How PB is PBR ?

It might be in theory physically correct, but I’ve never seen such a reflection ever. Am I doing something wrong?

When I am standing in front of a painted door, or the kitchen doors, or many other “standard” materials in real life that have a sharper glossy, I can see myself, or at least the contour of myself, or I see the environment at the back of me at facing angle.

Why does this pbr doesn’t show ANY reflections at facing angle. I think it’s not correct. I find also the gloss at grazing angle way to much.

I don’t know, I start to get the idea that the pbr happening is a bit over the top, and everybody is following the hype. I hope I am doing something wrong here in the node setup, because I would love to have a working PBR node setup.


What is that value of specular 0.033. in Cynicat pro’s nodes. Is that an artistic value or PBR ? If I use that then it looks OK.

I’m not sure what 0.033 is, but I suspect it is the minimum reflection happening at facing angles. It’s physically based, not physically correct. If I wrote it as PCR rather than PBR, I would assign the C to “convenient” rather than “correct” :). Some (most?) of us realize that fresnel is a more correct approach to blending, but for the sake of convenience we use facing^5 instead as it returns math that is far easier to control (exception being refractive materials which needs proper fresnel to blend correctly). The resulting math output is then also easy to convert blend reduction due roughness, I have no idea how I’d do this using proper fresnel.

The main idea of physical based is to prevent user from doing anything stupid (especially true for metallic workflow) by forcing energy conservation, and have a very good ratio between used texture channels and resulting flexibility - maximum control for minimum amount of channels (memory).

There are tons of things PBR approaches can’t handle even remotely, but it will normally be able to handle most of the surfaces in a scene - especially scenes in games where the costly stuff won’t be done at all or at least extremely faked. Also worth mentioning that one PBRs approach may not work in parallel with another, so i.e. you may not be able to use cheat sheets from one application to the next. (also check the links at the bottom of that page).

Ah, that changes my point of view.
I was indeed hunting for physically correct rendering, but I realize it is an odd idea, and not possible.

Ok, physically based. Ofcourse.
Btw, I forgot that materials have their base reflectivities, and indeed that is what the 0.033 ( adjustable) is for. It looks very odd without it.

Thanks for helping me out of the dream that I could do PCR where C stands for correct.