From where the pinkish color comes from ?
And why diffuse_albedo has falloff ?
Eventually, if there's an albedo channel ... shouldn't be [also a] direct and indirect _raw_ AOVs (ie. no colors applied there) to be able to do something with that albedo later in comp ?
(Settings are those in the screenshots plus a simple arealight on top)
If you change the Base Color, it should give you the correct results.
I think these are plain bugs. albedo and diffuse_albedo are inconsistent. and depending on what color is the base color other weird things can happen.
for example :
more a kind of bug report than a real question.
I have same problem on mtoa 2.0.2.3, but not only with specular, I have bug with clearcoat color...
yep it's everywhere.. even on the arnold maya blinn/phong etc.. not only AiStandardSurface.
However implementing our own materials .. it doesn't look like it's from the SDK or internal arnold AOVs buffers as here it works correctly.. it just looks arnold materials are a bit over-tweaked. Can't remember right now but I think there's also an inconsistency with metalness where diffuse/specular albedos are kinda switched.
Eventually, as I already said I can't see the point in providing albedos if then you don't provide build-in raw AOVs (that's a request 🙂