Seedance & Higgsfield Cinematic Shot Direction
Seedance/Higgsfield video prompts with a locked look, one motivated camera move per shot, reference-image roles, and a credit plan.
Before & after
Same prompt, same agent — the only difference is the skill.
You could describe the scene you want to create and include some visual details. For example, you might mention what's happening, what the setting looks like, maybe some colors or mood. You could try adding some camera direction like panning or zooming. Include any characters or key elements. It helps to be descriptive, but also concise so the AI can understand. You might want to think about what the final output should feel like and describe that too.
PROMPT TEMPLATE OPEN: Static wide on rain-soaked city rooftop at dusk. → Camera: Alexa Mini 4K, locked tripod, F2.8 → Lens: Zeiss Master Prime 35mm, slight vignette → Grade: Kodachrome 40 emulation, crushed blacks, +15 magenta → Runtime: 3.5s MID: Handheld following lead character crossing frame. → Camera: handheld operator-shake (0.8 intensity, 12Hz) → Lens: Zeiss Master Prime 24mm, 1.5x digital zoom in-motion → Grade: match open, add 2-stop knee lift → Runtime: 4.2s → Dutch angle: +12 degrees CLOSE: Product reveal, locked, color pop (product glows). → Camera: Macro setup, locked, F8 deep focus → Lens: Zeiss Mpe 65mm macro → Grade: desaturate BG to B&W, product full color → Runtime: 2.3s REFERENCES → Image 1: opening mood (position 1 of 3) → Image 2: character moment (position 2 of 3) → Image 3: product hero (position 3 of 3) TOTAL ESTIMATED RUNTIME: 10.0s CREDIT COST ESTIMATE: 270-310 credits
About this skill
Your AI video shots each come out looking a little different. The camera drifts around with no purpose. Reference images swap faces and outfits between shots. Words like "cinematic, 4K, masterpiece" just give you a generic glossy look. This skill writes Seedance and Higgsfield prompts the way a cinematographer preps a shoot. It locks one look, gives every reference image one job, and gives each shot one camera move with a reason behind it.
What it does
- Builds a look block you paste unchanged into every clip: lens, film stock, grade, and light source. This is the fix for shots that drift from each other.
- Describes each lens by what you see on screen (focus, background blur, compression), not just a focal length
- Gives each shot one camera move with speed, start, and end. Handheld shake is set to one of four levels, and Dutch angles get a specific degree
- Gives each @image reference one role and orders them so up to nine refs land in the right places
- Includes five look presets to start from: parking-lot noir, K-pop MV, documentary, editorial, action
- Catches problems before you spend credits: too many shots for the clip length, too many references, real-person likeness, and text in frame
What you get
Prompts you can paste straight in, with timecoded shots, a reference role map, and a generation plan. The plan estimates retries, gives a credit formula, and names the one line to change if a take goes wrong.
Best for
AI video creators, music video and short-film makers, and agencies producing multi-shot scenes in Seedance or Higgsfield who need shots that match each other.
SKILL.md preview
---
name: seedance-higgsfield-cinematic-shot-direction
… (sign up to view the full skill)More creator skills
View all Creator skills →Explainer B-Roll Prompt Pack for AI Video
Turns a VO transcript into a style-locked AI video prompt pack: coverage map, P1-P3 clips timed to the VO, exact data routed to code.
Seedance Scene Prompts: Action, Dialogue and Blockouts
Seedance prompts for scenes with people: weighty action, living performances, timed dialogue, multi-clip continuity, and blockout v2v.
Faceless 3D Animated Shorts Pipeline
Concept, timed narration, style bible, and per-scene image, motion, VO, and SFX prompts for faceless 3D shorts that stay consistent.
Seedance 2.0 Video Prompt Writer
Turns an idea into a timecoded multi-shot Seedance 2.0 prompt with camera moves, continuity, reference roles, and synced sound cues.