On Fri, 02 Jul 1999 20:19:35 +0300, Margus Ramst <mar### [at] peakeduee>
wrote:

>I've tried to model smoke with multiple containers (not that many, a few 1000s)
>but it was pretty slow. A particle system might need many more. I have doubts
>about the speed of this method. Perhaps you can give a few stats?
>
>Margus

Preliminary tests show that multiple particles is faster than multiple
media, and with better quality. For 100 particles render times were
like 2 vs. 5 minutes. Of course there's the problem with the high
max_trace_level and max_intersections that you might need...

I'll post more precise stats, test code, example images/anim tomorrow
(gimme some time to trace them)


Peter Popov
ICQ: 15002700

Post a reply to this message