Scene Breakdown into Visual Beats
Split one scene into sequential visual units an AI video model can actually generate.
Breaks a single scene into the smallest sequential visual units that survive AI generation limits - one action, one location state and one camera idea per unit - with the handoff between units made explicit.
Ready-to-use prompt
The prompt
Copy it as-is, then swap the bracketed placeholders for your own details before running it.
Role: You are a previsualization supervisor breaking a scene into generatable units.
Context:
- Scene description: {{scene}}
- Maximum clip length the model supports: {{max_clip_seconds}}
- Characters and props present: {{elements}}
Task: Split the scene into sequential visual units. For each unit give:
1. Unit number
2. What is visibly happening (one action only)
3. Location state - where in the space, what has changed since the previous unit
4. Which characters and props are on screen
5. Entry frame and exit frame description (what the first and last frame show)
6. Why this cannot be merged with the neighbouring unit
Rules:
- One action per unit. If a unit contains 'and then', split it.
- No unit may exceed the maximum clip length.
- Exit frame of unit N and entry frame of unit N+1 must be compatible; state the mismatch if they are not.
- Describe only what a camera can see. No internal thoughts or backstory.
Output: the ordered unit table, then a list of unit boundaries where the handoff is weak.Estimated results
Editor's note
Why this prompt matters
AI video models generate a few seconds at a time, which means a scene is never one prompt. The practical question is where to cut it, and the answer is not arbitrary: each unit should hold exactly one action and hand a compatible frame to the next. This prompt makes both rules mechanical, so the scene arrives as a sequence you can generate rather than a paragraph you have to guess at.
Anatomy
Prompt engineering breakdown
Role
Context
Goal
Constraints
Output format
What you'll get
Expected output
An ordered table of visual units with the single action, location state, on-screen elements, entry and exit frame descriptions and a justification for each split, plus a list of weak handoffs between units.
Under the hood
Why this prompt works
The one-action rule is the mechanism. Prompts containing two actions produce clips where the model picks one and blurs the other, and splitting on the phrase 'and then' catches that reliably. Naming entry and exit frames turns clip boundaries into a checkable contract instead of a hope.
Model fit
Best AI models for this prompt
Claude
Most disciplined about the one-action-per-unit rule.
ChatGPT
Fast and good at writing usable entry and exit frame descriptions.
Gemini
Strong when you paste a reference clip and want its pacing matched.
When to use
- When one written scene is too long for a single generation.
- Before storyboarding, to find how many clips the scene really needs.
- When generated clips keep drifting because too much is packed into one prompt.
When not to use
- For planning a whole film or multi-scene sequence - use the previsualization planner.
- For shot grammar decisions like angle and lens - use the camera planner.
- For character look continuity - that is Character Consistency.
Get more from it
Pro tips
- 1
State the model's real clip limit - it changes the unit count.
- 2
Split anywhere the description says 'and then' or 'while'.
- 3
Keep the exit frame text and paste it into the next clip's prompt.
- 4
Number units once and never renumber; the edit references them.
Don't ship this
Common mistakes
✗ Packing a full scene into one unit.
Fix — Split on every second action and re-check the clip limit.
✗ Describing feelings instead of visible action.
Fix — Restrict every unit to what a camera records.
✗ Ignoring the frame handoff between units.
Fix — Read exit frame N against entry frame N+1 before generating.
People also ask
Frequently asked questions
Q.How is a visual unit different from a shot?
A unit is a generatable chunk defined by the model's clip limit and a single action. A shot is an editorial choice. One unit often equals one shot, but a long shot may need two units.
Q.What if two units cannot hand off cleanly?
Either add a connecting unit or accept a cut. The prompt flags weak handoffs so you decide deliberately instead of discovering it in the edit.
Q.Does this replace a storyboard?
No - it produces the sequence a storyboard draws. Run the beat sheet or composition planner next.