Reflective limitation

[eungen]

I do not know how I could "extend" reflection inside the cylinder. It is a natural effect?

[maru]

I don't see a reason why this shouldn't work, but I didn't manage to get it working. Even with MSI set to 0. Strange, reflective caustics should be working fine.

[eungen]

I used Corona Volume Material to produce the fog. It should have used something else?

[maru]

No, it seems ok (except for some minor errors like the intersecting polys).

[eungen]

After set MSI to 0, PT-PT, it looks promising... Thank you, Maru!

[maru]

That's strange, I didn't manage to get it working even like this. :D

[cecofuli]

As we know, caustics in Corona 1.4 are very ugly and bad. 
You can try with PPM (photon mapping), but I'm not sure if  volume material is supported.
Eugen, you can render in PT+PT MSI= 0 for 10 hours, but you will never have good caustics.

[FrostKiwi]

We are scratching of what is possible with the current fast model of light transport.
Results like these: https://forum.corona-renderer.com/index.php/topic,2450.msg17928.html#msg17928
Are simply not possible in an efficient manner for the heavily optimized structure of Corona. You may try BiDir option for that, but it is no where near production ready in terms of efficiently averaging out the incredible intensity of some rays. Fireflies galore. Instead of prioritising the first direct bounce and the scatter from the volume, it samples every type of ray equally, which is why the fourth bounce of light in the cylinder will be just as noisy, as the first.
Chaosgroup has proposed and tested a more or less efficient manner for these types of reflections. Ondra mentioned looking into it aswell.
Maxwell Render has been known to do this rather well, which is why it is used in so many production shots of Rolex, diamonds etc. Although it is no where near as fast for interiors as Vray and Corona. Right tool for the right job I suppose.

[eungen]

OK, thanks for details. I will try Indigo Render and Maxwell for this particular situation :-/

[Juraj]

Is anyone using 3.3 Vray in production here ? It was supposed to do reflective GI caustics rather well.... does it ? (With regular BF/LC)