Why it is all right to use some metallicness for dielectrics.

Here a very short tip or trick. It started with my question why satin looks so metallic while it is a dielectric. In almost all cases only a metallic can have a colored gloss/reflection. All other materials not. There are maybe more reasons why a material appears a bit metallic to the eye, but here an evidence that real geometrie has a strong inluence on the appearance of materials. In this case the second illustration appears to be a bit metallic to me. What happens is that the light is bouncing on the ridges, and turn back a colored gloss. Since we cannot expect to build all geometrie I think we are allowed to trick a bit and use a Metallicness function in our shader ( colored gloss) even for dielectrics. Which is in contradiction with the theorie that only metallics can have colored gloss/relfections.
stc=1

No Image? Check here:
http://www.pasteall.org/pic/show.php?id=104456

Here a collection of pictures I took today of dielectrics in which I think to see some colored gloss. Could be caused by a mixture of properties like micro surface, translucency, colored transparancy etc.

http://blenderartists.org/forum/asset.php?fid=403382&uid=225119&d=1466628614

No image?
Check here: http://www.pasteall.org/pic/show.php?id=104454

It seems it is not possible to imitate this effect with a bump map; The light is not bouncing in the ridges this way. For that we need real geometrie: see illustration:

http://blenderartists.org/forum/asset.php?fid=402834&uid=225119&d=1466176209

No image ? Check here:
http://www.pasteall.org/pic/show.php?id=104455

Here some renders where I used rough colored gloss in dielectrics:

http://blenderartists.org/forum/asset.php?fid=402818&uid=225119&d=1466159236

No Image? Check:
http://www.pasteall.org/pic/show.php?id=104457

I agree with you they do look as though there is a rough colored gloss. Ive done this myself and used a rough colored gloss because I thought it looked better.

I think you’re wrong on many of these materials, except maybe silk and that gold-ish lid that looks made of a metal-look plastic. All the “tinted” speculars are just pale white that don’t reach 100%, so that you can see the diffuse (and colored) component. Many of these materials can be easily reproduced with pbr node groups without involving metals

I see the colorless reflections in the pictures ( collage), but what about a mix of the colorless pbr reflections, together with a tiny bit of colored rough gloss.

A few questions ( I hope you have some spare time for that):

  • As soon as I use a glossy node with some tinted color. Are we speaking of metallicness in that case?
  • If so are we able to make satin without a rough colored gloss? If so am am very curious about a node setup that can do that.

To get an impression, here the actual effect of the use of RCG ( rough colored gloss) in dielectrics.
The illustration forgets to mention that it makes use of RCG next to velvet, translancy,diffuse and coating.

No image? check here:
http://www.pasteall.org/pic/show.php?id=104492

Here an obvious one: On the inside a diffuse with colorless coat. On the outside a colored transparent layer ( real geometrie) with also a colorless coat. .
No metallicness used, but I see a colored gloss together with a colorless gloss. This could explain why some dielectrics appear to have colored gloss.

No Image? Check:
http://www.pasteall.org/pic/show.php?id=104495