I read the paper (very nice images in it).  But you'll notice that, unlike
the original work that Jensen did on photon maps (which ran on a low-end
linux box), the atmospheric stuff was rendered on an 8 (or was it 16?)
processor graphics workstation with many, many megabytes of memory.

I do plan to implement it, but it's not top priority right now.  First,
I want to work on finishing photon mapping for caustics, then make it
work with radiosity.  Also, I want to have a closer look at media to
try to get back the adaptive sampling method before introducing volume
photons.

-Nathan

Mike wrote:
> 
> The same guy who came up with the photon mapping idea wrote a paper
> decribing how to extend it to work with participating media too.  I
> don't know if Nathan has this paper, but it would be neat if it could be
> worked into POV sometime down the road.
> 
> It can handle caustics in media and even diffuse <-> volume light
> transfer!  :)
>

Post a reply to this message