Page 13 of 17

Re: Test scences GPU Path vs CPU BiDir

Posted: Wed Aug 29, 2018 9:02 am
by FarbigeWelt
B.Y.O.B. wrote: Tue Aug 28, 2018 5:46 pm the glossy translucent material scatters light uniformly, no refraction happens.

In LuxCore, materials are divided in three categories: diffuse, glossy and specular.
Rough glass is a "glossy" material.
Metal is also a "glossy" material.
Matte would be a "diffuse" material.

If you look at the source code of the glossy translucent material, the transmission event is "DIFFUSE | TRANSMIT", while the reflection event is "GLOSSY | REFLECT".
Thank you for clarifying this topic. I will consider material‘s proprties in source code for next studies.

Clear glass is not very transparent for scattered light

Posted: Wed Aug 29, 2018 10:09 pm
by FarbigeWelt
Did you ever point a laser at a scattering medium like fog or smoke or some turbid liquid? What did you see?
It is easy to simulate a laser beam in homogen air with lux core but it is very time consuming to render the beam visible if it is behind a flat glass, no matter how thin or clear it is. In fact even with reflection of 0.0 the glass seems to absorb the scattered beam.
This short animation shows pictures with 20 to 1080 Samples. As you can see beam gets the more visible the more sample there are. What happens to the scattered ray? For the moment I don't see any reason for this behaviour. Even with IOR set to 1.0001 glass aborbs the scattered beams.
Glass-and-Lasers
Glass-and-Lasers
Glass-and-Lasers.gif (467.82 KiB) Viewed 7010 times
The picture below shows a hollow cube. Its outer faces are glass, its inner faces are null. The null's interior volume is clear and its exterior is homogen with IOR of dense flint.
One can do lots of more or less meaningfull combinations, I know.
Outside Glass Inside Null Cube filled with Homogen 1146 Samples
Outside Glass Inside Null Cube filled with Homogen 1146 Samples
In comparision, here it is homogen with IOR 1.0001.
Outside Glass Inside Null Cube filled with Homogen Air 376 Samples
Outside Glass Inside Null Cube filled with Homogen Air 376 Samples

Glossy transluscent: cube size 1,1,0.003

Posted: Thu Aug 30, 2018 7:56 pm
by FarbigeWelt
One can see glossy transluscent scatters light on the surface and let light 'pass' to a certain depth.
Glossy Transluscent Cube 1,1,0.003 four lasers 1500 Samples
Glossy Transluscent Cube 1,1,0.003 four lasers 1500 Samples
If object is too thick light cannot pass as beam from entry to exit but backside can be expected to glow smoothly up to a thickness of 0.2 with settings for this material.
Glossy Transluscent Material
Glossy Transluscent Material

Visibility of scattered light

Posted: Thu Aug 30, 2018 9:45 pm
by FarbigeWelt
What ever happens here it fits not to my expections and experiences with glass and mirror in real life. There must be something wrong with the implementation of the interaction of scattered light with surfaces of some material settings.
Glass: Rough or Architectural works, default not.
Glossy: Rough 0.1 mirrors, Rough 0.01 barely, Diffuse not.
Mirror: works not.
Metal: Fresnel mirrors mediocre.
Matte: spreads well.
Visibility of scattered light
Visibility of scattered light
Maybe this helps to understand how the scence looks like.
10 Matte with lasers.png
10 Matte with lasers.png

Re: Test scences GPU Path vs CPU BiDir

Posted: Thu Aug 30, 2018 11:14 pm
by B.Y.O.B.
Renderengine and sampler?

Re: Test scences GPU Path vs CPU BiDir

Posted: Fri Aug 31, 2018 6:03 am
by FarbigeWelt
B.Y.O.B. wrote: Thu Aug 30, 2018 11:14 pm Renderengine and sampler?
CPU BiDir, random. (With path I don‘t see much of the laser beams in scattering medium.)

It happens at least for glass with sobol too, but it is a bit slower for these scenses.

Re: Test scences GPU Path vs CPU BiDir

Posted: Fri Aug 31, 2018 7:48 am
by B.Y.O.B.
Could you try with Bidir + Metropolis?

Scattered Lasers CPU BiDir, Metropolis

Posted: Fri Aug 31, 2018 10:28 pm
by FarbigeWelt
B.Y.O.B. wrote: Fri Aug 31, 2018 7:48 am Could you try with Bidir + Metropolis?
Behaves as random or sobol regarding glass.
CPU BiDir Lasers Scattering, Metropolis
CPU BiDir Lasers Scattering, Metropolis

Scattered Lasers CPU BiDir: start from scratch

Posted: Sat Sep 01, 2018 12:35 pm
by FarbigeWelt
Somehow materials got stranger with every change or use in the blender file used for the scenes above.
A good reason, to start with a new blender file and define objects and materials from scratch.
Now, glass is close to expected and experienced behaviour again. (It seems like material definitions can be screwed in a blender file leading to strange effects.)
CPU BiDir Null and 10 Glasses at least 600 Samples
CPU BiDir Null and 10 Glasses at least 600 Samples

Re: Test scences GPU Path vs CPU BiDir

Posted: Sat Sep 01, 2018 1:08 pm
by B.Y.O.B.
FarbigeWelt wrote: Sat Sep 01, 2018 12:35 pm Somehow materials got stranger with every change or use in the blender file used for the scenes above.
A good reason, to start with a new blender file and define objects and materials from scratch.
...wat?