Page 3 of 4

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 10:20 am
by Magog
Sharlybg wrote: Fri Nov 29, 2019 10:11 am So you don't have any caches on ?
Yes, I do.
You can see the screen in the first post..

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 10:31 am
by PetrT
PetrT wrote: Thu Nov 28, 2019 6:29 pm may be try to render with more resolution then needed and downscale to required
i mean to render in 4k or more dimensions if it possible
2000 samples is enough

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 10:34 am
by Sharlybg
Yes, I do.
You can see the screen in the first post.
Ok i see. You don't really need direct light sampling cache here. go uniform or log power. can we see your PGI all settings ( indirect light cache part) .

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 10:44 am
by Magog
Changing the Denoise [BCD] finally has a good compromise

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 10:46 am
by Magog

Ok i see. You don't really need direct light sampling cache here. go uniform or log power. can we see your PGI all settings ( indirect light cache part) .

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 11:03 am
by Sharlybg
Reduce brute force radius scale to 2 or 4.

The lower the faster. too low and you get light leak. i usually start with 2 and increase only if needed.

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 11:10 am
by Racleborg
Changing the Denoise [BCD] finally has a good compromise
That looks way better! :D

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 11:12 am
by Dade
@Magog, can you post a test scene with just a single couch pillow (if the major problem with the denoiser is there) ? So I can check if there is some problem with Oidn and where it comes from (or is just a limit of the Intel software, at the end of the day all denoisers have some limit).

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 12:21 pm
by Magog
Dade wrote: Fri Nov 29, 2019 11:12 am @Magog, can you post a test scene with just a single couch pillow (if the major problem with the denoiser is there) ? So I can check if there is some problem with Oidn and where it comes from (or is just a limit of the Intel software, at the end of the day all denoisers have some limit).
https://drive.google.com/open?id=1ovyDS ... jvJ6KeWhVN

If I set the render to 4k and after the calculation of the kernel, blender crash!
I also noticed that with the Particle System of the carpet and vegetation in the external area, the calculation does not start.

Re: Simple Test Scene

Posted: Fri Nov 29, 2019 12:25 pm
by Dade
Magog wrote: Fri Nov 29, 2019 12:21 pm
Dade wrote: Fri Nov 29, 2019 11:12 am @Magog, can you post a test scene with just a single couch pillow (if the major problem with the denoiser is there) ? So I can check if there is some problem with Oidn and where it comes from (or is just a limit of the Intel software, at the end of the day all denoisers have some limit).
https://drive.google.com/open?id=1ovyDS ... jvJ6KeWhVN
Thanks, checking.
Magog wrote: Fri Nov 29, 2019 12:21 pm If I set the render to 4k and after the calculation of the kernel, blender crash!
I also noticed that with the Particle System of the carpet and vegetation in the external area, the calculation does not start.
You may be running out of GPU memory, does it happens also if you try a CPU-only rendering (PATHCPU instead of PATHOCL) ?

Does it happen both with Oidn and BCD ?

Unfortunately, BCD denoiser requires a lot of RAM for each pixel so the total amount of ram used at high rendering resolutions is very high.