Page 1 of 1

Looking for suggestions with PMC grain

Posted: Sun Jun 14, 2015 7:28 pm
by dotcommer
I've gone through InLifeThrill's optimization video a couple times now, and i've tried various settings to keep grain down, but this scene i'm working on for the metaverse challenge is just not working out. Its insanely grainy. I have one massive light source which is the sun. Any other lighting in the scene is from mesh lights. I just can't get the dark areas to clean up at all even after leaving it over night for almost 8 hours. Looking for any suggestions. The scene is massive; over half a million polys and is completely enclosed. I know Octane has a difficult time with interior scenes, but this is crazy. Not even forcing PMC to be more biased helps. Here's a preview image for reference, as well as current render settings that are my best guess at the moment. Currently rendering at .25 Ms/Sec.
Observation_deck_v28_v01.jpg
Screen Shot 2015-06-14 at 12.24.03 PM.png

Re: Looking for suggestions with PMC grain

Posted: Mon Jun 15, 2015 5:47 am
by bepeg4d
hi,
my suggestion is to activate the multi-light and multi-pass options in order to better investigate what causing the noise. Maybe you have to play with the sampling rate of the mesh emitter materials :roll:
ciao beppe

Re: Looking for suggestions with PMC grain

Posted: Mon Jun 15, 2015 6:09 am
by dotcommer
Thanks for the suggestion Beppe. Would I find that under the info channels kernel?

Re: Looking for suggestions with PMC grain

Posted: Mon Jun 15, 2015 9:58 am
by bepeg4d
hi,
when you have find the guilty emitter, you can change the sampling rate for any individual emitter material in the attribute of the blackbody or texture emitter node. By default is set to 100 in c4d, you can increase for noised emitters till 10.000. Take care that changing the sampling rate, there could be some difference also with the power that must be corrected accordingly ;)
ciao beppe

Re: Looking for suggestions with PMC grain

Posted: Mon Jun 15, 2015 10:06 am
by Yambo
Why you went with pmc and not pathtracing?

Re: Looking for suggestions with PMC grain

Posted: Mon Jun 15, 2015 2:21 pm
by dotcommer
Yambo wrote:Why you went with pmc and not pathtracing?
PT created some weird results with the glass in the windows. I was getting upside down views through the windows with PT, so I switched to PMC which handles complex caustics and specular materials better.

Re: Looking for suggestions with PMC grain

Posted: Mon Jun 15, 2015 2:31 pm
by brasco
With dots and small non-noise related artefacts try reducing Ray epsilon, I've had some success with some issues I had.

cheers
brasc