 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Can you guess how many dispersion_samples I used in this image?
Post a reply to this message
Attachments:
Download 'dispersion_test.png' (35 KB)
Preview of image 'dispersion_test.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Greg Edwards" <edw### [at] hotmail comremovethis> wrote in message
news:78v### [at] 40tudenet...
> Can you guess how many dispersion_samples I used in this image?
42
~Steve~
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sat, 1 Mar 2003 00:37:22 -0000, St. wrote:
>
> "Greg Edwards" <edw### [at] hotmailcomremovethis> wrote in message
> news:78v### [at] 40tudenet...
>
>> Can you guess how many dispersion_samples I used in this image?
>
> 42
>
> ~Steve~
Heheh, that's what it looks like, but lower. ^_^
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
3e6000d6@news.povray.org...
>
>
> 42
>
> ~Steve~
>
>
I'd like to test Deep Thought as a render computer ( for long renders?)
Slartibartfast
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sat, 1 Mar 2003 08:15:06 +0100, Marc Jacquier wrote:
>
> 3e6000d6@news.povray.org...
>>
>>
>> 42
>>
>> ~Steve~
>>
>>
> I'd like to test Deep Thought as a render computer ( for long renders?)
> Slartibartfast
Actually, this scene only took a few minutes to render IIRC. That's the
beauty of this work. ^_^
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
5?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sat, 1 Mar 2003 22:57:52 -0000, Andrew Coppin wrote:
> 5?
Hehehe, a *very* good guess, but... lower! ^_^
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
The correct answer is:
2
In case the impossibility didn't hit you yet, I made a patch for POV-Ray
which adds a new speed and quality feature to POV's dispersion. It's
ever-so-creatively called dispersion jitter. Here's the interior{}
statement I used:
interior
{
ior 1.5
#if (!Photons)
caustics 1
#end
dispersion 1.1
dispersion_samples 2
dispersion_jitter 1
}
I believe the concept of jittering the dispersion samples and my
implementation are physically realistic and my implementation is compatible
with photon mapping (see attached image). I plan to make this patch
(currently a Windows binary) available as soon as possible.
I've attached a higher quality image with dispersion_samples 5 and photons
and radiosity enabled. This image took a little over 5 hours to render @
960x720 +a0.1 +r4 IIRC with a P3 871 under the noticeably slower VC++
compile.
Post a reply to this message
Attachments:
Download 'disp-jitter-test.jpg' (62 KB)
Preview of image 'disp-jitter-test.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
So... it only does 2 (random) colours per pixel, and AA blurs it so much it
looks clear again?
Andrew.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sun, 2 Mar 2003 12:56:42 -0000, Andrew Coppin wrote:
> So... it only does 2 (random) colours per pixel, and AA blurs it so much it
> looks clear again?
>
> Andrew.
The first image I posted had only dispersion_samples 2 and lots of AA to
clear things up but the patch can handle any number of dispersion_samples
and dispersion_jitter can be set to any float value but if it's out of the
0 to 1 range, a warning will be issued and the problem corrected. You can
also specify dispersion_jitter on which will set it to 1. The result is
similar to the jitter effect of area lights. Any number of samples can be
used but using many samples will just allow you to use less AA to get good
results. The second image used dispersion_samples 5.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
