Quote - That makes sense, but how do we know what values are "right?" 

Oy - this will take days to answer, but I'll try to give you the 10,000 foot level view.

First, you have to choose the right shading model. Then you have to tune it correctly. One way is to ask me, and you'll get the exact right answer, within the limits of what Poser can do. Sometimes I'll say there is no decent solution, but most times I have one. This is often how people have recently gotten good glass, water, leather, suede, gold, silver, copper, etc. for Poser. And this is how some have gotten decent  (not good) skin. I can only do decent with Poser as is.

Other ways can be almost as effective, or horribly pointless, depending on:

a) How good are you at recognizing what you see in real life or in real photos - i.e. are your observation skills actually good enough to isolate realistic from fake, and be able to say why?

b) How much do you know about the parameters - what effect will happen if you raise or lower each in isolation - and in combinations, how do they alter each other's effects?

In general, people guess, render, and refine. If they don't understand what the shader parameters do, the refine is usually just another guess. If they do understand, they observe the differences, and make a parameter adjustment that moves hopefully closer to realism.

Certain shading problems don't require any experiments at all. For example, the Fresnel reflection coefficient for a dialectric material (such as glass, or car paint) are perfectly given by a fixed, well known equation. Such equations are usually built into the shading system of a renderer, but unfortunately this is not the case in Poser. I have published shaders that replicate this in 26 nodes, but people get wiggy about that. I don't know why. If it was built in, but inside it really did require 26 lines of code, why would you care? So when I give you a glass shader that does the same thing in Poser, I have to give it to you in node form, and unfortunately you have the totally unnecessary but inescapable opportunity to see the guts and react with fear and loathing. Ignore the guts. Why do you care?

Now there are some shading problems that are so popular, that there are published sources with detailed physical values you can just look up and plug in.

For example, for SSS, there have been some decent actual measurements made and widely published as to the reflection, scattering, and extinction coefficients required in a dipole diffusion model approximation of skin, apples, potatoes, ketchup, various forms of milk, wax, etc. If you have access to a shader system that accepts these parameters, you can just type them in from published sources and you won't need to experiment at all.

As to Robyn's shader, either she made a mistake, or something's gone wrong when you switched textures, because that isn't even slightly realistic. There is no specular effect at all. That's the biggest problem at the moment. There are others, but that's like 80% of the fakeness right now.

---

Renderosity forum reply notifications are wonky. If I read a follow-up in a thread, but I don't myself reply, then notifications no longer happen AT ALL on that thread. So if I seem to be ignoring a question, that's why. (Updated September 23, 2019)