The velocity pass stores object movements as motion vectors. The main porpose for this is to allow you to add motion blur to your scene after rendering. You can also use this data to generate various effects based on object movement, for example make objects glow brighter when they are moving faster.
You can choose between 3 different types of velocity maps:
This generates 2 velocity vectors in camera space. This is the most accurate type to use for motion blur during compositing. It considers all object movement, camera movement and movement towards and from the camera while generating the final velocity data.
While the adjusted camera calculates movements from and towards the camera into its x and y directions this mode keeps a separate third channel for depth movement.
This mode does not take relative movement into consideration and the resulting vectors reflect how fast the corresponding objects are really moving regardless of the current view. This can be used for effects on objects that depend on their absolute speed.
Clamp Color Space
By default Spotmask Velocity stores vector data that goes far beyond the normal range expected for colors. Viewing the rendered file in an image viewer will often not show a recognizable image.
This behavior is fine for compositing but some software does not support the rendered file at all if its data exceeds the normal color range. By using the clamping option you can have all vector data translated to the visible color space that maps all channels into the range between 0.0 and 1.0. Only use this option when you need it for compatibility reasons.
This number defines the maximum velocity which gets mapped to the full intensity of the corresponding channel. If this number is smaller than the actual velocity you encounter in the scene your results will be cut off. If this number is significantly bigger than your actual velocity the results will only use a very small section of the available color space.
This specifies how the velocity vectors are mapped to the 3 color channels of the output image.
The default corresponds to the most widely used configuration and most software that utilizes velocity maps also includes options to swap channels on demand, so you’ll rarely need to adjust these options at all.