Mastering R2V References in Seedance 2.5: Your Secret Weapon for Motion Control
R2V (Reference-to-Video) is the most powerful and least understood feature in Seedance 2.5. Here's how to use green screen and white model references for production-level control.
R2V is the feature that separates Seedance 2.5 amateurs from pros. It lets you upload movement references — green screen footage, white model animations, or even simple phone recordings — to guide how characters move in your AI-generated video.
Think of it as motion capture without the expensive hardware. You show Seedance 2.5 how movement should look, and it applies that motion pattern to your generated characters while maintaining visual consistency with your other references.
This guide covers everything from basic R2V usage to advanced multi-reference workflows.
R2V Basics: How Reference-to-Video Works
At its simplest, R2V lets you upload a video clip that guides character movement. The model extracts motion patterns from your reference and applies them to the generated character. The result: movement that looks intentional and natural, not random.
You can use several types of R2V references: green screen footage (best for isolating character movement), white model/blockout animations (for 3D pre-visualization), and even regular video clips (the model can extract motion even without a clean background).
The key insight: R2V controls MOTION, not appearance. Your character reference images control what the character looks like. The R2V reference controls how they move. These are separate control layers that work together.
Creating Effective R2V References
You don't need professional equipment. I've created effective R2V references with my phone camera. The key is clear, intentional movement. Record yourself or a subject performing the actions you want the AI character to perform.
For green screen references: even a simple colored wall works. The model just needs to separate the subject from the background. Perfect green screen is nice but not required.
For white model references: if you work with 3D tools, create a simple blockout animation showing the spatial layout and movement paths. Seedance 2.5 can interpret these and apply them to the final render.
Tips for good R2V references: keep movements deliberate and clear, avoid rapid or erratic motions, ensure the subject is fully visible in frame, and match the general spatial scale to your intended output.
Frequently Asked Questions
What is R2V in Seedance 2.5?
R2V (Reference-to-Video) lets you upload movement reference footage to guide character animation in AI-generated videos. Upload green screen footage, white model animations, or regular video clips to control how characters move.
Do I need professional equipment for R2V references?
No. A phone camera recording works fine. The key is clear, intentional movement. Green screen helps but isn't required — the model can extract motion from regular video too.