Weird colorful noises Tweet

[darrenblack]

Hello!

I'm testing a scene that contains lots of specular materials and having weird colorful noises.
It doesn't go away with any settings
These are the settings I tried but none of them fixed the issue...
There is no scatter/SSS materials in the scene
Also, denoiser didn't really do anything but getting rid of grainy noises
Am I missing something???

Path Tracing / PMC
Max. samples 1024-4096
Diffuse / Specular depth 10-50
GI clamp 1-10
Caustic blut 0-1
Denoiser
Adaptive sampling ON (I also tried tweaking its Min samples higher)
Noise threshold 0.01-0.05
Hot pixel removal

[promity]

Try to play with Hot Pixels visibility.

[darrenblack]

Thank you for replying
I forgot to add it! I have already tried but it didn't work...

[elsksa]

Path Tracing / PMC
Max. samples 1024-4096
Diffuse / Specular depth 10-50
GI clamp 1-10
Caustic blut 0-1
Denoiser
Adaptive sampling ON (I also tried tweaking its Min samples higher)
Noise threshold 0.01-0.05
Hot pixel removal

Hi, some of the settings are irrational. Genuinely curious.
• Why would PT have a caustics blur of 0 but not in PMC?
• The GI Clamping is quite low. Does it make a difference to set it at 100 or 1000?
• Is 50 Specular Depth necessary? More or less around 15-20 would normally suffice in such scene.
Side note: the render setting extend to beyond this section.
This sampling guide will come handy.



Scene related questions:
• What is the light source?
• How many light sources?
• Any extreme values for light or material?


I don't think it's a sampling issue. I'm unable to reproduce it. I think it's, as mentioned, a faulty light source.
Last but not least, Filmic AGX will greatly reduce these, basically said, "strong clipped colors" whenever they appear naturally (unlike this case).

[darrenblack]

Oh, sorry for the confusion haha. The settings I wrote were just "range" I tried.
Not like, seperating into two parts (Path tracing and PMC)
Some of options I tweaked seemed not related but I just needed to anything I can to fix the issue.

Apparently you are right! It wasn't an issue related to sampling issue.
It was the lighting issue... I used colorful HDRI as a light source.
When I removed it, it was fixed.
I wasn't even thinking light sources could be an issue in this case since I used only 2 lights (HDRI and Daylight)
Thank you for helping me!