> It's probably not actually a realistic effect, but a pale
> simulation of one...;)
>
> (the ultralow confidence/ultrahigh variance is how to optimize
> the speed of media when you're playing around with the OTHER
> numbers and are doing several dozen renders of the same thing
> in a row)

Interesting. Was this on 3.1 or 3.5? (So that I can know what sampling
method it was.) I'm beginning to think that method 1, if used well, is more
powerful than method 3. Sometimes I'd rather have a grainy artifact than the
more noticeable artifact that method 3 sometimes gives with too few samples.

 - Slime
[ http://www.slimeland.com/ ]

Post a reply to this message