Honestly, I didn't think of using an isosurface.  Even though the
functions are simpler for a POV programmer to write down, isosurfaces
are inherently slower to render than the "more efficiently coded" poly
objects .

	I mean there is a large amount of overhead in just evaluating the math
expressions using an interpreter, not to mention poor renderings due to
max_gradient issues. (Check out the 1/r factor, which will cause the
gradient to blow up close to the z-axis).

	But in this case I think you're totally right.  Max gradient isn't an
issue and the increased time probably isn't a factor.

	Ben
	<><

Post a reply to this message