Is this parameter an actual Abbe Number parameter? That is, does this parameter represent the actual Abbe Number range between 1 and 100? If I have a physically accurate Abbe Number value for a glass material, and I put this value into this slot, will the results be physically accurate? Or, ... does this parameter represent more of an approximate slider for some relative degree of dispersion?
>If I have a physically accurate Abbe Number value for a glass material, and I put this value into this slot, will the results be physically accurate?
I would hope so. Have you tried it? Does it look right?
From the docs:
"Specifies the Abbe number of the material, which describes how much the index of refraction varies across wavelengths. For glass and diamonds, this is typically in the range of 10 to 70, with lower numbers giving more dispersion. "
Yes, before posting this question, I experimented with it as systematically as I could (after reading the available documentation you've cited above). I set up a "measured" light stage affair with checkerboard plane behind a sphere - all with regular metric sizes and standard camera settings. Worked out the scale settings of everything to make the testing rig as sensitive to slight changes in the IOR values as possible. And then started a series of systematic experiments on the Abbe Numbers parameter, holding everything else in the Transmission settings constant, of course. But, the Abbe parameter seems relatively "insensitive" - that is, it takes rather large changes in the value to develop a discernible change in the amount of visible dispersion in the edges of the checkerboard blocks. (And, as you can imagine, the statement in the documentation above that "for glass and diamonds, this is typically in the range of 10 to 70," when the total scale range is only 1-100 to begin with, doesn't really tell me if we are dealing with the classical Abbe Number scale of 1-100.) I'm not meaning to be critical, rude or disparaging. Just trying to ask for a little bit more information.
To get a match for a given material you would:
specular_IOR
to that, There could be other issues at play here, like flipped surface normals, self intersecting geo, thin shells instead of closed mesh, etc.
Thank you Lee. I'll give that a try and return with the results. Am coming down with some flu, so it might be a day or two before I can reply to your kind answer.