REFERENCES @image3 controls only the man's face, hair colour, skin, eye colour and build. Do not copy wardrobe, background, pose, framing or lighting from @image3 . @image1 controls only the pack geometry, colour, finish and all label text. Do not copy background, hands, lighting or scale from @image1 . SCENE Late afternoon on a quiet country road. @image2 and @image3 are sitting side by side on the hood of a classic late-1980s car, @image3 on the left, @image2 on the right, both turned slightly toward each other with their knees almost touching. The car's paint, badge and grille read clearly below them. Behind them the road falls away into blurred trees and a low sun, and the frame holds only the two of them, the hood, and the treeline. WARDROBE @image3 wears a faded blue denim button-up open over a plain white T-shirt, high-waisted relaxed blue jeans and a worn leather belt. His hair has authentic late-1980s volume. @image2 wears a loose oversized white blouse with rolled sleeves tucked into high-waisted light-blue denim jeans, and a small vintage wristwatch. Her hair is styled with late-1980s volume and body, colour and length exactly as @image2 . Both are dressed and groomed as real people in 1988, never as a costume. LIGHT One condition, unchanged across every beat. Low warm afternoon sun from frame-right, flaring gently at the edge of the lens. Warm rim on the sides of their hair and jawlines. Soft bounce off the pale car hood filling the shadow side of both faces. Subtle atmospheric haze between them and the treeline. No fill light, no studio lighting. CAMERA Locked-off eye-level two-shot on a 50mm lens, chest-up on both of them, with one very slow push-in across the whole clip. Shallow depth of field so the treeline stays soft. Slight handheld weight, never a visible operator move, never a cut. 0:00-0:03 @image3 turns toward @image2 and looks at her with a small affectionate expression. @image2 is already looking back at him, her hand flat on the hood between them. The low sun catches the edge of her hair. Neither of them speaks. A single slow blink from @image2 . 0:03-0:05 @image3 leans in slowly toward @image2 . @image2 reads it, lifts her chin, and leans in to meet him. Both sets of eyes lower. They stop a few centimetres apart. Her fingers curl slightly against the hood. Neither of them speaks. 0:05-0:06 @image2 stops dead. Her eyes snap open and she pulls back one sharp inch, her nose wrinkling. @image3 stays exactly where he is, eyes still closed for a beat, then opens them. 0:06-0:08 @image2 holds @image3 at arm's length with one flat palm on his chest and looks straight at him. Her mouth moves only for her own line. At 0:06 she says, flatly and a little appalled: "Wait, did you eat garlic?" 0:08-0:09 @image2 turns her face away an inch and fans the air once with her free hand. @image3 stays where he is, still leaning, and starts to realise what she means. At 0:08 she says, short and disgusted: "Ew." 0:09-0:10.5 @image2 keeps the palm on his chest and digs into her front jean pocket with the other hand. @image3 opens his mouth to answer and closes it again. At 0:09 she says, without looking at him: "I'm not kissing that." 0:10.5-0:12 @image2 brings @image1 up between them and thumbs one piece free of the pack. The label on @image1 crosses the frame in focus for this beat and is sharpest here. @image3 watches her hands, eyebrows going up. At 0:10.5 she says, business-like: "Here, chew this first." 0:12-0:13.5 @image2 puts the @image1 straight into @image3 hands. He lets her, too surprised to stop it. She sits back and folds her arms, while @image3 still in @image3 hands. One object only, never a second pack anywhere in frame. At 0:12, starting before opening @image1 , @image3 says, defensively: "It was one piece of bread." 0:13.5-0:15 @image3 opens @image1 , slowly, and looks at @image2 completely bewildered. @image2 raises her eyebrows at him and does not move. At 0:13.5, @image3 says: "One." From 0:14.2 to the last frame neither of them speaks. They hold the look, the sun flares once at the lens edge, and the car and treeline are unchanged. PERFORMANCE Played straight, never mugged. The first three seconds are a genuine romantic beat with no comedy in them, and the turn at 0:05 is the whole joke, so it must be sudden and physical rather than acted out in advance. @image2 is practical rather than cruel about it. @image3 is not offended, only confused, and stays confused to the last frame. Real weight shifts, real breathing, small unconscious adjustments on the hood. VOICE @image2 : female, warm and bright, unhurried American. @image3 : male, dry and slightly flat American, lower placement. Each mouth moves only during that person's own lines. Realistic lip articulation, no exaggerated mouth opening, no head turns while speaking. Natural breaths. REALISM Pore-level skin realism on both faces, fine texture, faint natural shine on the forehead in the warm light, real eyebrow and lash detail, wet living eyes with catch-lights and natural asymmetric blinking. Flyaway hairs catching the low sun. Denim shows real weave, creasing and wear. The car hood shows fine swirl marks and dust in the raking light. PHYSICS Both bodies carry weight on the hood and the suspension settles very slightly when @image2 sits back at 0:12. Hair moves with real inertia when she pulls back and a few strands fall across her face. Cloth creases and drapes naturally under the lean. @image1 has real weight and correct contact shadow against her hand. Correct contact between her palm and his chest. FILM LOOK Authentic 35mm analog film capture. Warm muted vintage colour, slightly faded highlights, gentle brown and beige tones, low-to-moderate contrast, subtle halation around the sun and the brightest edges, fine authentic film grain, mild lens imperfections, slight softness away from the centre, faint gate weave. It should read as a genuine home-shot film clip from 1988, imperfect and warm. AUDIO Natural ambient only: warm outdoor air, distant birds, a car passing far off, the faint creak of the hood under their weight, the small crackle of the pack. Ambience ducks slightly under each line. LOCKS Same face, hair colour and length, skin and eye colour as @image2 for the woman in every beat, first frame to last. Same face, hair colour, skin and eye colour as @image3 for the man in every beat, first frame to last. Same wardrobe on both from first frame to last, fully covered throughout. @image1 stays one object, correct one-hand scale, label text correctly spelled and legible whenever in focus, sharpest in the 0:10.5-0:12 beat. Exactly two people in frame, four arms and four hands total, anatomically correct with five fingers each. Same car, same road, same warm low sun from frame-right, first frame to last. Spoken lines exactly as written, in order, at the stated timings, with accurate natural lip-sync, and each line spoken by the named person only. AVOID No cuts. No slow motion. No repeated action. No second or duplicate product. No modern cars, no modern clothing, no modern hairstyles, no contemporary styling. No digital sharpness, no CGI look, no glossy studio lighting, no retouching, no skin smoothing, no commercial polish. No on-screen text, no captions, no subtitles, no music. No logos other than the product's own packaging. No third person in frame and no reflection of the camera anywhere in the shot.
Drag the clip sideways
Create a cinematic 10-second premium advertisement for a grey fabric sphere speaker with a light ring, vertical 9:16. a grey fabric sphere speaker with a light ring is the exact speaker in the reference: a grey fabric sphere on a white circular base with a glowing cyan LED light ring around the bottom. Same shape, mesh, base and ring in every shot. One object. Ignore any brand marks — describe the object, do not name Alexa/Amazon/Echo. 0–2s: Dark modern room. Camera slowly pushes toward a grey fabric sphere speaker with a light ring as the ring softly illuminates cyan. 2–4s: Macro of the glowing ring, smoothly shifting ambient colors (cyan, warm white, magenta). 4–6s: Camera follows a grey fabric sphere speaker with a light ring being placed on an elegant outdoor table at sunset. Ring still lit. 6–8s: Modern bedroom, a grey fabric sphere speaker with a light ring on a nightstand, cozy ambient ring light. 8–10s: Hero, slow 360 around a grey fabric sphere speaker with a light ring, cinematic lighting. Ring is the star. Photorealistic commercial, smooth camera, shallow depth of field, no on-screen text. AUDIO required: soft whoosh, a gentle chime when the ring lights, quiet room tone. No voiceover. Duration 10s. Aspect 9:16. OPENING: already in the dark room with the speaker — no grey still.
TWO LEADS — PHOTOCOPY BOTH STILLS. Do not invent faces. Do not swap them. @image1 = FEMALE LEAD. Photocopy her face, bone structure, skin, long blonde hair, green eyes, body. ALSO photocopy her outfit from that photo (the prompt has no wardrobe for her): burgundy corset bodice with black leather harness straps and silver rings, high-slit burgundy skirt, ivory chiffon cape that ombrés to blood-red at the hems, black gladiator lace-up sandals, small hoop earring. She fights with a jian / straight sword. Same woman and same clothes every frame. @image2 = MALE LEAD. Photocopy his face, bone structure, skin, wavy dark-brown hair to the jaw, body. ALSO photocopy his outfit from that photo: wide black kasa/sugegasa straw hat, black kimono and hakama, black sash, katana at the left hip, black tabi. Same man and same clothes every frame. The hat can tilt in motion but stays on. Never give him her burgundy dress. Never give her his kasa. Never a third human. Never OPENCLIPS.AI type. SETTING: vast ornate temple hall from the source film — gilded dragon columns, wet reflective stone floor, thousands of candles, teal mist, high vault. 16:9. 30 seconds. One continuous action sequence. No dialogue, no subtitles, no on-screen text. LIGHTNING (required on every clash): whenever the two swords collide, a bright white-blue lightning arc flashes along both blades and a fork of electricity cracks the air at the contact point — not ordinary orange sparks alone. Every bind, parry and strike that meets steel = lightning. CORE: zero-delay chaining; scene changes only via their bodies (slam a column, leap from a pillar, crash through candelabra); real weight — blocked hits shove the other back, feet slip, bodies twist. Camera follows limbs; never orbit a frozen pose. BEATS (30s) 0–2s already sprinting from opposite ends of the hall, full-speed clash at centre. Blades meet with a lightning burst; both recoil half a step; floor tiles crack. No establishing pause. 2–7s close melee: he sweeps a wide horizontal cut; she drops almost to the floor, blade skimming her nose, a lock of blonde hair cut; she springs up with a straight thrust to his chest; he spins, wrists reverse to bind. Swords lock, shoulder to shoulder, then both shove off with lightning at the bind. Two steps back. 7–9s the shove sends him into a dragon column; he rebounds, runs vertically up the pillar; she explosive-jumps after. Two or three lightning clashes on the column, blades scarring the gilt. 9–12s he is forced to the beam; her heavy blow knocks him off; he mid-air flips, diving slash; she drops from the column to meet it. Landing clash — lightning + sparks, arms visibly shake. Now they are on the hall floor again. 12–17s densest round: she three accelerating cuts (he sidesteps, then floor-dodge, then ducks under and slashes at her waist). Her cape corner splits. He lunges; she barely slips, fabric tears at the shoulder; she counters so hard his blade rings and the tsuba cracks; he staggers two steps. 17–21s bind at close range, simultaneous break, charge again — they smash a row of candelabra; candles snuff then relight from the sword-wind; they keep fighting through fallen candles and oil, lightning every contact, water-mist from the wet floor. 21–26s peak speed: cut, thrust, parry, counter in two–three seconds, each hit a lightning flash plus body feedback. Teal-green sword qi spirals. No frozen frame. 26–29s simultaneous all-out strike; blades meet; a hall-wide teal-white lightning shockwave hurls both back several metres; tiles and columns crack; candles gutter; camera shake then soft defocus. 29–30s smoke thins. Both still on their feet, one knee or standing, chests heaving, sword-hands shaking from exhaustion. Remaining candles relight. Wide two-shot. Hold. No titles. AUDIO: steel, thunder-crack on every clash, stone, fire, breath, shockwave. No BGM. No speech.
subject_definitions: @image_1 is the first performer. Gender, face, hair, skin and body proportions come entirely from @image_1 and not from this text. @image_1 defines identity only - its clothing, pose, framing, background and lighting are ignored. @image_2 is the second performer. Gender, face, hair, skin and body proportions come entirely from @image_2 and not from this text. @image_2 defines identity only - its clothing, pose, framing, background and lighting are ignored. summary: [reference generation] The target video is a 15-second 16:9 opening title sequence in which @image_1 and @image_2 live separate nights in the same city, never meet, and are connected only by one small yellow paper ticket, cut together as a punk editorial collage and ending on a single title card and a two-face poster frame. retention_analysis: @image_1 (appears in [Shot 1], [Shot 2], [Shot 5], [Shot 6], [Shot 8]): fully_preserved - gender, face, hair, skin and body proportions stay exactly as in @image_1 in every shot. @image_2 (appears in [Shot 1], [Shot 3], [Shot 5], [Shot 6], [Shot 8]): fully_preserved - gender, face, hair, skin and body proportions stay exactly as in @image_2 in every shot. detailed_description: The target video is live-action and photographic, cut and printed like an underground magazine: real people and real rooms, then torn paper edges, xerox grain, rough halftone and offset misprint laid over them. The palette is restrained and graphic - deep black, dirty white, cobalt blue and one sharp yellow. Light comes from fluorescent tubes and street lamps only, never from glow effects. @image_1 wears a heavy dark work jacket over a plain pale shirt; @image_2 wears a dark wool overcoat over a plain pale shirt. Both outfits are fixed in cut and colour and only their drape adapts to the body; the gender, face, hair and skin of each performer come from their own reference and never from this text. The two performers never share a room and never touch. [Shot 1] A split composition holds two panes of glass side by side. On the left @image_1 stands inside a late-night laundromat behind its window; on the right @image_2 stands behind the window of a passing tram. Reflections of buildings slide across both faces at the same speed, so the two panes read as one continuous street. The camera holds a static shot. [Shot 2] At 00:02.000, the shot cuts to a close shot of @image_1's hands lifting a folded shirt from a machine, and a small yellow paper ticket falls out of it onto the metal counter. The camera pushes in slowly onto the ticket. [Shot 3] At 00:03.800, the shot cuts to an identical yellow paper ticket held between @image_2's fingers against the tram window, matched in size, angle and position to the previous frame. The camera holds a static shot. [Shot 4] At 00:05.600, the shot cuts to a fast burst of fragments printed as torn magazine panels: an eye, a hand on a rail, two pairs of shoes walking in opposite directions, tram doors closing, a washing drum spinning, overhead wires against a black sky. Photocopied city maps and rough black marker strokes slide between them. The camera shakes slightly at fast speed. [Shot 5] At 00:08.000, the shot cuts to @image_1 and @image_2 standing in the same frame but in separate torn panels. Each figure breaks apart into xerox slices and offset print layers, then reassembles in the opposite corner of the frame. Both faces stay whole and readable while everything around them fragments. [Shot 6] At 00:10.000, the shot cuts to a wide street view as the tram passes directly in front of the laundromat window. For one instant @image_1 and @image_2 are aligned on opposite sides of the glass, both in profile, looking past each other. The camera trucks left with large amplitude at fast speed. [Shot 7] At 00:11.800, the image freezes and the surrounding paper fragments, maps and marker strokes collapse inward and assemble into a single title card. Cut and pasted letters spell "NIGHT SHIFT" across the middle of the frame in dirty white with the second word in sharp yellow. This is the only text in the entire video. [Shot 8] At 00:13.400, the shot cuts to a final poster frame: @image_1 on the left and @image_2 on the right, both turned to face the camera straight on, evenly lit, faces clean and unfragmented and filling the frame, the pane of glass between them and the small yellow ticket centred on it. The camera holds a static shot through the final frame. overall_soundscape: A late city night: a washing machine drum turning and stopping, tram wheels on rail joints, doors hissing shut, a fluorescent tube buzzing, footsteps on wet pavement, paper tearing and the flat snap of a photocopier. non_diegetic_music: A driving instrumental built on a dry drum machine, a single distorted bass line and one repeating minor piano figure at 118 BPM. The arrangement stays at one even, moderate level from start to finish, with hits landing on the ticket falling, the fragment burst, the tram passing and the title card assembling.
CHARACTER LOCK — HIGHEST PRIORITY: The person in every shot is exactly the person in @image1. Preserve their facial identity, facial structure, eye shape and spacing, nose, mouth, exact skin tone, hairline, hairstyle, hair length, texture and colour, freckles and distinctive marks, apparent age, build and body proportions — identically, shot to shot. Preserve their gender presentation exactly as shown; do not masculinize, feminize or reinterpret it. Do not beautify, slim, de-age or restyle them. The attached image outranks every word below it: wherever the image and the text disagree, follow the image. Frame 1 is already in the scene — never a studio portrait, grey backdrop, or the attached still as an opening plate. Use they/them throughout. No other people. A warm hand-painted pastoral anime sequence in a quiet apple orchard at the start of autumn. The person wears a soft cream blouse and a rust-colored skirt and walks beneath trees heavy with red apples. Golden afternoon sunlight, first hints of autumn, rich painted textures, gentle breeze, nostalgic countryside atmosphere, simple graceful actions, highly detailed. No on-screen text. Shot 1: Wide establishing shot, slow push-in. Rows of apple trees glow in warm autumn sunlight. Frame 1 is already in the orchard — never a studio portrait. Shot 2: Low-angle shot, static. Red apples hang among green leaves against a pale blue sky. Shot 3: Medium rear shot, gentle tracking. The person walks slowly between the trees. Shot 4: Close-up, static. Their skirt brushes softly against tall grass. Shot 5: Medium profile shot, static. They pause beneath one tree. Shot 6: Close-up, static. A ripe red apple hangs from a low branch. Shot 7: Medium shot, slight push-in. They reach toward the apple. Shot 8: Close-up, static. Their hand gently picks it from the branch. Shot 9: Close-up portrait, static. They look at the apple and smile. Shot 10: Wide side shot, slow pan. Leaves move gently around them. Shot 11: Medium rear shot, static. They continue walking through the orchard. Shot 12: Close-up, static. Sunlight flickers across fallen leaves on the grass. Shot 13: Low-to-high shot, gentle tilt. Branches sway overhead. Shot 14: Wide rear shot, slow pull back. They walk toward the far end of the orchard. Shot 15: Far wide shot, static. The orchard rests beneath soft golden evening light. CAMERA: slow, simple, graceful. No teleporting. AUDIO: soft breeze, leaves, quiet countryside. No voiceover. NEGATIVE: no live-action photoreal, no modern objects, no text, no watermark, no extra people.
A warm, heartwarming beach scene at golden hour: @image1 gently feeds and befriends a small stray cat by the shoreline, coaxing it closer until it eats from their hand and settles beside them. Photoreal, tender, soft natural light, gentle waves and a light music bed, no dialogue.
ONE SHOT. VERTICAL. ONE PRODUCT ONLY. LUXURY PRODUCT FILM. ASSET MAP - every reference is bound here and cited by tag below: @image_1 - the product. This is the only product in the film; its shape, colour, finish, material, proportions and every detail come entirely from @image_1 and never from this text. PRODUCT LOCK - HIGHEST PRIORITY: the product on camera is exactly the product in @image_1. Preserve its silhouette, proportions, colour, finish, surface texture and every constructed detail identically from the first frame to the last. Do not restyle it, do not redesign any part of it, do not change its colour or its material, and do not add a marking, a logo, a label or lettering that is not already on it. One single product, alone, for the whole film. ATTACHED IMAGES ARE MANDATORY AND FINAL: if a reference is attached, that is the product - reproduce it exactly, in every frame, and never invent a different one. The attached reference outranks every word written below it. WORLD: a high-end studio with no visible walls, floor or horizon - only deep shadow falling away behind the product. Large soft key light from the upper left and a narrow rim along the back edge, so the surface carries one long clean specular highlight that travels as the camera moves. A tight contact shadow holds the product to the ground. Behind it, a single panel of heavy silk drifts slowly across the darkness, catching light in soft folds and never touching the product. Fine dust-fine sparkle particles hang in the air and drift, lit only where the key catches them. The colour of the film comes from the product itself: whatever dominant colour @image_1 carries, the silk, the rim light and the particles take that same colour, and the rest stays deep near-black. Shot 1: THIS IS THE ONLY SHOT AND IT RUNS FROM THE FIRST FRAME TO THE LAST - there is no cut anywhere in this film. The product is already in frame, already lit and already large at the very first frame, held upright and perfectly still. The camera begins in a slow macro push-in toward it, close enough that the surface texture and the light travelling across it are readable, then without stopping eases into a smooth orbit around the product, circling it at the same slow speed while the specular highlight slides along the surface and the contact shadow rotates with the move. The silk behind keeps drifting, the particles keep floating, and the product itself never moves, never tilts, never spins and never leaves the centre of frame. The camera comes to rest on a clean three-quarter view and holds there to the final frame. AUDIO: a very low room tone, one soft sustained pad that stays at a single even level, a faint high shimmer as the particles drift, and the quiet rustle of heavy silk moving in the dark. No voice, no narration, no lyrics. Do not generate subtitles, captions, on-screen text, a caption card, a timestamp, a recording indicator, an interface overlay, a logo or a watermark. No second product, no packaging, no box, no hands, no person, no readable lettering or numerals anywhere in frame, no cuts, no reflections that resolve into a room or a face, and no change to the product's shape, colour or finish at any point.
CHARACTER LOCK — HIGHEST PRIORITY: The person in every shot is exactly the person in @image1. Preserve their facial identity, facial structure, eye shape and spacing, nose, mouth, exact skin tone, hairline, hairstyle, hair length, texture and colour, freckles and distinctive marks, apparent age, build and body proportions — identically, shot to shot. Preserve their gender presentation exactly as shown; do not masculinize, feminize or reinterpret it. Do not beautify, slim, de-age or restyle them. The attached image outranks every word below it: wherever the image and the text disagree, follow the image. Frame 1 is already in the scene — never a studio portrait, grey backdrop, or the attached still as an opening plate. Use they/them throughout. No other people. @image1 is identity only — face, bone structure, skin, long blonde hair, green eyes, age, body. Not wardrobe. Do not keep the white crew-neck t-shirt from the photo. Do not give them dark hair or bangs. WARDROBE (from the source film): white sleeveless fitted ribbed tank; light-blue wide-leg jeans; single-strand pearl choker; thin silver rings; a delicate silver bracelet on the left wrist. Bare arms. Same outfit the full 30 seconds. SETTING: bright modern bedroom, unmade white duvet, white pillows, cream wall, golden-hour sun from camera-right throwing a sharp silhouette of the person on the wall. Handheld smartphone UGC, subtle shake, cinematic commercial grade. One person only. No on-screen text, no logos, no subtitles, no watermarks, no end card. PRODUCTS (match the source clip, no brand marks on the devices): - A book-style foldable phone: closed it is champagne-silver on the lower back with a glossy black camera bar, vertical triple cameras on the left of that bar. Opens like a book to a large inner screen. - A white rectangular retail box on the bed. - A gold-case smartwatch with a white sport band. 30 SECONDS. Native room tone + their spoken lines only. No BGM. BEATS AND LINES (lip-sync these exact lines): 0–5s — already sitting on the bed facing camera, box and watch on the duvet. Smiling. Says: "I genuinely wasn't expecting to love this foldable iPhone this much." Picks up the box, opens it, lifts the folded phone. 5–9s — holds the closed phone toward camera, rotates it slowly (thin frame, hinge, camera bar). Says: "The design already feels incredibly premium." Then: "The design is so sleek, and I really love how the foldable screen opens up." Unfolds it. 9–16s — phone fully open, colorful app grid on the inner display. Scrolls, opens camera, points it around the room. Says: "And the camera quality is seriously impressive. Everything looks so sharp and detailed." Closes and opens the phone once. Says: "Plus, the battery is supposed to be better too, which is honestly a huge win for me." 16–23s — sets the phone on the bed, picks up the gold-and-white smartwatch, holds it close to camera (case, crown, strap). Says: "And the new Apple Watch is actually really nice too." Buckles it onto their left wrist, raises the wrist to camera, scrolls the face. Says: "I love how clean it looks, and having all these features right on my wrist is so convenient." Turns the wrist in the sunlight. 23–30s — picks the folded phone back up. Holds phone in the right hand and shows the watch on the left wrist. Looks into camera, warm smile. Says: "Honestly... this might be one of my favorite Apple upgrades yet." Soft laugh, glance down at both, then holds the phone beside their face with the watch clearly on the wrist. Slow push-in. Last frame still in the bedroom — no fade to black, no titles. Frame 1 is already on the bed in the scene — never a studio portrait or the attached still. Photoreal skin, real hands, five fingers. Same foldable and same watch the whole time.
Lovers QuarrelRecreate workflow