OTOY Forums

Here I have posted some examples that describes how ies lights are rendered with all rendering kernels available in Octane for Max
In this scene ies lights (8 instances) are boosted with 100 samples

To simulate the emitter I created a self illuminated disc and I placed below one ies emitter with a diffuse+blackbody with a power of 0.1
Probably the ies geometry reflects the low disc light creating a lot of fireflies
All renderings have a 0.25 fireflies reduction.

I wouldn´t use a subD mesh for so a simple type of a wall.
To many polygons on a very small area. Maybe you can try to reduce the polycount and see what happens...

face

It´s also possible, on that small areas on your openings, that the normal angle can make trouble.
Some edges should be hard, some smooth, that would explain the shading issue.

Have overlap the wireframe with the render result and it seems, that the dark areas hit an edge face

face wrote:I wouldn´t use a subD mesh for so a simple type of a wall.
To many polygons on a very small area. Maybe you can try to reduce the polycount and see what happens...

face

the problem is not in the mesh but in the render algorithm, in particularly in the pmc kernel.
As you can see PT renders correctly that mesh.

The issue happens only in low light conditions and with the pmc kernel only.

Hard to see because the PT is realy grainy.
It seems that you have also issues with DL. I have try to do similar with bad and good normals, but only with dl.
If your angle between two polygons is flatter and you have more polygons, and you have, the issue is looking like yours. I would try to render without subD and show what happens.

face

Can you check with the deep_channel_kernel if the normals are correct at those bezels?

--
Roeland