Scott Jewell wrote in part:
> ... I was really surprised that POV doesn't seem to have the
> ability to output some sort of triangle mesh output (or does it?). ...
> I guess I expected at least a 3DS sort of thing since it seems that
> POV at some point obviously knows the coordinates of the object.

Well, sorta; some points on the object's surface are directly measured,
when a ray intersects it.  But that information does not translate
directly to a mesh, because there's no topology in it.  

You could use a program similar to PoV-Ray to parse an object's
description and make a list of intersection points (or interior points)
to whatever resolution you specify; but the resemblance to PoV-Ray ends
there.  You'd then need to determine how these points connect to form
triangles.  There's apparently a good algorithm for that, but still it's
bound to make mistakes with an object like
http://ogre.nu/images/spider.jpg -- some gaps will be counted as bridges
and vice versa.

-- 
Anton Sherwood  --  br0### [at] p0b0xcom  --  http://ogre.nu/

Post a reply to this message