Cool idea!

> >>The transmission of light is nothing that makes sense to simulate in
> >>POV-Ray since it has no effect on the image POV-Ray renders (light itself
> >>isn't visible).  Therefore using emitting media for the glow and placing a
> >>spotlight at the end is the way to go.

A MUCH faster way to get the "points of light" at the ends of the fibers
(like in a flower sculpture) might be to attach spheres at the end the same
diameter as the cable.  Give them a high ambience.

>
>  That being said, I did find out that the coating layer has a lower IOR
> than the fiber core.  This is explained by that link you posted. When
> passing through one medium strikes a medium with a lesser IOR, "the ray
> is bent away from the normal." Thus, the angle of reflection will be
> greater than the incident angle. (I can picture this.  That doesn't mean
> that I completely understand it.)

This might actually be made to work with photons, although be prepared to
have your swap file limit increased to an astonomical size and a LONG
render time!  I'd love to experiment just now but MY machine is 15 hours
into a major render with photons and my swap file just expanded to over two
gigs :o(  One potential problem I forsee is the dreaded co-incident surface
bug.  Photons only refract when entering (or leaving, depending on your
point of view) an object.  An object can only have one ior in it's
interior.  So to simulate two ior's I'd put a solid sphere sweep with high
ior inside a larger, hollowed out (NOT hollow) sphere sweep with a lower
one. Make sure the inner one is a little bigger than the hollow in the
outer one is, which prevents a co-incident surface, and then merge the two
together to remove interior surfaces.  A lot of this is personal conjecture
on my part but it's where I'll start next week when I get POV back again.

Post a reply to this message