On Sun, 13 Jan 2002 12:24:55 +0200, Eugene Arenhaus <eug### [at] avalon-netcoil> wrote:
> How? Easy. You write a Trace routine for transparency shader that will
> create two more ray objects (passing and bouncing), cast them (by
> calling Scene's Trace routine), combine the result in its single ray and
> return that. 
> 
> Or, in short: You can cast a zillion rays if you want simply by
> recursively calling Trace.

Yeah, I can see that... But I have to wonder if POV is not already doing something
like this then
how is it doing it? I see no way of returning an accurate value for any single point
unless it
already impliments something of this sort. If it doesn't, then there is bound to be a
good reason
why it doesn't. As for my not thnking... Maybe, but then I could say the same, since
as I said
this time, I don't get how the current behaviour could differ significantly from what
you suggest
and still work, unless there is some very specific optimization or feature that
requires it be done
some other way. And in that case I can't even imagine what they did do instead of a
single
trace() routine. But then I haven't exactly looked at the code either. I have enough
problems
figuring out why I did something after leaving program code alone for a coupla years,
let
alone trying to figure out how or why someone elses (far larger and more complex) code
does
anything. lol

Post a reply to this message