Seedance 2 Prompt Guide — The Choreographer

REFERENCE ROLES:
@Image1 defines the room, its exact furniture, materials and lighting.
@Audio1 defines the rhythm — one build-up and three clear impact beats.

SEQUENCE — 12 seconds:
0–3s: The finished room from @Image1, calm and photoreal. On the first beat, everything silently unlocks: the entire room lifts a few centimeters and separates vertically into clean floating layers — rug and flooring at the bottom, furniture level above it, then shelving and art, then lighting fixtures, like a living exploded axonometric diagram. Every object holds its exact plan position in the air.
3–8s: The camera performs one smooth orbit around the floating exploded room. Thin light lines briefly connect each piece to its position, like drawing callouts made of light. Gentle rotation of individual pieces, slow and technical, deeply satisfying.
8–12s: On the final beat, all layers slam back down into place in a fast controlled cascade — floor, furniture, decor, lights igniting on impact. One cushion settles. The final frame matches @Image1 exactly. Hold for one second.

CAMERA: One continuous elegant orbit, then a locked hero frame.
NEGATIVE: no collisions, no warped furniture, no bent architectural lines, no people, no text, no logo.

REFERENCE ROLES:
@Image1 defines the exact stone: its veining, color and texture.
@Image2 defines the kitchen where that stone lives.
@Image3 defines the building exterior.

SEQUENCE — one single continuous pull-back, 12 seconds, no cuts:
0–3s: Extreme macro inside the marble veining from @Image1 — golden mineral rivers in white stone, light glinting across the polished surface like a landscape from a plane.
3–6s: The pull-back reveals the vein is the kitchen island countertop from @Image2; a hand places a coffee cup on it as we pass. The kitchen assembles into full view around the island, warm morning light.
6–9s: Continuing back through the apartment — living room sliding past in perfect layered parallax — and out through the balcony glass, revealing the full facade of the building from @Image3, its windows glowing.
9–12s: The pull-back rises into a gentle aerial: the building among its street, trees, evening city lights coming on. The camera settles. One window — ours — glows warmest of all. Hold.

CAMERA: One continuous accelerating-then-settling backward move. Straight architectural lines stay ruler-straight at every scale.
NEGATIVE: no cuts, no warped transitions, no bent buildings, no fisheye, no text, no logo.

REFERENCE ROLES:
@Image1 defines the person — identity, outfit, exact stance. They are the only fixed constant.
@Image2 defines the finished space the site becomes.

SEQUENCE — 15 seconds, one slow orbital camera move around the motionless person:
0–4s: Construction chaos in accelerated time: the person stands perfectly still and sharp at frame center while workers rush around them as elegant motion trails, scaffolding assembles and vanishes, dust clouds bloom and clear. Daylight arcs across the space in seconds — day, night, day.
4–9s: Walls rise plank by plank in stop-motion rhythm, screed pours and dries in a breath, windows glaze themselves, the light arcs keep spinning. The person slowly turns their head, calmly surveying the storm of progress. Seasons flick past the window openings — rain, sun, dusk.
9–13s: The blur decelerates: flooring sweeps in, paint rolls up the walls, furniture assembles, the motion trails of workers thin out and disappear one by one. Warm lighting fades up.
13–15s: Time lands at normal speed. The finished interior from @Image2 stands quiet and complete. The person exhales, finally moves — one step forward — and smiles. Hold.

NEGATIVE: no identity drift or motion blur on the person, no warped walls, no bent lines, no text, no logo.

REFERENCE ROLES:
@Image1 defines the room, furniture set and final arrangement.
@Image2 defines the person's exact identity, hair and outfit — keep identical throughout.
@Audio1 defines the timing: long build-up, one hard drop.

SEQUENCE — 12 seconds:
0–7s (build-up): Inside the room from @Image1, gravity is off. The sofa drifts slowly at shoulder height, armchairs rotate lazily, books and ceramics float in constellations, a pendant lamp swings free of its cable, cushions tumble in slow motion. Dust motes sparkle in window light. The person from @Image2 walks calmly through this floating world, completely unbothered, gently pushing a drifting cushion aside like a swimmer. The camera tracks them in one smooth lateral move.
7–8s: They stop at the room's center, look at the camera, and snap their fingers.
8–12s (drop): On the drop, everything falls at once — every piece landing in its exact designed position with soft weighty impacts, rug flattening, books aligning on shelves, lamp catching its cable and igniting. The final frame matches @Image1 perfectly, with the person standing composed at its center. Hold.

NEGATIVE: no identity drift, no collisions, no broken objects, no warped furniture, no bent lines, no text, no logo.

REFERENCE ROLES:
@Image1 defines the space and every final material: tiles, wood, metal, fabric.
@Audio1 drives the chain reaction — every impact lands on a beat.

SEQUENCE — 12 seconds, one continuous chain reaction across the empty room shell:
0–3s: Macro shot: a single ceramic tile tips forward and triggers a domino run of tiles snapping flat across the floor in a curving path, each click on a beat. The camera races low alongside the domino line.
3–6s: The last tile flips a row of oak planks upright; they fall in sequence like piano keys, laying the wooden floor in a spreading wave. The wave slaps the wall and paint rolls upward in a clean band.
6–9s: The paint wave knocks brass handles and hooks off a rail; they fly in precise arcs and click onto cabinets and doors one by one. A rolled rug is bumped, unrolls at speed, and its edge flicks a switch.
9–12s: The switch ignites all lighting in a cascade while a fabric wave — curtains, cushions, throws — settles softly over everything. The camera pulls up and back to reveal the completed room matching @Image1. Everything stills. Hold.

CAMERA: One continuous flowing move chasing the chain reaction, ending in a raised wide hero frame.
NEGATIVE: no chaos or breakage, no warped geometry, no bent lines, no people, no text, no logo.

REFERENCE ROLES:
@Image1 defines the person's exact identity and outfit — keep identical throughout.
@Image2 defines the room and the final fabrics: sofa upholstery, curtains, bedding.
@Audio1 defines the rhythm — three beats, three transformations.

SEQUENCE — 12 seconds:
0–3s: The person from @Image1 stands in the unfurnished room holding a single linen swatch. They release it into a beam of window light. Audio: rising wind under the music.
3–6s: The swatch multiplies — dozens of fabrics in the project's palette spiral out of it, orbiting the person like a controlled silk storm. Cloth ripples in slow motion, catching light; the person turns calmly inside the vortex, arms slightly raised like a conductor.
6–10s: On each beat, part of the storm dives and becomes real: BEAT 1 — fabric wraps a bare frame and tightens into the upholstered sofa. BEAT 2 — two long streams rise and hang themselves as curtains, swaying once. BEAT 3 — the last wave rolls across the bed and settles as perfect layered bedding.
10–12s: The air clears. The finished room matches @Image2. The person catches one final swatch out of the air and smiles at the camera. Hold.

NEGATIVE: no identity drift, no fabric tearing, no chaotic tangles, no warped furniture, no bent lines, no text, no logo.

REFERENCE ROLES:
@Image1 defines the phone screen: the real Instagram post showing the project photo.
@Image2 defines the actual room inside that photo.

SEQUENCE — 12 seconds, one continuous camera move:
0–2s: Over-the-shoulder macro: a thumb scrolls a feed on a phone in a cozy evening cafe. The scroll stops on the project post from @Image1. The thumb hesitates. The camera pushes toward the screen.
2–4s: The push accelerates through the glass — the flat photo gains depth, parallax opening around the lens as the image becomes a real three-dimensional space. Screen glow gives way to true window daylight.
4–9s: We are inside the real room from @Image2. The camera glides through it in one graceful arc — past the textured stone wall, low over the sofa, brushing the linen curtains as they breathe in a soft draft, catching dust motes in the light. Everything is alive: a candle flame sways, reflections track across brass.
9–12s: The camera keeps moving toward a glowing rectangle ahead — and pulls back out through another phone screen, revealing a different person holding it, eyes wide, in a bright design studio. She looks up from the phone to camera and smiles. Hold.

NEGATIVE: no warped screen transition, no UI gibberish text, no bent room lines, no fisheye, no logo.

REFERENCE ROLES:
@Image1 defines the exact furniture piece — its material, joinery, color and proportions.
@Image2 defines the corner of the room where it will live.

SEQUENCE — 10 seconds, macro-to-wide, physics-perfect:
0–2s: A flat cardboard box lies in the middle of the room from @Image2, morning light raking across it. The straps snap open by themselves; the box unfolds flat like a blooming flower. Audio: crisp cardboard sounds, a soft click.
2–7s: The furniture piece from @Image1 assembles itself in one continuous origami choreography: legs pivot and lock with satisfying clicks, the frame rises and squares itself, joints slide together with millimeter precision, the seat drops in and settles, wood grain aligning perfectly at every joint. The camera circles low and close, celebrating each connection in shallow depth of field. Every motion obeys real physics — weight, momentum, tiny settling bounces.
7–10s: Fully assembled, the piece slides gently into its exact place in the corner, the flattened box folds itself small and tucks away, and warm light lands on the final composition matching @Image2. One slow push-in. Hold.

NEGATIVE: no floating parts, no impossible bends in solid wood, no warped joinery, no people, no text, no logo.

REFERENCE ROLES:
@Image1 defines the living room and the person — keep both identical in architecture and identity throughout. Only light, styling details and atmosphere change.
@Audio1 provides three sharp accents — each one is a clap.

SEQUENCE — 12 seconds, locked-off wide camera, the person centered:
0–2s: The person stands relaxed in the finished living room, neutral daylight. They look at the camera, raise their hands — and CLAP.
2–5s (SCENE 1 — MORNING): Instantly: bright soft morning light floods in, curtains open, a breakfast tray and newspaper appear on the coffee table, a throw blanket tossed casually, warm family energy. The person gestures around proudly. CLAP.
5–8s (SCENE 2 — DINNER PARTY): Instant switch: golden evening, pendants dimmed low, candles lit on a styled bar setup, wine glasses catching light, jazz-warm atmosphere, cushions rearranged sharp. CLAP.
8–11s (SCENE 3 — MOVIE NIGHT): Instant switch: deep blue night, hidden cove lighting glowing softly, TV wall alive with color reflecting on the ceiling, popcorn bowl and soft blankets, cozy cinematic dimness.
11–12s: The person shrugs at the camera with a smile: "your room, every hour." Hold the final frame.

NEGATIVE: no architecture or furniture layout changes, no identity drift, no flicker between scenes, no text, no logo.

REFERENCE ROLES:
@Image1 defines the room's geometry and furniture in a pure white state.
@Image2 defines the exact color and material palette — use only these tones.
@Audio1 defines three deep beats — one per color wave.

SEQUENCE — 12 seconds:
0–2s: The all-white room from @Image1, silent and gallery-like. A single drop of warm clay-toned liquid falls in slow motion onto the center of the floor and lands exactly on the first beat.
2–5s (WAVE 1 — WALLS): From the drop, liquid color surges outward and climbs the walls in slow elegant waves — the palette's wall tone from @Image2 flowing like paint underwater, stopping in razor-sharp lines exactly where the design says it should.
5–8s (WAVE 2 — WOOD & JOINERY): On the second beat, a deeper wave of walnut tone flows across cabinetry and shelving, grain patterns blooming through the liquid as it thins; brass details condense out of golden droplets.
8–11s (WAVE 3 — TEXTILE): On the third beat, a soft matte wave rolls over sofa, curtains and rug, soaking them in the palette's fabric tones; the liquid sheen evaporates upward as fine mist, leaving perfect matte finishes.
11–12s: The room stands complete, dry and exact. Slow push-in to the hero frame. Hold.

NEGATIVE: no color bleeding outside designed zones, no glossy wet leftovers, no warped geometry, no bent lines, no people, no text, no logo.