Alright,

I've thought about both of the options I outlined in the previous post, and
it seems that speed-wise #2 is the better option, because we just figure
out UV mappings and build an image for the UV coordinates to map
into...once...and then we can use it to handle any deformation of the mesh.

BUT, there could be artifacts due to scaling/shearing of triangles...which
#1 is immune to.

I have implemented option #1, and will post it, along with some example
images to show what I'm talking about, tomorrow (too late to do it
tonight).

If anyone still thinks they have a better way to do it, please let me know.

This is probably something that could quite easily be bundled into POV-Ray
itself.

Cheers,
Joe

Post a reply to this message