I would like to take the results of a QCRenderer and feed them to a
different pipeline as a Core Image. I was thinking about the
following approach:
- attach a QCRenderer to a Pbuffer-based CGLContext
- create a CVOpenGLBuffer and attach the CGLContext to it (not
clear from CV docs as to what happens when you attach though)
- create the CIImage using the constructor that takes a CVImageBuffer
Sounds a good approach - you can actually render directly the
composition in a CVOpenGLBuffer, which is a subclass of
CVImageBuffer, using a QCRenderer. See /Developer/Examples/Quartz
Composer/Performer.
