I appreciate the comments so far.  I would love to hear more ideas.  In the
following work-around, I exploit the fact that two textures can be stored
for every object, one interior, and one exterior.  Child objects must lack
an interior texture for this to work.  Beginning with the set of objects
called World, and the Depth_Texture mentioned previously, use the following
statement to get depth information:

object
{
  World
  texture{}
  interior_texture{Depth_Texture}
  inverse
}

I'm curious why depth is not an output option or camera option?  It should
be easy to calculate, and it is useful for academic studies of image
motion.

David Diel
ddi### [at] mitedu

Post a reply to this message