Storyboard
Feature Snapshot
| Field | Value |
|---|---|
| Feature name | Storyboard |
| Primary goal | Plan video shot-by-shot and generate AI video content |
| Availability | All users |
Key Facts the Model Must Know
- Storyboard is the second mode in the workflow: Pre-Production → Storyboard → Production
- Users plan their video by creating scenes (individual shots)
- Each scene can have multiple takes (AI-generated video variations)
- Two view modes: Table view (spreadsheet) and Card view (visual gallery)
- Timeline at the bottom shows scene sequence and allows reordering
- Video takes are generated through the AI Director chat for each scene
Core Concepts
Scenes
A scene represents one shot in the video sequence.
| Property | Description |
|---|---|
| Name | Identifier (auto-generated, editable) |
| Duration | Length in seconds |
| Narrative | Text description of what happens |
| Visual descriptions | Reference images for the look |
| Custom fields | User-defined metadata |
Takes
A take is an AI-generated video for a scene.
| Status | Meaning |
|---|---|
| Pending | Currently generating |
| Ready | Complete, can view/download |
| Failed | Error occurred |
Common Operations
Creating scenes
- From Mito Director: Drag a script to auto-generate scenes
- From scratch: Click "Add Scene" for blank scenes
- Duplicate: Copy existing scenes for variations
Generating video takes
Video generation uses the AI Director chat interface for each scene:
- Open a scene's video generation dialog
- Choose a generation type:
- Image to video — Animate a still image (provide start frame, optional end frame)
- Video to video — Edit or transform an existing video (provide video source)
- Write a prompt describing what you want
- Select a model and duration
- Optionally adjust aspect ratio and toggle audio generation
- Click generate — cost is shown before submitting
Multi-shot generation
Some models support multi-shot, where you define multiple prompts for different segments:
- Each shot has its own prompt and duration
- Total duration is the sum of all shots
- Each shot can reference different assets with @mentions
- Useful for scenes with distinct movements or transitions
Prompt enhancement
Choose between Fine-tune (enhanced) and Raw (as-written) prompt modes.
Organizing
- Reorder: Drag scenes in the timeline
- View modes: Toggle between table and card view
- Custom columns: Add metadata fields
- Preview: Play through all scenes in sequence
Supported Scenarios
- Guide users through creating their first storyboard
- Help with scene organization and reordering
- Assist with video generation prompts and settings
- Explain generation types (image-to-video, video-to-video)
- Explain multi-shot generation
- Explain view modes and when to use each
- Troubleshoot generation issues
Example Conversations
Creating scenes
User: How do I start my storyboard?
Assistant: In Storyboard mode, you have two options: drag your script onto the Mito Director area to automatically generate scenes, or click "Start from Scratch" to create blank scenes manually. Each scene represents one shot in your video—add a narrative description and reference images, then generate video takes.
Generating video
User: How do I generate a video for my scene?
Assistant: Open the scene and go to the video generation dialog. You'll see the AI Director chat where you can configure your generation. Choose a type (image-to-video or video-to-video), provide source frames, write a prompt, and select a model and duration. Click generate and your take appears once ready. You can generate multiple takes to explore different directions.
Multi-shot
User: What is multi-shot generation?
Assistant: Multi-shot lets you break a single video generation into multiple segments, each with its own prompt and duration. This is great for scenes where the action changes—for example, a 4-second shot where the camera starts wide and then pushes in. Each shot can reference different assets. Enable it in the generation input when using a model that supports it.
Common Issues
| Problem | Solution |
|---|---|
| Generation taking too long | Video generation takes a few minutes; status indicator shows progress |
| Take failed | Try a different prompt, model, or duration |
| Video doesn't match expectations | Refine prompt with more specific details, provide better source frames, try a different model |
| Multi-shot not available | Only some models support multi-shot — select a compatible model |
| Character limit reached | Each model has a prompt character limit — shorten your prompt or use a model with a higher limit |
Related Resources
- User article:
docs/user/storyboard/overview.md - User article:
docs/user/storyboard/guides/generating-videos.md
Update Log
- 2024-12-02 — Initial documentation for Storyboard feature
- 2025-03-09 — Rewritten for AI Director chat-based video generation (gen types, multi-shot, source inputs, audio)