LuxCoreRender Forums

Dade wrote: ↑Sat May 08, 2021 12:02 pm But where does Blender takes the VRam usage ? I doubt it ask it to BlendLuxCore. It may be a totally wrong number.

It is retrieved from LuxCore stats: https://github.com/LuxCoreRender/BlendL ... cs.py#L102

Awe-inspiring, outstanding work!
I'll embed this in the "new in v2.5" page as an example of what can be done with the new light group workflow.

Nice models!
I like the plants and ground in kintuX' edit, but the sky is too saturated/contrasty there in my opinion.

Dez! wrote: ↑Mon May 03, 2021 9:27 am I note that the increase in Max Epsilon badly affects the preparation of rendering when the GPU is enabled.

Each different epsilon value requires a re-compilation of the rendering kernels, maybe you mean that? (If the epsilon is no longer changed after that, the kernels will be re-used of course)

Dade wrote: ↑Thu Apr 29, 2021 2:22 pm LuxCore "v3.0" will require no compilation at runtime, not even the first time.

Awesome to hear!

Can't you just set all object material holdout to "1" if the button on the object panel is set and vice versa ? It is different from Cycles behavior but it makes sense. Cycles can have two objects with the same material, where one object has holdout enabled and the other not. I don't see ...

The problem is that LuxCore has the holdout toggle only in the material settings. Cycles has it both for materials and for objects. So if you have for example 5 objects with the same material, and you enable holdout on 2 of those objects, that's simple in Cycles: just enable holdout in the object se...

I would also find such a feature useful, but AFAIK it's currently not possible in Lux. (Although the use case for such a split is usually the opposite: you want to use a high-res LDR for reflections and a low-res HDR for lighting, to save memory compared to one high-res HDR for everything). What I w...

Looking at the code, I assume this may happen because BlendLuxCore is setting the properties as string or float (?), not as integer so it should be fixed in BlendLuxCore. This value is the output of the round() function. According to the Python documentation it returns an integer if the ndigits arg...

Can this be improved by enabling cache or maybe issue with Space Partitioning to be more particular? Or maybe if we assign floating point decimal places base on distances from origin maybe it will smooth out the blocks noise? Just guessing for now. The linked article mentions a possible solution: r...