> the stats show intersection tests against a bounding object and a bounding
> box. Why the bbox when the shape is already bounded manually?

I'm willing to bet that the ray first tests to see if it hits the bounding
box of the bounding object. If it does, then it tests to see if it hits the
bounding object. If it does, then it tests to see if it hits the object
itself.

This probably speeds it up a bit, since you could technically specify an
isosurface or other slow-to-render object as the bounding object if you
wanted, and testing the bounding box of *that* object may speed up the
render.

Just a guess.

- Slime
[ http://www.slimeland.com/ ]
[ http://www.slimeland.com/images/ ]

Post a reply to this message