settles into the center of the frame. Both share the frame through the transition. The cup tumbles toward the camera exactly as @Image4 shows it, dark coffee arcing out of it in a rope of liquid, droplets suspended along its side.",
"reference_check": "@Image4 again in close-up: its surface, colors and whatever graphics @Image4 shows on it are sharp and match that image exactly as it turns."
},
{
"t": "0:10.1-0:10.8",
"speed": "slow motion",
"camera": "Focus racks onto the flip phone; the camera begins to pull back while staying on it.",
"action": "The cup drifts out of the bottom of the frame and the flip phone floats into its place from the side, tumbling end over end with its lid flopped wide open, the little screen and brushed silver hinge catching the low sun."
},
{
"t": "0:10.8-0:11.3",
"speed": "returning to normal",
"camera": "The pull-back accelerates straight through the flip phone, opening up into a wide shot.",
"action": "The flip phone softens in the foreground as the wide shot opens up behind it: she is straightening up in the middle of the road, catching her balance, both hands raised palms-up, bare and empty, fingers spread, her face bare and her eyes visible. All three objects are still airborne a few centimeters above the asphalt."
},
{
"t": "0:11.3-0:12",
"speed": "normal",
"camera": "Holds wide.",
"action": "The objects land. The coffee bursts across the asphalt in a dark radial splash with flying droplets. She stares down at the mess, then glances back over her shoulder toward where the man went — eyes bare, eyebrows lifted, lips slightly parted, completely at a loss. Her hands stay empty and open at her sides. She keeps breathing and shifting her weight. Behind her the man in the olive jacket walks away with a small black dog.",
"final_frame": "Wide, with the avenue, the receding towers and a glowing street lamp in the background."
}
],
"sound": {
"ambience": "Quiet evening big-city ambience, distant traffic on other avenues, a faint far-off siren, faint birds.",
"normal_speed": "Sharp heel clicks on asphalt.",
"impact": "A dull heavy thud of shoulder against shoulder and a short scrape of a heel skidding on asphalt, then the sound drops into a low muffled slow-motion rumble.",
"slow_motion": "Soft overlapping air whooshes as each object passes the lens.",
"landing": "Wet heavy splash as the coffee hits the road and the hard hollow clatter of a plastic flip phone bouncing on asphalt. Distant dog collar jingle at the end.",
"music_and_voice": "Ambient sound design only, instrumental at most."
},
"must_hold_true": [
"The whole video is one continuous shot. Every change of framing comes from camera movement and from objects passing close to the lens.",
"At each handoff both objects share the frame at the same time, one softening as the other comes into focus, inside one unbroken camera move.",
"The @Image3 eyewear rests on her face over her eyes in every frame from 0:00 through 0:05.5, seated on the bridge of her nose with the temple arms over her ears.",
"The eyewear leaves her face at 0:05.5, knocked off by the shoulder impact, and stays off for the rest of the video. It appears in exactly one place in every frame: on her face before 0:05.5, away from her after it.",
"Each of the three objects appears in exactly one place in every frame. From 0:05.5 onward the cup and the phone are airborne or lying on the asphalt, and her hands are bare and empty — an object that is in the air is only in the air.",
"Her hands are loaded only between 0:00 and 0:05.5. From 0:05.5 to the end of the video both hands are empty, fingers open, palms clear, in every frame she appears in.",
"While falling, the eyewear rotates slowly and continuously with its temple arms held open, so that both temple arms, the hinges and their surfaces are individually readable and match @Image3. Any mark or text on the temples is the one @Image3 shows, at the size and position @Image3 shows it.",
"Her left hand holds the @Image4 cup and her right hand holds the flip phone, from 0:00 to 0:05.5. These are the only two objects she carries.",
"From 0:05.5 to the end her hands are empty and open and her eyes are bare, with her expression fully readable.",
"The cup is a copy of the cup in @Image4 and takes every visual detail from that image, identical at the 0:02 reveal, at the 0:05.5 release, and in its 0:09.2 close-up. Any graphics on it are the graphics @Image4 shows, at the size and position @Image4 shows them.",
"Her clothing is the @Image2 outfit in every shot she appears in.",
"The phone is a chunky clamshell that opens on a hinge, with a physical rubber keypad and blank unprinted surfaces. She holds it in her palm and looks down at it.",
"The eyewear from @Image3 gets the longest and closest screen time of any falling object and is the only one given a full hero close-up.",
"The man's shoulder makes real physical contact with hers. He walks on at a constant pace, facing forward, and exits the frame.",
"Every object leaves her because the impact tears it loose. Her grip is broken by force.",
"Her expression runs three phases: a light easy smile while walking from 0:00 to 0:05, a startled open-mouthed reaction at the impact, and blank bewildered confusion with lifted eyebrows in the final wide shot.",
"Her face stays the @Image1 face with natural head-to-body proportion throughout.",
"Five fingers on each hand with correct anatomy. If the @Image2 outfit includes gloves, they fit the hands naturally.",
"She keeps small natural motion in every shot, including while the objects fall.",
"The avenue is invented, anonymous modern-city architecture with abstract unreadable signage.",
"Slow motion runs from 0:05.5 to 0:10.8. Everything before and after runs at normal speed.",
"Wind moves her hair; her clothing hangs still and holds its shape.",
"The frame is clean live-action photography edge to edge, carrying picture only.",
"She appears exactly once in every frame, as a single person.",
"Both of her hands stay in view with five fingers each and natural anatomy whenever they are framed."
]
}