LuxCoreRender Forums

Dade wrote: ↑Tue Apr 17, 2018 7:28 pm

kintuX wrote: ↑Tue Apr 17, 2018 6:24 pm 4. BiDir, Clamp Output (The higher the Max. Bright. value, the smaller the fireflies appear - IIRC, known issue)

That is a very strange problem, can you post a test scene ?

Same for me. I've the same problem in my test scene http://forums.luxcorerender.org/viewtopic.php?f=2&t=210
Yesterday I tried using an emitter material for the LED part of the light lamp and If I set a clamp value, the reflections seems broken after some samples. The issue starts with small black pixels and then they became more visible with wide black pixels area.

The strange thing is that today the pixel are white and yesterday was black...anyway it seeems there is something wrong.

marcatore wrote: ↑Wed Apr 18, 2018 6:53 am The issue starts with small black pixels and then they became more visible with wide black pixels area.

The strange thing is that today the pixel are white and yesterday was black...anyway it seeems there is something wrong.

Black pixels are likely to be NaN (Not A Number) and white probably Inf (Infinite). I'm going to check.

Dade wrote: ↑Wed Apr 18, 2018 8:14 am

marcatore wrote: ↑Wed Apr 18, 2018 6:53 am The issue starts with small black pixels and then they became more visible with wide black pixels area.

The strange thing is that today the pixel are white and yesterday was black...anyway it seeems there is something wrong.

Black pixels are likely to be NaN (Not A Number) and white probably Inf (Infinite). I'm going to check.

Strange problem, it happens only with IES lights and clamping (IES lights alone and clamping alone are working fine). I'm tracking where is the problem.

Dade wrote: ↑Wed Apr 18, 2018 9:38 am Strange problem, it happens only with IES lights and clamping (IES lights alone and clamping alone are working fine). I'm tracking where is the problem.

I should have fixed the problem. You can download an updated pyluxcore.pyd for Windows (with OpenCL) from here: https://drive.google.com/file/d/13bkQ3J ... sp=sharing

Dade wrote: ↑Wed Apr 18, 2018 1:21 pm

Dade wrote: ↑Wed Apr 18, 2018 9:38 am Strange problem, it happens only with IES lights and clamping (IES lights alone and clamping alone are working fine). I'm tracking where is the problem.

I should have fixed the problem. You can download an updated pyluxcore.pyd for Windows (with OpenCL) from here: https://drive.google.com/file/d/13bkQ3J ... sp=sharing

b.) BiDir, Clamp Output: Enabled, Set Suggested Value: 0.208361 (original post w/ bug)

Solution confirmed!
TYVM

Playing around, trying to understand few things. Are these correct?

1. Only Homogeneous & Heterogeneous Volumes allow multiscattering, but not Clear Volume.

2. Glossy Coating Material doesn't show Gloss (Reflection) if Null (Transparent) is used as Base. Material is still only transparent.
Here I expected Gloss to be visible on surface based on IOR value (similar to what you get with Cycles).
- Is there an alternative to create such material?

3. Material node doesn't read Auto Texture Space from Curved Object using Bevel.
- Must make it a mesh?

... working really nice.
TIA

kintuX wrote: ↑Sun Apr 22, 2018 10:38 pm 1. Only Homogeneous & Heterogeneous Volumes allow multiscattering, but not Clear Volume.

Clear volume does not have scattering at all, that's why it's named "clear".

kintuX wrote: ↑Sun Apr 22, 2018 10:38 pm 3. Material node doesn't read Auto Texture Space from Curved Object using Bevel.
- Must make it a mesh?

Please upload a test scene that shows the problem.

kintuX wrote: ↑Sun Apr 22, 2018 10:38 pm 2. Glossy Coating Material doesn't show Gloss (Reflection) if Null (Transparent) is used as Base. Material is still only transparent.
Here I expected Gloss to be visible on surface based on IOR value (similar to what you get with Cycles).
- Is there an alternative to create such material?

This makes sense because NULL is like a material with 100% transparency: the rays will just go trough NULL like if the object doesn't exist at all (so even the coating is not visible).

Instead you could a Glass material of NULL with IOR 1.0 and (0.0, 0.0, 0.0) reflection and (1.0, 1.0, 1.0) refraction plus Coatting. It should provide the result you were looking for.

B.Y.O.B. wrote: ↑Mon Apr 23, 2018 6:04 am

kintuX wrote: ↑Sun Apr 22, 2018 10:38 pm 1. Only Homogeneous & Heterogeneous Volumes allow multiscattering, but not Clear Volume.

Clear volume does not have scattering at all, that's why it's named "clear".

kintuX wrote: ↑Sun Apr 22, 2018 10:38 pm 3. Material node doesn't read Auto Texture Space from Curved Object using Bevel.
- Must make it a mesh?

Please upload a test scene that shows the problem.

Within the scene file here there's a model (bottle > circle) where the node that needs UV (ie. Glossy Coating Material) states: "No UV map" (That's specifically what i was wondering about.)

BTW Thanks for clearing my scattered mind

-----------------------------------------------------------------

Dade wrote: ↑Mon Apr 23, 2018 8:29 am

kintuX wrote: ↑Sun Apr 22, 2018 10:38 pm 2. Glossy Coating Material doesn't show Gloss (Reflection) if Null (Transparent) is used as Base. Material is still only transparent.
Here I expected Gloss to be visible on surface based on IOR value (similar to what you get with Cycles).
- Is there an alternative to create such material?

This makes sense because NULL is like a material with 100% transparency: the rays will just go trough NULL like if the object doesn't exist at all (so even the coating is not visible).

Instead you could a Glass material of NULL with IOR 1.0 and (0.0, 0.0, 0.0) reflection and (1.0, 1.0, 1.0) refraction plus Coatting. It should provide the result you were looking for.

Thanks. I assumed that would be the case... and in the end came up exactly with what you described. Great.

kintuX wrote: ↑Mon Apr 23, 2018 8:51 am Within the scene file here there's a model (bottle > circle) where the node that needs UV (ie. Glossy Coating Material) states: "No UV map" (That's specifically what i was wondering about.)

Is the mapping working correctly in the rendered image?
The check only looks for a UV map in a mesh datablock, it probably does not even work for curves. If the mapping is working correctly and only the warning is shown, then I will have to correct the code for the warning.