Method
Selects the motion estimation algorithm.
Motion-vector analysis between frames
Analyzes pixel movement over time and generates motion-vector information for retiming, interpolation, vector blur and deformation workflows.
OpticalFlow estimates where image details move between neighboring frames. The resulting vector field can feed temporal or vector-based tools rather than forcing those tools to infer motion from color alone.
Overview
OpticalFlow estimates where image details move between neighboring frames. The resulting vector field can feed temporal or vector-based tools rather than forcing those tools to infer motion from color alone.
Think of OpticalFlow as converting visible motion into arrows: each region gets a direction and magnitude that downstream nodes can use.
Flow anatomy
The port structure shows how this node fits into a Fusion graph.
Source image for the operation.
The processed result from this node.
Inspector
Start with these controls before reaching for more complicated workarounds.
Selects the motion estimation algorithm.
Uses Classic for method.
Uses Advanced for method.
Controls repeated warping passes during vector estimation.
Controls refinement iterations.
Balances detailed local motion against a smoother field.
Speeds analysis by estimating at reduced resolution.
Stores vector data in layers for compatible downstream tools.
Enables more detailed quality controls.
Standard Fusion node settings for processing, masks, blending and animation.
Practical setup
Use a representative motion-heavy section.
Choose Classic for speed or Advanced for difficult shots.
Keep iteration/warp values modest at first.
Inspect the field with VectorVisualization.
Feed the vectors into the required temporal or distortion node.
Examples
Generate motion vectors and use them for directional blur.
Analyze motion, inspect it, then build a controlled vector warp.
Troubleshooting
Production notes
Related nodes