In SubsurfaceBurleyNormalized.ush, in GenerateSampleInfo(), you used disk sampling right? But why is the PDF of it is R(r) * r * 2 * pi which I think is for sampling radius only? Shouldn’t it be R(r) * r, which is for sampling radius and angle(disk)? And in Unity they used R(r) * r. I’m confused sorry but correct me if I’m wrong.
The Slab node allows you to get (colored) transmitted light, though the control of MFP and thickness. If you want to control the pass-through (aka the old opacity input), you probably need a coverage node at the end of your slab, to control how much this slab covers. Maybe this is already something you already had?
/Charles.
Good question. This is orthogonal to Substrate and I’m not sure the reason behind that. I will ask around.
/Charles.
Charles, thank you for the reply. I still haven`t tried the Coverage node in that case. Gonna do it right now.
This does not ring a bell, and seems related to shadow cache/update rather than Substrate?
Interesting thing. The PT and Lumen show different behaviour when using Directional light. The PT kinda ignores it. Is it as supposed to be? First one is PT and thse second Lumen
One more fact. When unchecking CastRaytraycedShadows this strange thing happens. One Directional light, but seems like it is two.
using an hdri skylight with a “hot spot”? this may cast it’s own shadow into the lumen scene volume data.
Yeah I`m using HDRI Skylight. As you said it was the case. But why this shadow from windows visible only in path tracer mode?
Thank you for the help! I’m looking forward to hearing you back!
i figured it out, the main issue is that the stationary light shadows seems to break down on objects that have a large bounding box
I came across a really basic render issue with substrate this morning, but can find no reference online to anything similar. It’s SO basic that I was sure this meant I made a stupid error somewhere - but I’ve re-created it in a fresh download of the ‘Electric Dreams’ project in 5.5, so it seems to be legit.
The issue:
Layered substrate slabs exhibit a black edge / halo at the layer boundary when lit only by a skylight.
Close up - it’s only 1 pixel wide whatever the distance from camera:
Even closer:
To replicate this:
- Create the simplest possible layered slab material - if it uses a ‘vertical’ layer blend, the black line appears on the outer edge of the boundary, if it’s a ‘horizontal’ blend, it’s on the inner edge, but still visible
-
Create a new default level
-
Turn the skylight to ‘specified cubemap’, and select the default epic courtyard HDRI
-
Turn the directional light off
-
Add a mesh to the default level, and assign the material. In this test case, I used the ‘Electric Dreams’ project - the mesh is ‘Rock_Formation_04’ and the mask texture is ‘T_NoiseMask_RGB’, but any mask or mesh shows the same thing.
Is there a magic CVAR I’m missing?