Wan 3.0
Combine reference images, video, and audio into one clip.
alibaba/wan3.0-video/omni-video
Examples
representative generations · starting points to seed the playground
00:00 / 00:00
Use Image 1 as the scene, the character and the camera world: the same woman — same face, same dark bob, same deep cobalt blue shirt — seated at a formica table in the same empty cafeteria with pale butter-yellow tiled walls, and the SAME SMALL GREEN PARAKEET from Image 1. Use Image 2 ONLY as the reference for the six blue birds: five in saturated cobalt and indigo blue, one pale ice-white-blue, with those exact proportions, plumage and wing shapes. Do NOT bring the jungle, the banana leaves, the red flowers or the dark background of Image 2 into the film — the setting stays the pale cafeteria of Image 1 throughout. CONTINUITY RULE, CRITICAL: the small GREEN parakeet from Image 1 is present and clearly visible in ALL THREE shots, including the very last frame of the film. It is bright green against her deep cobalt shirt. It never flies away, never leaves the frame, and is never replaced by or turned into a blue bird. There are seven birds in total by the end: the one green parakeet plus the six blue birds. A 10-second film in THREE distinct shots with hard cuts between them. SHOT 1 (0.0-3.0s) — medium shot, waist up, camera almost still. The small green parakeet already on her wrist walks up her forearm and hops onto her RIGHT SHOULDER, where it settles and grips, turning its head once. She stays composed and glances toward an open window across the room, never at the camera. CUT. SHOT 2 (3.0-6.0s) — the camera looks past her toward a tall sash WINDOW set in the pale butter-yellow tiled wall, its lower sash pushed up and open, daylight and a soft blur of green outside. Her shoulder is in the near foreground at the edge of frame, slightly soft, with the GREEN PARAKEET still perched on it and visible. The six blue birds come IN THROUGH THE OPEN WINDOW: the first drops over the sill and flicks left, two more follow close behind, one hesitates on the wooden sill for a beat then ducks under the sash, the pale ice-white bird comes last and clips the frame with a wingtip. They arrive raggedly, at uneven intervals, real and slightly awkward — NOT a clean slow-motion flock, NOT stock nature footage, NOT a tidy formation. Ordinary shutter speed, wings genuinely beating, a thin net curtain lifting as they pass, shadows skidding across the tiles. CUT. SHOT 3 (6.0-10.0s) — wider shot of her at the table. The GREEN PARAKEET IS STILL ON HER RIGHT SHOULDER, in sharp focus and unmistakably green — this is essential. One by one the six blue birds land ON HER around it: two on her forearm, one on her left shoulder, one settling on her right shoulder BESIDE the green parakeet, one on her raised hand, one on the back of her chair. They fold their wings and settle. She holds very still, chin level, arms steady, calm and entirely deadpan, looking off to the side, completely unsurprised. HOLD THE FINAL FRAME with all seven birds on her — six blue and the one green parakeet at her shoulder — as one last bird tucks its wings. AUDIO (native, diegetic only): flat cafeteria room tone, a faint fluorescent ballast hum, faint street air through the open window, then the clatter of small wings arriving in ones and twos through the opening, a light scuff of claws on the wooden sill, claw-taps on fabric and on the formica table, a handful of short dry chirps overlapping, her quiet steady breathing. No music, no dialogue, no narration. Look: match Image 1 exactly — soft, fully diffused even light, no bare flash, no harsh sun, no sharp-edged shadows, shadows soft but present so texture holds in fabric, skin and feathers, low-to-moderate contrast. Real unretouched skin texture — pores, fine hairs, natural variation. Every feather resolved. Color: soft but vibrant — never oversaturated, never desaturated or washed out (muted is off-brand, soft is not dull). Full color range; saturated reds, oranges, blues, greens welcome and a bold color can lead. Build one complementary pairing — a lighter tone against a deeper saturated one, warm against cool — for depth and to separate subject from background: the pale butter-yellow tiled room against the deep saturated cobalt and indigo of the birds and her shirt, with the green parakeet as the one bright accent. Colors true; no cast, no teal-and-orange or stylized tint. Not over-saturated; contrast controlled, low-to-moderate. Natural restrained grade, slightly creamy, light natural film grain. Camera moves slow and smooth on a tripod or fluid head — no handheld shake, no whip pans, no zoom punches, no lens distortion, no slow motion. The birds are dignified and photographed straight, never cute, comic or twee; surreal but played completely straight, documentary in feel rather than polished stock footage. No text, no lettering, no logos, no watermark.
00:00 / 00:00
A high-production contemporary dance music video, glossy and editorial. Wardrobe and casting as in Image 1: a troupe in long fluid trench coats, muted oat and rust tones. Stage as in Image 2: a grand empty parquet ballroom with tall windows, stacked bentwood chairs against the walls, dust-gold light. Eight dancers, razor-sharp synchronised choreography. 0-3s: the troupe stands in a tight cluster, backs to camera echoing Image 1; on the first beat every head snaps to the lens in unison. 3-7s: they break into a sweeping unison phrase, coats flaring, a slow camera orbit around the group. 7-11s: formation change — dancers fan into a diagonal line; a low tracking shot skims the parquet as feet slide and stamp on the beat. 11-15s: two soloists lift a third; fabric and hair carry real weight and momentum; the camera cranes up. 15-19s: the whole troupe drops to the floor and rolls in a wave, coats fanning like petals, overhead shot. 19-23s: they rise into a spinning canon, one dancer after another, chairs and windows streaking behind in the whip-pans. 23-27s: a sudden freeze — everyone holds a sharp sculptural pose, dust glittering in the light shafts. 27-30s: they surge toward camera in a tight V and stop dead on the final beat, coats settling. Cinematic, shallow depth, smooth gimbal and crane moves, strong rim light through the windows, believable cloth and hair physics. Image 1 governs wardrobe and dancers; Image 2 governs the room.