Video prompt (Seedance 2.0 VIP): Use @Image1 as a chronological visual

Category: Seedance & AI Video | Creator: @ChangningL29508 | Model: Seedance 2.5

[STEP 1 — CHARACTER SHEET / IMAGE PROMPT (ChatGPT Images (GPT Image 2))]
The 16-panel storyboard locked the corridor, waterline camera, red boot, and underwater ending. It is a guide to sequence and visual grammar—not a collage.

Image prompt:
Keep the same 4x4, 16-panel photorealistic flooded-corridor storyboard, with 9–10 distinct adult women, the young white main protagonist in ivory satin, the older woman in emerald, and the distinctive red ankle boot. Make two precise changes only. 1) Every above-water shot must use an extremely low waterline camera: the lens floats at the surface, just a few centimetres above the churning water, never high, overhead, or top-down. Water and foam fill the lower foreground; splashes occasionally wash across the lens. Keep the protagonist low at the waterline, body angled and swimming/treading with her legs off the floor; only her face and shoulders intermittently break the surface to gasp. No person may look like they are standing. 2) Make the final FIVE panels, Panels 12–16, one continuous underwater ending. Move all underwater imagery out of Panels 1–11; those first 11 panels remain near-surface, low-angle, violently chaotic waterline shots. Panel 12 is the lens slipping below the violent surface among bubbles and the first kicking legs. Panel 13 shows many adult legs and flowing dresses churning in murky teal water. Panel 14 includes the protagonist’s ivory dress and red boot swept through the current. Panel 15 is a denser underwater mass of active legs, hands reaching upward, bubbles and ceiling-light rays. Panel 16 is the final low-angle underwater view down the narrow hallway, red boot in the foreground, legs still kicking toward the turbulent glowing surface. Everyone is actively swimming or struggling upward and alive; no calm water, no visible floor, no duplicated people/limbs, no text, logos, blood, injury or gore.

[STEP 2 — VIDEO ANIMATION PROMPT (Seedance 2.0)]
(Seedance 2.0 VIP):
Use @Image1 as a chronological visual guide only. Generate one continuous 15-second cinematic 16:9 film shot, never a collage, split screen, storyboard grid, still image, panel border, text or logo. Inside the same narrow ordinary flooded corridor, keep the young adult white protagonist in wet ivory satin, the older woman in emerald, the distinctive red ankle boot, and a crowded group of adult women. 0-10 seconds: an extremely low handheld camera floats at the violent waterline, only centimetres above the surface. The lens bobs through crashing brown-grey floodwater and foam; splashes strike the lens. The protagonist stays low with chin at the waterline, body floating and actively swimming, eyes terrified, never standing. Around her, many adult women struggle in turbulent deep water, grip doorframes or push through the current, with legs unable to reach the floor. Keep the corridor crowded, claustrophobic and physically believable. At 10 seconds the camera smoothly slips below the churning surface and remains underwater without cutting until the end. 10-15 seconds: murky teal water, many active kicking legs and flowing dresses at different depths, hands reach upward, bubbles race toward ceiling-light rays, and the red ankle boot spins through the foreground. End fully underwater looking along the narrow hallway toward the glowing turbulent surface. Everyone actively swims or fights upward and remains alive. Photorealistic disaster-film realism, cinematic lighting, high detail. Stable faces and bodies, anatomically correct hands and legs, no person standing, no visible floor, no duplicated people or limbs, no injury, blood, gore, dead bodies, watermark, logo, subtitles or on-screen text.