Just a quick question:  Is it possible to modify regular scattering media to
work with MCPov?  The true montecarlo approach would be to spontaneously branch
a ray into the original and a deflected ray according to a scattering function,
a mean free path, and a density function.  As in gas dynamics, the mean free
path would be inversely proportional to the density.  This branching could be
done just once or even recursively, as long as a maximum branching limit or
adc_bailout were specified.  Is this unrealistic?  It's certainly slow, but it
seems like this would be very easy to implement.  Perhaps there are better
optimizations?

 - Ricky

Post a reply to this message