|
 |
Thanks very much, this fixed my problem! It still confuses me how the
clipping effect could result from insufficient sampling, but oh well.. I
would expect it to simply make the data look more blurry.
Now I can get to my real renderings involving radiosity. Expect more
questions. :)
Thanks,
Eric
"Tim Nikias v2.0" <#macro timnikias (@) #local = "gmx.net" #end> wrote in
message news:40b123c4$1@news.povray.org...
> > The relavent POV code is below. Does anything look awry?
> > media {
> > intervals 15 //more needed? -> 100
> > //ratio 0.5
> > samples 2,6
> > method 3
>
> Well, there are three methods for media. I don't know anything about
method
> 1, since POV 3.5 I've just been using methods 2 and 3 (mainly method 3).
> With method 2, you want to use an amount of intervals which return
> sufficient results, and have to take care that "samples" uses the same
value
> for min and max, otherwise, the result might look grainy.
> With method 3, you want to use only one interval, but varying values for
min
> and max for "samples". I'd try method 2 for this volume first though.
>
> Now, given the amount of data in your volume, I'd say that you're using
way
> too few intervals. What might happen is that when the ray travels through
a
> dense region, it might sample before and after the bright region. Since
> those two samples are nearly identical, the adaptive-algorithm assumes
that
> between those two samples no significant changes occur, and hence, the
> bright spot is lost.
>
> Another note regarding your clipping-theory: when colors get clipped in
> POV-Ray, they get clipped to an upper or lower boundary, I've never heard
of
> values getting clipped via some modulo function (e.g. when clipping
between
> 0 and 1, 1.2 would turn out as 0.2).
>
> Regards,
> Tim
>
> --
> "Tim Nikias v2.0"
> Homepage: <http://www.nolights.de>
>
>
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.682 / Virus Database: 444 - Release Date: 5/11/2004
Post a reply to this message
|
|
