Object inside glass globe behaving odd?

[GilmourMotionDesign]

I'll start saying I'm not very experience with physically based renders, so maybe this is normal, if that's the case please let me know haha

I'm experimenting with a snow globe and depending on the refractive index on the reflection channel, the highlights on a glossy object inside the globe will disappear.
The globe was created with a simple sphere object and placing a standard corona texture on it, with only reflection and refraction enabled.
There is 1 large area light directly above the globe.

I attached screenshots of what I'm seeing.  Is this normal behavior?
I also attached my C4D project file if anyone is interested in taking a look.

Any help would be appreciated, thanks!

[spencerp92]

I'm wondering if it is because of the way you have it modeled now, the entire thing is a solid within a solid. A real snow globe's glass would have a thickness to it, and then whatever was inside of it would be free floating or suspended in liquid.

I'd say give it a try with the globe having a thickness to it as opposed to a solid sphere and see if that fixes your problem.

[maru]

I am not sure, but guessig this could be the "dark highlights behind solid glass" issue, which is already fixed in 1.5:
https://coronarenderer.freshdesk.com/support/solutions/articles/5000516180-known-bugs-and-limitations
One solution is to use thin glass mode. Another one is using rayswitch (both solutions are explained in that guide).

[GilmourMotionDesign]

I'm wondering if it is because of the way you have it modeled now, the entire thing is a solid within a solid. A real snow globe's glass would have a thickness to it, and then whatever was inside of it would be free floating or suspended in liquid.

I'd say give it a try with the globe having a thickness to it as opposed to a solid sphere and see if that fixes your problem.

I thought that too, but when I do that, there is no refraction at all, no matter what settings I use. See attached picture and project for example of results... and again, maybe I'm doing something wrong?

I am not sure, but guessig this could be the "dark highlights behind solid glass" issue, which is already fixed in 1.5:
https://coronarenderer.freshdesk.com/support/solutions/articles/5000516180-known-bugs-and-limitations
One solution is to use thin glass mode. Another one is using rayswitch (both solutions are explained in that guide).

I'm actually on C4D Alpha 5, so maybe that's something coming later?  One of the Corona guys put on a demo a few days ago and mentioned C4D Alpha 6 is due any day now.

[Ondra]

it may be actually physically correct that the reflections disappear, since what actually matters is the IOR ratio of 2 surfaces, not any single IOR - for most materials in air the IOR ratio is mtl ior / air IOR, where air IOR is 1.003 (+-), which is basically 1 - so you dont need to divide. But as soon as you have something inside water, or glass, 1.52 / 1.33 or 1.33 / 1.52 is quite different from 1.52 or 1.33, respectively.

too tired to word it better, perhaps somebody will understand it :D