This page gives information on the Cryptomatte Render Element.
Cryptomatte is a method developed by Psyop for efficiently encoding accurate mattes. It uses typically three to five render elements (which are automatically generated) with OpenEXR files, which removes the hassle of creating dozens, or hundreds, of s. Using the Cryptomatte plugin for NUKE or Fusion, mattes can be extracted through directly picking objects or entering object names.
Compared to the, Cryptomatte offers the following:
- Does not require a setup with object IDs, etc.
- Only requires a fixed number of additional render elements, typically four.
Starting with V-Ray 3.60.04, single-channel EXR files using DWA compression are supported. V-Ray 3.60.03 and earlier versions only support Multichannel EXR files.
Currently, the Cryptomatte Render Element requires the Bucket Image Sampler type and outputs set to OpenEXR files.
UI Path: ||Render Settings window|| > Render Elements tab > Cryptomatte
The parameters for this render element appear in the Attribute Editor under Extra VRay Attributes.
Enabled – When enabled, the render element appears in the V-Ray frame buffer.
Filename suffix – The text added to the end of the rendered file, when saved as a separate file (e.g. myrender.cryptomatte.exr).
ID type – Specifies how the ID mattes are determined.
Node name – Creates mattes by node names.
Material name – Creates mattes based on the materials in the scene.
Node name with hierarchy – Creates mattes by node names and takes the node hierarchy into account for linked or grouped objects.
V-Ray user attribute – Creates mattes with objects that include the V-Ray User Attribute specified in the User attribute name field.
Asset (reference scene) name – Creates mattes based on Maya References made within the scene. Note: This does not create mattes based on Maya Assets, only file references.
User attribute name – Specifies a user-defined attribute for creating ID mattes when ID type is set to V-Ray user attribute.
Num levels – Specifies the number of levels for the render element, which determines the number of objects that can be distinguished per pixel. If multiple objects intersect at a single pixel, higher-order elements will contain non-zero results. The default value of 6 is normally enough for most cases.