@image1 is the first frame and @image2 is the last frame. The product's identity comes entirely from @image2 — its shape, colour, material, finish and every marking on it are taken from there and are never invented, substituted or restyled. The product, its position, the tabletop, the reflection and the framing are identical in both frames and do not change at any point. One continuous locked-off take. The product is revealed by light arriving on it. The product holds as a dark silhouette while a soft key light begins to come up from the front, catching only its nearest edge. The key keeps rising and the product's form separates out of the shadow — curvature, edges and depth become readable while the surface is still dark. The light climbs across the face of the product and the true colour, material and finish of @image2 emerge out of the dark. The key reaches full and the product stands exactly as it appears in @image2, completely lit, colour true and every marking clearly readable, the tabletop reflection matching it below. The backdrop stays dark and even throughout and does not brighten, bloom or form a glow behind the product. The product keeps its position and orientation, the camera holds its frame, and the light rises continuously and evenly from first frame to last. No cut, no zoom, no push, no orbit.
Drag the clip sideways
@image1 is an absolute product lock: all three pieces in it keep their exact form, proportions, material, finish, colour and every surface detail, and must not drift, restyle, merge or swap with one another at any point. Exactly one piece is on screen at a time, and each appears once. Follow @image2 exactly — its shot order, framing progression, hand position and staging. Soft editorial product film, shot on a long lens wide open. High-key white, blown-out daylight from above, heavy bloom, warm-neutral white balance, very shallow depth of field, delicate tissue and fabric texture held in the mid-tones, no hard shadows anywhere. Calm, intimate, unhurried. One continuous take, no cuts. The whole film is shot from inside a bag of crumpled translucent white material, looking out through one irregular soft-edged opening torn through it. That material fills the foreground on every edge, far out of focus and glowing, and the opening is blown out to pure white. The camera stays down at the bottom of the enclosure with the lens angled steeply up at that opening, and creeps steadily nearer it across the film so it grows in frame, never pulling back and never leaving the enclosure. RIGID ATTACHMENT RULE: once the fingers close on a piece, that piece is locked to the hand and moves ONLY because the hand moves. It never turns, spins, rotates, tilts, drifts, floats, sways or changes its angle on its own. Every change in how it faces the camera comes from the wrist rolling or the fingers rolling it — hand and piece travel as one rigid unit, with the piece keeping exactly the same grip position in the fingers from the moment it is picked up to the moment it leaves frame. If the hand is still, the piece is still. THE HAND IS NEVER STILL AND NEVER POSES. It is always travelling — descending, closing, rising, or leaving frame. No held beats, no static presentation, no hovering in place. The action is one simple cycle, repeated three times: the hand reaches DOWN into the bag toward the camera, closes on a piece, lifts it UP and carries it OUT of frame to the side — leaving the bag empty for a beat before the hand comes down again for the next one. The hand comes down through the opening toward the lens, filling the frame and covering the light so the picture falls to near-black for a moment. The fingers close on the first piece and the hand rises without pausing, the wrist rolling over as it climbs so the piece it is gripping swings through the light with it, and carries it up and away out of frame. The frame is left empty — folded white material and the bright opening, nothing else. The hand comes down again into the same brief darkness, the fingers close on the second piece, and the hand draws straight back up, the wrist turning as it rises so the gripped piece comes round with the hand, then takes it up and out of frame to the side. The bag is empty again. The hand descends a third and last time, closes on the final piece and climbs again, the fingers rolling it slightly as the wrist turns, then carries it out of frame after the others — leaving the empty bag and the softening light to hold to the end. Each descent covers the lens and darkens the frame, and each beat ends with the piece leaving frame upward. Nothing is ever set back down, no piece reappears once it has gone, and only one piece is ever on screen at a time. Do not animate any product independently of the hand. No spinning jewellery, no floating or hovering pieces, no product rotating in mid-air, no piece turning while the hand holds still, no piece slipping or shifting inside the grip. Audio: soft ambient room tone, the faint rustle of tissue and bag, and a quiet sparse music bed low in the mix. No dialogue, no voiceover. Do not generate any text, subtitles, captions, logos or watermarks.
A cinematic product-creation film follows one maker making the exact object shown in @image1 (product plate) completely from scratch. MATERIAL TRUTH — this governs every stage below: Read the object's materials from @image1 and make it the way that object is genuinely made. The pieces, joins, tools and hand actions in every stage must be the ones its materials actually require. Leather, textile and canvas: cut panels, skiving, linings, interfacing, glue, stitching by hand or machine, edge finishing and burnishing, rivets, zips, buckles, straps and hardware set by hand. Wood: cut stock, joinery, glue, clamps, sanding, oil or lacquer. Ceramic, glass and metal: forming, shaping, joining, grinding and polishing appropriate to the material. Cosmetics, food and liquids: vessel, contents, seal, closure and label applied by hand. Rigid and electronic goods: chassis, internal components, wiring, fasteners, screws and press-fits. A chassis, circuit board, wiring, brackets, screw bosses or screws may appear ONLY if @image1 is itself a rigid or electronic product. If it is not, none of them appear at any point, and no screwdriver ever touches it. If a stage below has no equivalent in this object's making, spend that stage on the nearest real step in its craft rather than inventing a part it does not have. REFERENCE PRIORITY: @image1 is an absolute lock on the finished object's design: overall silhouette, proportions, geometry, curvature, part and hardware placement, seams, materials, surface finish, colors, lighting zones, logo placement, and all other distinctive visible details. Match these features as closely as possible. No drift or reinterpretation in any shot. Ignore @image1's background, framing, lighting, hands, desk, and unrelated objects. ENVIRONMENT: @image2 (environment plate) is the work surface and lighting for the entire film. Take the surface, its finish, the light direction and the falloff from it exactly. STORYBOARD: @image3 (storyboard) is to be followed exactly — matching shot order, framing, hand position and stage of making panel for panel. CAMERA: Use a stable front three-quarter overhead macro product view, approximately 35-45 degrees above the work surface. Keep the object centered and fully readable throughout. Shallow but usable depth of field; the developing object remains sharply readable. ONE MAKER RULE: There is exactly one maker throughout, represented only by the same consistent left and right hands and forearms. Maximum two hands visible at any moment. Every piece must be physically carried, positioned, joined, glued, stitched, fastened, pressed, fitted, or adjusted by these hands. TIMELAPSE: The work occurs as a fast, precise timelapse, returning to normal speed only for the final reveal. Hands move rapidly and purposefully with restrained natural motion blur. Use short jump cuts to compress repetitive work, but ONLY after the current physical action has visibly completed. Every cut must preserve exact continuity. The object must retain all pieces and progress established in the previous shot. No piece may magically appear between cuts. FIRST PIECES: One hand enters frame already holding the object's first and lowest piece — the first cut panel, the base, the first formed piece, whichever its making actually begins with — and lays it precisely at the center of the surface. The second hand steadies it. The next piece is brought into frame by hand and positioned beside or upon it. The object begins at its earliest stage, flat or loosely formed, nothing like the finished shape. BODY: The maker joins the first pieces so the object takes its basic form, working each join by the method its materials require — stitched, glued and clamped, fitted and fastened, or formed and set. Small pieces are brought into frame between two fingers and set into place. Everything remains physically believable and correctly scaled. CLOSURES AND HARDWARE: The maker sets whatever the object closes, fastens, carries or hangs with — the closures, fittings, trims and hardware visible in @image1 — each worked in with the correct tool for that material. The maker tests any moving part with one finger to confirm it works. Nothing activates or moves independently. OUTER SURFACE: The recognizable body of the object now takes shape. The maker works the main outer surface into place over what is beneath it — lowered and pressed, wrapped and smoothed, stitched down or joined, as the material demands. Separate panels, sections, inserts, trims or straps seen in @image1 are fitted individually. Every seam, edge and join follows the geometry of @image1. Any textured, perforated, woven, glossy, matte, metallic or transparent areas in @image1 are produced through visible physical hand interaction. FINISHING: The maker performs rapid final detailing with the tools that material calls for — a needle, a burnisher, a glue brush, a mallet, tweezers, a cloth, and a small screwdriver only if @image1 actually has screws. The hands refine alignment, gaps, seams, edges, surface transitions and small decorative elements. Any remaining visible details present in @image1 are set — inserts, accent pieces, hardware, closures, cabling or labelling. The exterior progressively reaches the exact proportions, colors, materials, surface finish, and distinctive appearance of @image1. One hand steadies the object while the other performs every final adjustment. REVEAL: The maker makes one final precise adjustment, sets the tool down, gently wipes or straightens the object once, then withdraws both hands completely from frame. The accelerated timelapse smoothly returns to normal speed. If @image1 is a powered object it now switches on and any lighting zones visible in @image1 softly illuminate; if it is not, the light on the surface lifts instead and the object settles into its finished pose. The camera performs a subtle, restrained cinematic push-in toward the finished object. Hold on a pristine premium product hero shot, matching @image1 as closely as possible. PHYSICAL CONTINUITY RULES: Every new piece enters the frame already held by one of the maker's hands. Every piece must be visibly placed before becoming part of the object. Pieces remain exactly where they were set. No piece floats or moves independently. No spontaneous material changes. No instant construction. No magical morphing. No disappearing or duplicated pieces. No unexplained changes between cuts. No machined chassis, circuit board, metal rails, brackets, screw bosses or screws unless @image1 is itself a rigid or electronic product. No screwdriver touching leather, textile, canvas or any other soft material. Nothing that resembles an electronics teardown or a repair bench. No more than two hands. No assistants. No detached or floating tools. No drawing or sketching phase. No software interface. No cursor. No menus. No annotations. No labels. No text overlays. The entire sequence must feel like an impossibly fast but physically believable piece of premium making, filmed as a cinematic macro product commercial.
17 seconds, 9:16. Intimate GRWM. Dimly lit room, neutral background. ONE hard cut. Front-facing. ELARA LOCK: @image1 is Elara — photocopy her face, hair, eyes, top, jeans and earrings exactly as in @image1. Adult. Same woman both scenes. IGNORE any logo in the corner of @image1. COMPACT: @image2 is the blush — photocopy this compact exactly as in @image2 — case, lid, band and the shade inside. Hold THIS compact, not a generic one. She LIP-SYNCS English. 0.0–11.0s CLOSE-UP. She taps the blush from @image2, in its own shade, onto her cheekbone with a brush, smooths it softly, turns her head to show the glow, then POINTS at her cheek. Spoken: "More people need to be talking about this blush. Look at how pretty it is. It blends so well, but do you see this sheen? It's like a glass finish blush." HARD CUT. 11.0–17.0s She holds up the compact from @image2 and gestures to it. Spoken: "It's so buildable and it comes in several beautiful shades." HOLD. AUDIO: clear spoken English, quiet room. No songs. NEGATIVE: overlay type, extra people, hair recast away from @image1, generic compact, outfit different from @image1, watermark on her face.
Photoreal live-action, saturated neon night: magenta and cyan practical light, glossy specular highlights on wet ground, shallow depth of field, rich deep blacks, fine film grain, smooth cinematic camera with a trace of handheld life in it. Not animation — no cel shading, no anime, no cartoon, no illustration, no ink outlines, no flat banded shadows. @image1 is an absolute character lock: face, body, hair, wardrobe, colours and materials are taken from it exactly and must not drift or be reinterpreted at any point. @image2 is the location lock. @image3 is the storyboard — follow it exactly, matching its shot order, framing, camera height and pose. ONE UNBROKEN TAKE. No cuts, no transitions, no jumps. A single camera travelling continuously from the first framing to the last; every change of framing below is achieved by the camera physically moving. They walk toward the lens down the centre of the passage, unhurried, shoulders swinging with the stride, one hand skimming lightly along the wall beside them. The camera pushes in to a medium as they plant both feet, drop their weight onto one hip and throw a quick pose, chin tipping up toward the lens. The camera dips low to the ground and holds on the feet as they bounce through a short footwork step, and fine glitter lifts off the wet surface around the soles. The camera rises and glides backward into a wide as broad wet paint strokes sweep in across the frame from both edges, laying fresh colour over the surfaces behind them. They hold the centre of the wide and dance — arms carving outward, a spin on the ball of one foot, hands snapping back to the lapels — while more strokes bloom outward around them. The camera pushes in to a tight close-up as they turn away and then look back over one shoulder into the lens, mouth curving. Audio: an upbeat electronic track with a steady kick driving the whole take, plus close ambient detail — footsteps landing on wet ground, the rustle of fabric as they move. No voice, no dialogue, no narration. No on-screen text, no captions, no subtitles, no logo, no watermark.
30 seconds, 9:16. Handheld STREET INTERVIEW. Natural, easy-going, slightly messy. Daylight cobblestone, blurred pedestrians. Eye-level. HARD CUT between groups. LANA LOCK: @image1 is Lana — photocopy her FACE. IGNORE the clothes in @image1. OUTFIT: @image2 clothes ONLY — every garment, colour, fabric and accessory exactly as shown in @image2, never changing. IGNORE the studio backdrop. She wears this on the STREET. Lana is IN EVERY SHOT. She holds a BLACK handheld INTERVIEW MICROPHONE. She FACES the girls. The girls look at LANA, not the camera. Two-shots and over-the-shoulder. Never a talking-head into the lens. 0.0–3.0s Lana mic toward a young blonde in denim jacket, white pants, sunglasses on her head. They face each other. Lana: "Can you rate your hair out of ten?" Woman looking at Lana: "Rate my hair? Yeah. Seven." HARD CUT. 3.0–6.0s Lana with mic, two women: brown leather trench + orange pants, beige coat + grey hoodie. They look at Lana. Beige to Lana: "I'm gonna give it a solid six." Brown to Lana: "Like a seven." HARD CUT. 6.0–14.0s Lana with mic. Olive quilted jacket + cap (gestures), beige coat + green scarf. They look at Lana. Olive to Lana: "I'm going to say an eight. Just because it's very low maintenance. I don't really have to do anything to it." Beige to Lana: "For me, it's six." HARD CUT. 14.0–22.0s Lana with mic, blonde curly hair, beige trench, white top, light blue jeans. Woman looks at Lana, smiles, shrugs. Woman to Lana: "I love my hair, but having, like, naturally curly hair, like, sometimes it's just like a wreck. So I feel like I can't give it higher than a six. I'm always dealing with it." HARD CUT. 22.0–30.0s Wider two-shot. Lana still holding the mic, hands a plain WHITE tube of curl cream. Woman smiles and takes it. Lana: "We'd love to give you this curl defining cream, which will really help you maintain your hair and really nourish it. So here you go." Woman: "Thank you." HOLD. AUDIO: street, two voices, mic presence. No songs. NEGATIVE: girls looking into camera, Lana missing from a shot, no microphone, studio backdrop, overlay type, talking-head vlog.
@image1 is an absolute lock for the pieces: it holds two opposing sets, banded upper and lower, and the same pieces, the same number of them, the same colour, finish, thickness, proportion and detailing are taken from it. Nothing is added, removed, restyled or substituted in any shot, and nothing from one set is blended into the other. @image2 is the board: its wood tones, square size and strict alternation are taken from it. Follow @image3 exactly — same beat order, same framing, same hand entries, same piece positions. Look: live-action product film, photographed not rendered. Soft even overhead daylight with no hotspot. Real material behaviour — specular highlights sliding along each curve as it moves, crisp contact shadows sitting directly beneath every piece and travelling with it. Real skin texture, real wood grain. Restrained natural colour, shallow contrast, no grade push. The camera is locked directly overhead, square-on to the board, and holds that exact position and height throughout — it never pans, tilts, pushes, rises or rotates. All movement in frame comes from the hands and the pieces. This is a real game in progress, not product arranged on a checked surface. Every piece stands centred on its own square, one piece per square, never straddling a line and never between squares. The first side's pieces are gathered toward the lower part of the frame and the opposing side's toward the upper part, the two fronts meeting across the middle, about eight to twelve pieces on the board at once with a denser knot near the middle and a few still held back at each side's own end. A short row of already-taken pieces sits just off the board at the right edge of frame and grows as the game goes on. Each move is a single legal-shaped move — straight along a rank, straight along a file, or along a clean diagonal — travelling a whole number of squares and landing centred. Between moves NOTHING on the board shifts except the one piece being played; every other piece holds exactly the square it was on, and the position only ever changes by the move just made. A woman's hands work the board — short neat natural nails, two slim gold rings and a fine gold chain bracelet, the same hands from the first shot to the last. They enter only from the frame edges and withdraw out of frame between moves; they are never shown whole and no face, arm or body ever enters frame. A hand reaches in from the lower edge and advances a first-side piece one square straight forward along its file, landing it centred, then withdraws. A hand reaches in from the upper edge and advances an opposing-side piece two squares straight down its file toward the middle, then withdraws. A hand slides a first-side piece along a clean diagonal, three squares, stopping centred on a square beside an opposing piece, then lifts away. A hand lifts that threatened opposing-side piece straight up off its square and carries it to the row of taken pieces at the right edge of frame, setting it at the end of the row, then withdraws. A hand slides the capturing first-side piece sideways into the square just vacated, fingertips resting a moment where it lands, then withdraws. The last hand draws back out of frame, leaving the position mid-game, the taken row one piece longer than it was at the start. Each beat carries one entry, one move and one withdrawal, and they are spread evenly across the film — none rushed, none held longer than the rest. Audio: soft instrumental music only, low and unhurried, with the faint tap of metal meeting wood as each piece lands. No voiceover, no dialogue, no narration. Do not generate: text, captions, subtitles, on-screen words, logos, watermarks, brand names, price tags, packaging, substituted or restyled product, extra pieces that are not in the reference, sets blended together, pieces between squares, pieces on square lines, two pieces on one square, pieces sliding to random positions, the position rearranging itself between moves, hands held still enough to freeze, faces, bodies, moving camera, tilted or perspective angles, illustration, 3D render, cartoon, cel shading, outlines, flat banded shadows, floating pieces, detached shadows.
@image1 is an absolute character lock for the first dancer and @image2 is an absolute character lock for the second dancer: face, body, hair, costume, materials and tones are taken from them exactly, with no drift or reinterpretation at any point. @image3 is the storyboard and is followed exactly — the same order of moments, the same framing, the same poses, the same camera positions. High-contrast monochrome film look throughout: deep blacks, luminous whites, fine film grain, no colour at any point. One single continuous take from first frame to last, with no cuts, no transitions and no jumps. They dance a hypnotic Argentine tango in a classic close embrace — only two people, two bodies, two heads, four hands, four feet — their steps precise and continuous through a sequence of ochos. The embrace never breaks. Then the surface beneath them unravels into silk-like vapour ribbons and drifting smoke, and they carry a controlled giro through it without pausing. Then they fall through it without interrupting the dance, bodies still locked together, feet still drawing tango lines in midair. Then they land softly and open into a slow volcada, holding the same unbroken embrace and the same fluid rhythm. Head and neck motion stays small and natural throughout. Faces stay oriented toward each other inside the tango hold. No sudden head snaps, no 180-degree head spins, no whipping turns, no broken necks, no impossible rotations. The camera orbits them slowly from a high three-quarter angle, then descends with them in one smooth unbroken move. Camera motion is steady and fluid. No whip pans. No jump cuts. No abrupt orbit reversals. Identity lock for the entire clip: the same two dancers, the same faces, the same clothing, the same hair, the same tango hold, from first frame to last. Sound: a raw bandoneón tango, close footsteps on vapour and then on stone, wind shearing through the fall, distant city hush. Audio fades out as the screen goes black. Negative: no ghosts, no doubles, no extra people, no extra couple, no extra limbs, no extra heads, no extra arms, no transparent clones, no motion-ghost afterimages, no face morph, no sudden head spins, no whip turns, no broken anatomy, no colour, no cuts, no text until the final black frame.
ASSET MAP — the references are bound here and cited below: THE AVATAR — the attached avatar is the sole source of truth for the person: their face shape, eyes, brows, nose, lips, jawline, ears, hairline, hair length, texture and colour, their exact skin tone, their build and body proportions and their apparent age, all exactly as shown. Reproduce them exactly in every frame they appear in. @image1 — the pet food pack. This defines the product: its exact shape, size, material, finish, colours, printed artwork, typography and logo, all exactly as shown. Reproduce it exactly; never redesign it, never re-letter it, never substitute another pack. ATTACHED IMAGES ARE MANDATORY AND FINAL: the attached images outrank every word written below them. Wherever an attached image and this text disagree, follow the image. Never invent a different face, hairline, hair colour, skin tone, body or age for the person, and never invent different packaging. THE CAT — one specific animal, identical in every frame: a dense plush semi-long coat, golden-brown classic tabby with dark striping down the back and flanks, heavy striping across the forehead, a broad white chest bib, white front paws, a long very fluffy tail. A round broad face with a short muzzle and a small pink-brown nose. Small rounded ears set wide apart. Long white whiskers. Large, round, wide-set amber-copper eyes. WORLD — one lived-in living room, photoreal: a plain pale fabric sofa against a plain wall, soft daylight from camera left. Nothing stylised, nothing cinematic-graded, nothing staged; shot as if on a phone. No phone, camera or tripod is visible anywhere in frame. THE PACK STAYS UPRIGHT — the pack@image1 is an opened 10 kg sack of dry kibble. It stands VERTICAL, base down, opening upward, in every frame: never tipped, never tilted, never rotated toward horizontal, never on its side, never above the person's head, never inverted. It stays roughly the height of the person's head and never grows, never comes toward the camera, and never passes in front of the person's face. If keeping the printed front square to the camera would require tilting the bag, do NOT keep the front square — upright always wins. Nothing spills out of it. THE RIGHT HAND PUSHES THE CAT, IT DOES NOT SIGNAL — from 0:14 onward the person's right hand is PHYSICALLY TOUCHING the cat. The heel and palm of their right hand rest flat against the cat's chest and shoulder, fingers pointing upward, and they push the cat gently but firmly back. Their palm faces the CAT; the camera sees the BACK of their hand, never the palm. From the camera's point of view the hand OVERLAPS the cat's body — the cat's chest is directly behind the hand, they are in contact, and there is no gap of empty air between them. This is a person fending off an animal that is leaning on them, not a gesture made to anyone. PROPS AND HANDS — the pack@image1 is held in the person's LEFT hand and their left hand only, for the entire 18 seconds; it never changes hands and the cat never holds it. Their RIGHT hand takes exactly one kibble from the pack in the second beat, and from 0:14 onward touches only the cat. GEOGRAPHY — the person sits at frame left, the cat is on the same sofa cushion at frame right. The pack is carried sideways toward the left edge of frame, away from the cat, never upward. Nobody swaps sides and nobody leaves frame at any point. CAMERA CONTRACT — handheld at chest height. One continuous medium two-shot from 0:00 to 0:06; three instantaneous snap zooms IN at 0:06, 0:07 and 0:08; one instantaneous snap zoom OUT at 0:10; then one continuous medium two-shot from 0:10 to the end. Every snap is an instant change of scale between fixed framings, each held rock steady. Never a smooth continuous zoom, never a crash dolly, never a whip pan, no other cuts anywhere. STARTING STATE at 0:00 — the person is settling back onto the sofa at frame left. The cat is already sitting upright on the same cushion at frame right, haunches down, front legs braced, calm and unbothered, looking off toward camera left. The person holds the pack@image1 upright in their left hand at chest height with its printed front turned square to the lens, unobstructed, in sharp focus and fully legible. 0:00-0:02 — the person settles into the cushion beside the cat and rests the pack@image1 on their knee, its printed front still square to the lens. The cat does not react; it stays calm, blinks once, looks away. Camera holds the medium two-shot, all three clearly readable. 0:02-0:04 — the person reaches into the pack@image1 with the two fingers of their right hand, takes out one single small dry kibble pellet, lifts it to their mouth and puts it in. Completely casual, unhurried, not looking at the pack, eyes on a television off-frame. The cat is still unbothered. 0:04-0:06 — the person chews the kibble, jaw working, an ordinary snacking motion, still staring off-frame, totally untroubled. On the first chew the cat turns its head sharply toward them and freezes. 0:06-0:07 — HARD SNAP ZOOM, step one: the frame jumps instantly from the medium two-shot to a chest-up shot of the cat. The cat's eyes are wide, pupils blown fully open, mouth fallen open, one white front paw lifted and frozen in mid-air at chest height. 0:07-0:08 — HARD SNAP ZOOM, step two: another instantaneous jump to a tight close-up of the cat's head and shoulders. The paw stays up and frozen. The expression holds, absolutely still. 0:08-0:10 — HARD SNAP ZOOM, step three: a final instantaneous jump to an extreme close-up of the cat's amber-copper eyes and open mouth, filling the frame. Hold dead still with the faintest handheld tremble. The expression never resolves, never softens, never blinks. 0:10-0:12 — HARD SNAP ZOOM OUT: the frame jumps instantly back to the medium two-shot, the person at frame left and the cat at frame right, both fully in frame with the pack@image1. The cat goes from frozen with one paw in the air to standing on all four paws on the cushion, and walks across the cushion toward the person. The pack@image1 is back on the person's knee in their left hand, upright. 0:12-0:14 — the cat rises onto its hind legs and plants both front paws flat on the person's left forearm, leaning its weight into them. The person leans their upper body away to frame left and lifts the pack@image1 off their knee to shoulder height, still perfectly vertical, base down, the same size in frame as before. 0:14-0:16 — the person holds the pack@image1 out to the left at shoulder height, near the left edge of frame, upright and unchanged in size, away from the cat. The cat leans and reaches across to frame left after it and swings one white front paw at the bag exactly three times, missing every time. At the same time the person brings their right hand onto the cat: their palm lands flat on the cat's chest and shoulder, in contact with the fur, and pushes the cat back. The hand overlaps the cat's body in frame, the back of the hand toward the camera. 0:16-0:18 — the standoff holds: the pack@image1 stays out to frame left at shoulder height, upright, base down, same size. The person's right palm stays pressed against the cat's chest, still touching, still pushing back. The cat stays up on its hind legs, straining forward against that hand, stretched toward the bag. Both are looking at the bag. The person's face is fully visible, mouth closed, brows raised, refusing. Nothing else moves. The camera holds this exact framing, dead steady, until the last frame. CONTINUITY — the person's face, hair and clothing are identical in every frame they appear in. The cat's coat, markings and eye colour are exactly as described above in every frame from 0:00 to 0:18, and identical across all four changes of framing — the same animal at a different distance, never a different cat. The pack@image1 keeps the same artwork and the same lettering every time it is on screen. Only one person and only one cat exist in this scene. Both the cat's shock and its begging are a real cat's: real feline anatomy, real fur, real pupils, real claws, real balance. Never a cartoon, never humanised eyebrows or hands, never a human mouth, never CGI, never a speaking animal, never walking upright like a person. NEGATIVE — no stop gesture to the camera, no palm facing the lens, no raised hand in empty air, no gap between the hand and the cat. No tilted bag, no tipped bag, no bag overhead, no bag on its side, no bag growing larger, no bag in front of the face, no spilled kibble. No phone, no smartphone, no camera, no tripod anywhere in frame, including the foreground and the bottom edge. The live cat in the room is NOT the cat printed on the pack@image1: never copy the printed kitten's coat, colour, markings, face or expression onto the real animal, and never let the printed artwork replace or blend with it. No on-screen text, no captions, no subtitles, no title cards, no watermark, no logo overlay, no second person, no other animals, no hands entering from off-frame, no slow motion, no fade in or out, no black frames, no split screen. AUDIO: diegetic sound recorded in a quiet room, plus orchestral stings on the snaps and one comedic sting at the end. 0:00-0:02, faint room tone and the soft rustle of the sofa cushion. 0:02-0:04, the dry papery rustle of a hand in the pack. 0:04-0:06, one loud, hard, dry CRUNCH as the kibble breaks between their teeth, close and unmistakable, then a second slower crunch. 0:06, 0:07 and 0:08, a short low orchestral sting on each of the three snap zooms — one deep brass hit each, decaying. 0:10-0:12, the soft thump of the cat's paws crossing the cushion. 0:12-0:14, fabric scuffing as the cat leans on their sleeve, and one short high begging chirp from the cat. 0:14-0:16, claws scratching dry against the foil of the bag on each of the three swipes, and a second short begging chirp. 0:16-0:18, quiet room tone, and over the final one and a half seconds a classic comedic SAD TROMBONE: a single muted trombone playing a slow descending wah-wah-waaah slide, four falling notes, ending flat and deflated exactly on the last frame. No music bed, no song, no other instruments, no dialogue, no words, no narration, no voiceover, no meowing beyond the two chirps, no purring, no laugh track.
30 seconds, 9:16. Cute 3D animation. Static camera per scene. HARD CUT. Lip-sync English. Photocopy the characters from the stills. Do not invent new gadget designs. @image1 = the angry old phone vs the smug smartphone, exactly as drawn in @image1, in the setting shown there. @image2 = the angry old PC vs the smug laptop, exactly as drawn in @image2. @image3 = the angry floppy vs the smug USB stick, exactly as drawn in @image3. @image4 = the angry filament bulb vs the smug LED bulb, exactly as drawn in @image4. 0.0–7.0s @image1. Old phone angry: "I survive wars, you panic at 5% battery." Smartphone smug: "Yeah, but people love touching me, not bricks." Cartoon kick. HARD CUT. 7.0–15.0s @image2. Old PC angry: "I built empires of code before you even had a battery." Laptop smug: "And I run circles around you without needing half the desk." Fist slam, steam puffs. HARD CUT. 15.0–23.0s @image3. Floppy angry: "I stored legends when you did not exist." USB smug: "And needed 10 copies for one song. Move aside, fossil." USB points. Floppy shocked. HARD CUT. 23.0–30.0s @image4. Filament bulb angry: "I sparked the modern world while you sip electricity like a minimalist intern." LED smug: "And I outlast you by decades while you burn out chasing nostalgia." Filament bulb sparks and dims. HOLD. AUDIO: two cartoon voices per scene. No songs. NEGATIVE: overlay type, extra gadgets, humans, recast characters that don't match the stills.
Street ChatRecreate workflow