Atomic cell

[Vutshi]

Hi guys,

I'm doing rendering of a cell with atomic vapor which is illuminated by a laser. There is some scattering in the air to see the laser and atomic cloud in the cell also scatters light. The problem is that if I use rough glass then everything looks fine one can see the atoms inside the cell. However, if I switch to glass2 material (in the example I used it with negative ior, but I guess It shouldn't matter) it looks like the atoms in the cell are not going to appear at all. Is glass2 transparent for scattered light?

Rough glass n = 1.6

Glass2 n = -1.4

[jeanphi]

Hi,

Can you try with positive IOR? Since negative IOR has received almost no testing and is quite rare in real life, there might still be some bugs here and there but it might also result in some unexpected behaviour (like creating more total internal reflections that could mask the scattering).

Jeanphi

[Vutshi]

Hi Jeanphi,

I'm rendering it now with glass2 and positive n. It's still noisy, but one can see already that there is light on the floor under the cell which was scattered off the atomic cloud, however, we can not see this light directly.
This is done in the latest build of luxrender for mac from the forum.

[jeanphi]

Hi,

Would you mind posting the blend file?

Jeanphi

[Vutshi]

Here is the blender scene

[Vutshi]

Silly me! I didn't setup interior and exterior properly for the cell I set it to air (which does't scatter in this scene). That's because I wanted no scattering inside the glass cell except the atomic cloud.

[Vutshi]

Hmmm... Properly defined interior and exterior help a bit but not entirely. Now the glass cell is transparent for the light scattered in the media outside of the cell. Still there is no light from inside. I tried to set inside of atomic cloud as scattering media, exterior - air (no scattering), the exterior for inner surface of the cell is also air (normals are oriented toward the center of the cell). Also I experimented with a simpler case: there is no atomic cloud inside the cell, only exterior for inner surface was set to the same scattering media as outside of the cell. The result looks similar.

[jeanphi]

Hi,

When inside the glass, it is extremly difficult to see the light (classic SDS sampling problem), since the laser is extremely directional, the probability of sampling a scattering direction that falls on the laser is nearly 0. With roughglass, the algorithm is able to connect to path vertices on the glass surface so it is easier to get something visible.

Jeanphi

[Vutshi]

Does it mean that it is absolutely imposible to see the scattered light from inside of a glass object, or it is just a question of rendering time with current rendering algorithm?
Is there a way to add some roughness to glass2 material?

[J the Ninja]

SPPM would be able to do it I think, but it doesn't support scattering yet (and has a few weird shading bugs). Glass2 has no roughness atm. As far as adding that feature....uh...I haven't got a clue. Jeanphi?