Colored glass prisms

[alberto_royo]

Hello everybody.
We have a problem trying to generate an image that a client proposes to us. Does anyone think about how to solve it?
I enclose the reference photograph, a capture of our scene and an attempt (unfortunate).
Thank you very much in advance,

[jrgby]

Pretty sure that's not glass but cast resin with a tilted pour - see right block with diagonal line running through, and also the slightly less than flat faces.
The tilted pour is a kind of trick to achieve gradients in acrylic that isn't just laminating an acetate print onto one surface.

I've tried to render this effect before and I'm not sure it's totally possible - see attached image. This is just gradient map in volumetric absorption colour slot, which almost gets you there but the refactions aren't really right. Maybe good enough for what you need though?

[alberto_royo]

Wow that's great, I did not know the tilted pour technique, but I get the feeling that you're right. Thank you very much, I will try and show you the result.

[alberto_royo]

Well, inspired by what you have told me, I have prepared this test that I have sent to the client to see it ... but instead of using the gradient map, I have built it in geometry, 2 bodies one inside the other.

[romullus]

I wonder how did you get that soft transition between mediums if it's  two separate geometries? Looks pretty close to reference.

P.S. i think you should move glass cubes slightly up, it can be seen where they intersect with the ground plane.

[alberto_royo]

The truth is that I do not know where the transition comes from, you can see how I built the geometry with one body inside another, more or less how a glass with liquid is built ...

[romullus]

Did you lowered refraction glossiness on any of those materials? I think that would explain blurry transition.

[alberto_royo]

No, as you can see they are normal glasses, without anything special ... they also have an edge bump plus a noise to produce slight undulations ...

[jrgby]

Looks to me like the transition is from the angle of the camera. At the meeting of the two materials in that angle the red material is thinnest and therefore paler.

[jrgby]

Here's a sample we've got kicking around, hopefully clear from the change in angles.

If there was a nice way to do this with just material, without having to slice geometry I'd love to know!

[alberto_royo]

Really, that would be the ideal, only material, but in this case it seemed to us that it was solved better with geometry. And on the other hand, it does seem that the gradient comes from the change of point of view

[alberto_royo]

And I am very grateful to jrgby for enlightening us in a field that is totally unknown to us, thank you very much!!

[jrgby]

No problem. Actually, thank you for suggesting body within body because this is a much better method. I just tried it now and it looks true to life from all angles, which is great.
I had only thought of slicing 1 body and capping open boundries, which doesn't really work.

[maru]

I am pretty sure this could be done with just a material using the "inside volume" mapping mode, however that is possible only for the Corona Volume Mtl, which cannot have surface effects such as reflectivity or refraction...
Interesting topic! Caustics in practice!