Albedo - Proposal?

[Dionysios.TS]

Hi all,

I used till last September Nvidia iray in my professional workflow and I decided to start using Corona and personally I am very excited and happy with the results! Is a very powerful engine! Something I really miss and I bet all of you too, is a simple tool where we can set the correct Albedo values to our RGB colors and textures. Back in the iray beta testing cycle, I asked at the dev team to add the Reflectance Values & Transmittance Values which you can find in the material editor to their shaders too and they made it and it always worked out!

So, why we can't do the same for Corona? Those values are extremely useful and they work with both RGB colors and texture maps. Also I remember the iray team told me that the Reflectance values was calculating also the Reflection values as well!

What do you think?

Thanks in advance,

Dionysios -

[Ludvik Koutny]

In corona, you do not need to watch albedo of reflection. Only shading components that should have reasonable albedo (below RGB 220) are diffuse color, translucency color and scatter color. You can add CShading_Albedo render element to check your scene for any materials with problematic albedo (they will show red on the render element)

[romullus]

Rawalanche, could you explain a bit more about translucency and scatter colours? Never heard about that. IIRC in latest material tutorials you emphasized several times that translucency should be set to pure white and now you're saying that it should be kept under RGB 220. I'm a bit confused.

[Ludvik Koutny]

Rawalanche, could you explain a bit more about translucency and scatter colours? Never heard about that. IIRC in latest material tutorials you emphasized several times that translucency should be set to pure white and now you're saying that it should be kept under RGB 220. I'm a bit confused.

Ah right, yeah... For translucency it applies only in case of thin objects, like leaves, paper or curtains. If you use translucency on thick objects to reveal SSS, then that rule doesn't apply.

[romullus]

So, for thin objects translucency should be kept lower than RGB 255, and for thick objects it should be at RGB 255, right? Or the other way?

[Ludvik Koutny]

So, for thin objects translucency should be kept lower than RGB 255, and for thick objects it should be at RGB 255, right? Or the other way?

Translucency should have reasonable albedo whenever you use it on it's own. As you basically emulate object that is very thin, but still has some thickness, and light energy gets lost as it passes through that small thickness. If you use translucency along with SSS, then it's a different story, ray will pass through the surface and will keep 100% of its energy, but that distance in which it keeps its energy is extremely small because as soon as it passes the wall, it will start to lose energy in SSS volume. So there we no longer try to "fake" some very small thickness, but actually simulate it using SSS volume.

That being said, perhaps even when using translucency with volumetric SSS, it could be beneficial to use not completely white translucency color. But I haven't tested it yet. You could if you want. I mentioned my workflow just because it's been a rule of thumb that always worked for me in realistic scenario, not because it's an ultimate truth ;)

[maru]

Could it be simplified to saying that translucency and scattering colors can be perceived the same way as diffuse color, but not of the object's surface, but the tiny particles that fill its inside in real life?

Another thing - I remember Ondra saying that it's fine to leave albedo at some extreme values provided that it's not applied to large flat areas, such as walls, that contribute much to GI. So applying a very bright material to some small detail would be fine. Can someone confirm this?

[romullus]

Thanks for explanation Rawa! I trust your experience enough to adapt your workflows straight away without testing ;]

[Ludvik Koutny]

Could it be simplified to saying that translucency and scattering colors can be perceived the same way as diffuse color, but not of the object's surface, but the tiny particles that fill its inside in real life?

Another thing - I remember Ondra saying that it's fine to leave albedo at some extreme values provided that it's not applied to large flat areas, such as walls, that contribute much to GI. So applying a very bright material to some small detail would be fine. Can someone confirm this?

It will do less harm, but it will still do some. It's like with caustics glass. Fine to apply on a vine glass on a table, but destructive when applied onto window glass. Small vase on a table with extreme albedo won't slow things down much, but it will still slow down things a little bit + it will reduce realism of that objects and objects surrounding it. So while it won't cause disaster, there's no point of doing that.