User Guide > Rendering > Mantra   

Deep Rasters


When rendering an image using Mantra, you can add extra channels that create Deep rasters. The Houdini .pic format supports these extra channels which can be used to store channels for variables such as Surface Position or Surface Normals. These channels can later be used to achieve compositing effects.


Setting up a Deep Raster Rendering

Set up a Render Output Operator for rendering out a sequence of images. Click on the Deep Raster tab and from the VEX Variable arrow [>] choose a new channel type. Here are a few of your options:

Surface Depth [Pz] - This provides the zdepth information as measured from the camera. This can be used for zdepth compositing or adding effects such as depth of field or fog.

Surface Normal [N] - This provides information about the surface normals. This can be used to relight a scene in compositing.

Surface Position [P] - This contains information about the xyz positions of points in space. This can also be used for relighting.

Once you have these set then rendering an image will produce extra channels that can be used by the compositor.