LuxCoreRender Forums

alpistinho wrote: ↑Fri Apr 05, 2019 12:43 am I am facing issues to install AMDGPU pro drivers...
I think I will create a dual boot with Ubuntu 18.04 in order to solve, since linux Mint is unsupported

I can look into the OpenCL part if you want, it is really intricate and cryptic code anyway.

"loopsubdiv" takes as input a mesh and further subdivide the surface: for instance, it takes a sphere made of 100 triangles and return a "smoother" sphere made of 400 triangles. It can easily generate millions of triangles if you subdivide too much. It is not supported by LuxCore...

Piita, welcome back

Are you following the procedure described here https://github.com/LuxCoreRender/WindowsCompile

Tried to render the blender scene "Sharlybg_Food.blend" ,I can see the render on the viewport but when i try to do a final render this error shows up opencl device selection string has the wrong length, must be 3 instead of 1 I guess the scene was saved on a PC with 3 OpenCL devices, just...

May be, you could try to use glass material and apply the texture with the finger prints to bump map slot

The texture should probably have noise inside the finger prints.

Could we test it just downloading the next daily build? Yes, if you just set the radius manually to 0.0 (if it is allowed by BlendLuxCore). However, I have not yet changed anything else (I started from more obvious things to borrow from PhotonGI): it is a only "quality of life feature". P...

This means that one could have an light guide effect of a glass tube surrounded by some black acrylic cladding? At least for a very narrow (almost parallel) beam pattern it should work... I'll investigate that a bit... Yes but be careful with very near surfaces, it is a situation easily prone to nu...

DLSC now automatically estimated the best radius if lightstrategy.entry.radius is set to 0.0. So it works like PhotonGI and there is no more need to worry about picking the right radius by hand.

I will write the OpenMP version later today, and will set up the OpenCL drivers to compile this version. If there is no convergence calculation on the GPU side, it should be trivial to implement the new test there too. The new test uses two comparisons and a subtraction, while the old uses just a c...