Locked Rooftop Staff Duel from Storyboard Poses

An image-to-video prompt that turns a nine-panel storyboard contact sheet into one unbroken locked shot of a staff-training duel on a city rooftop at dusk. The camera stays bolted down while the action passes through nine ordered key poses, with strict continuity rules for props, positions and character states.

This clip could not be loaded — it may have been removed from X.
Borderless 🌐 @Borderless_DeFi ·
View on X ↗

Prompt

Use the first reference image (9 panel storyboard contact sheet whose cells are labelled TAKE 01 to TAKE 09) as ordered key poses TAKE 01-TAKE 09 for ONE unbroken LOCKED shot: the camera never moves, the action passes through TAKE 01 to TAKE 09 inside a fixed frame, without cuts or holds; completely ignore its line-sketch style, panel frames, gutters and any text.
LOOK & WORLD: Modern 2D cel-shaded anime with painted backgrounds and clean confident linework. On a city rooftop at dusk, two fighters in simple training clothes trade staff work inside a chalk circle. Warm low dusk light, long shadows on concrete, cool blue city haze behind. One fixed 24mm lens, deep focus, cinematic. The lens never changes.
CONTINUITY (binding, not flavour): Locked frontal wide at waist height, unchanged from TAKE 01 to TAKE 09. Stage anchors hold their positions throughout: rooftop water tank screen-left, chalk circle center, city skyline screen-right, crate at right. The chalk circle is unbroken through TAKE 04; scuff marks start at TAKE 05 and persist. One staff drops at TAKE 06 and stays on the ground for the rest of the take. The screen-left fighter is upright through TAKE 07 and grounded from TAKE 08 on. Stage anchors hold their screen positions throughout, the camera height never changes, and every state pinned to a TAKE number is true from that TAKE onward and false before it. Never invent a shot that the sheet does not have.
ACTION & CAMERA: the camera is bolted down. No cuts, no zoom, no pan, no tilt, no tracking, no shake, no reframing. Only the action moves, entering and leaving a fixed frame.
(TAKE 01) locked, no movement, Facing off: both fighters stand at opposite edges of the chalk circle, staffs low.
(TAKE 02) locked, no movement, First approach: the screen-left fighter steps in, staff rising to guard.
(TAKE 03) locked, no movement, First clash: the staffs meet high in the center of the frame.
(TAKE 04) locked, no movement, Bind: bodies close, staffs crossed, both leaning into each other.
(TAKE 05) locked, no movement, Break: the screen-right fighter pivots away, first scuffs on the circle.
(TAKE 06) locked, no movement, Disarm: the screen-right fighter staff spins out to frame-right and lands on the ground.
(TAKE 07) locked, no movement, Empty hands: the screen-right fighter raises open palms, the other still guarding.
(TAKE 08) locked, no movement, Takedown: the screen-left fighter is swept and lands flat inside the circle.
(TAKE 09) locked, no movement, Hand up
? How to use this prompt ▾
  1. Copy the prompt using the green "Copy prompt" button above.
  2. Open your model — launch Seedance 2.0 on its official interface.
  3. Paste & generate — keep the verbatim prompt text for best adherence. Set aspect ratio to 9:16 and clip duration to 15s.

What makes this prompt work

Every insight below is verified against benchmark generation parameters, not asserted. How our prompt engineering benchmarks work →

CAMERA
"zoom" is a core cinematic camera technique for establishing spatial immersion and subject framing. Also present: "pan", "tilt", "tracking".
LIGHTING
"warm tone, cool tone" establishes luminance depth, volumetric shadow falloff, and tonal contrast across the generation.
MOOD
"cinematic, dramatic" — Mood and aesthetic tone are explicitly directed to maintain emotional consistency throughout the scene.
DURATION
At 15s, this clip is optimized for high temporal coherence and fluid, realistic motion physics.
MODEL
Seedance 2.0 (ByteDance) provides the underlying diffusion and motion architecture. Camera, lighting and duration cues transfer across models — what differs is how literally each model interprets motion.
PROMPT LANGUAGE
The prompt is written in English. This library keeps every prompt in its original language rather than translating it — the exact words are what was conditioned during generation.
Seedance 2.5 BytePlus Cloud Partner

Tell bigger stories with Seedance 2.5

Create AI videos up to 30 seconds long with richer image, video, and audio references. Ideal for complex ideas that need consistency and control.

Explore on BytePlus Official ByteDance Enterprise Cloud