 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I quickly patched mlpov to include the halton distribution for radiosity.
You can download the win binary :
http://195.221.122.126/mlpov82-win-bin.zip (923 ko)
I'm not good at building radiosity test scenes (I'm not good at building
scenes at all :) so I can't really be sure that halton is better , So feel
free to try it with your radiosity scenes for low/mid/high count, your
comments are very welcome !
To activate the halton sequence instead of pov3.5 distribution :
#version unofficial mlpov 0.82;
radiosity {
halton
..
}
if you use halton you can specify count > 1600
More test images with the scene from Jerome Colin :
http://195.221.122.126/samples/erreur_images2.html
(Warning, 6 png images = 350 ko, no jpg to avoid any compression artifacts)
to see differences between images you can download them and use an image
browser to quickly switch from one to another
M
Post a reply to this message
|
|
| |
| |
|
|
From: Christoph Hormann
Subject: Re: mlpov0.82 with halton for radiosity
Date: 7 Jan 2003 14:30:19
Message: <3E1B2ACA.5FC32313@gmx.de>
|
|
|
| |
| |
|
|
Mael wrote:
>
> I quickly patched mlpov to include the halton distribution for radiosity.
> You can download the win binary :
> http://195.221.122.126/mlpov82-win-bin.zip (923 ko)
>
> I'm not good at building radiosity test scenes (I'm not good at building
> scenes at all :) so I can't really be sure that halton is better , So feel
> free to try it with your radiosity scenes for low/mid/high count, your
> comments are very welcome !
I have made some more tests including the halton distribution:
http://www.schunter.etc.tu-bs.de/~chris/files/rad_test.html
The directed sample sets only contain samples on the upper part of the
hemisphere - this leads to less artefacts because the samples are more
dense but to a quite different overall appearance.
Christoph
--
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 31 Dec. 2002 _____./\/^>_*_<^\/\.______
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> The directed sample sets only contain samples on the upper part of the
> hemisphere - this leads to less artefacts because the samples are more
> dense but to a quite different overall appearance.
visually pleasant but probably not physically correct isnt it ?!
if we get rid of "most photorealistic" constraint, it became another problem
:)
M
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
If you ask me, none of them looks good, even with 300 samples. It seems to
get very noisy, not smoother.
Also, it seems like the Halton has a lower constrast / higher brightness
than the conventional sample. It's debatable whether this is a good or a bad
thing.
--
Anthony Bennett
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
