LuxCoreRender Forums

I really like this direction!

I've found an easy to understand comparison video between VCM and Path Space Regularization:
https://www.youtube.com/watch?v=Hc9zu5-O7Eo

So, if I understand correctly, the degree of the "blur" is adjustable, and it is the quality vs. speed variable here.

It works fine now!

Dade wrote: ↑Fri Aug 23, 2019 8:50 am It wasn't a glossy translucent specific problem but something affecting all diffuse/glossy and transmitting materials.

With the previous version I could reproduce the problem with rough glass too, but that is also good now.

Hmm, it doesn't work for me, and the file dates in the build are yesterday morning...

I think the deployment has failed:
https://dev.azure.com/LuxCoreRender/Lux ... mentId=396

Thank you!
I'll test it when it's released.

I thought it was quite visible in the original scene...

Without bump:
With bump:

Not sure if it matters, but I rather mix them after both of them are transformed to bump:

Just another side idea: If cache evaluation is dividable why not start rendering with the first bunch of entries and refining step by step. This would reduce waiting for first result, i.e. fast but inaccurate then improving stepwisely accuracy with the advantage to stop at any time. Something like ...

Dade wrote: ↑Thu Aug 22, 2019 5:15 pm ...it would be a lot easier to investigate with a simpler scene.

Here is a simple test scene.

The black leaf looks like something with flipped normals. I need a simpler set up. The three leaves share the same mesh data and the "black" one is scaled -1 on one axis to produce more variability. You could just delete that one. (I'm away from my computer at the moment, but will be back...

I made some tests, and I also think it works nicely.

...except the problem with the translucent materials:
http://forums.luxcorerender.org/viewtop ... f=4&t=1325