reflection values

[soulburn3d]

In general, IORs are a good way to control the reflectivity and refractivity of a material. Go to google and look for "ior tables" and you'll find many useful lists of ior values. And don't be afraid to eyeball the value either for materials you don't have, since cg is really just an approximation of the real world to begin with.

- Neil

[mister3d]

In general, IORs are a good way to control the reflectivity and refractivity of a material. Go to google and look for "ior tables" and you'll find many useful lists of ior values. And don't be afraid to eyeball the value either for materials you don't have, since cg is really just an approximation of the real world to begin with.

- Neil

Thank you, Neil!
One archviz guy told me the same - use your eyes, because if to follow only scientifically, everything (lights, camera etc) must be scientifically correct too, Maybe in Maxwell this is possible.
But I found some interesting values in one article about fresnel effect:
reflectivity for metals is somewhere not lower than 80(vague) for 0 degree face(but some people say you can ignore fresnel for metals)
most glasses and plastics have about 4% of reflectivity for 0 degree face(and strive for 100% for 90 degree)
almost nothing reflects 100%, even a mirror(it's better to use a grey color for a mirror than pure white)

The only thing I'm not sure is if you must use black for diffuse for any kind of metals?
In your tutorials you say about IOR values like 10-20, but I didn't meet such big values in tables. :)

[soulburn3d]

reflectivity for metals is somewhere not lower than 80(vague) for 0 degree face(but some people say you can ignore fresnel for metals)

Well, you shouldn't ignore fresnel for metals, you should just realize that the complex IOR value for most metals is very high, which means it's highly but not 100% reflective on the front facing angle. For example, in 3dsmax, when I make chrome, I generally place my IOR value between 10 and 20, which is almost 100% reflective on the front, but not quite.

The only thing I'm not sure is if you must use black for diffuse for any kind of metals?

If you're after that pure, super polished perfect metal, then go with a black diffuse. But if your metal needs to look impure, maybe worn a little, you can start bringing a little diffuse color into the mix. Remember, as you bring your diffuse value up, you should compensate by bringing your reflectivity down a bit, since Diffuse + Specular = Your total energy reflecting off the surface (1). Some materials like the the Brazil Advanced Material that ships with the Brazil Renderer performs this calculation for you automatically.

In your tutorials you say about IOR values like 10-20, but I didn't meet such big values in tables. :)

Ya, those values are highly unscientific :) But from a visual sense, it's doing exactly what you mention abive, it makes the front facing area highly but not perfectly reflective.

All of CG is an approximation. CG REALLY becomes an approximation when you start getting art involved. From a scientific perspective, a shader realy should be one thing, should perform things a certain way. Scientifically, many of the standard shading models we have totally miss the mark. However, assuming you want to work in the entertainment field (and that is an assumption, please ignore all this if you're more interested in another field), we need shaders that start off with a good approx of reality, and then have the controls to vary reality based on artistic concerns. If you tell a director "I can't make that more green, because that's scientifically inaccurate", you will be fired :) So my recommendation is to get a general sense of how a certain material reacts to light, know some of the physics behind that, then get a good approximation of it and eyeball the rest. There are much better shading models out there for metals, but many aren't available for the standard renderers (which is a shame), and some are not art directable, so if you understand the basics, and know what looks good, you should be fine no matter what shading model you have at your disposal, and no matter what requests you get to change stuff.

- Neil

[noizFACTORY]

Remember, as you bring your diffuse value up, you should compensate by bringing your reflectivity down a bit, since Diffuse + Specular = Your total energy reflecting off the surface (1).
- Neil

Hi Neil,

Does that mean in Maya, if my diffuse is set to 0.8, then my eccentricity and specular roll off should together be around 0.2 to bring the sum total to 1? What about the specular color and reflectivity then? Or is this just a numerical approximation to preserve the conservation of energy thingy? Was always a little confused about these things so thought would pop in the question to you over here.

-Sachin

[soulburn3d]

Does that mean in Maya, if my diffuse is set to 0.8, then my eccentricity and specular roll off should together be around 0.2 to bring the sum total to 1? What about the specular color and reflectivity then? Or is this just a numerical approximation to preserve the conservation of energy thingy? Was always a little confused about these things so thought would pop in the question to you over here.

Well, assuming you're using the maya blinn shader, eccentricity controls the size of the highlight, not it's intensity, so that can be whatever you want. Your "Specular Roll Off" and your "Specular Color" are multiplied together to give you your specular component. So assuming you keep your "Specular Color" at pure white, you'd want to set your "Specular Roll Off" to 0.2 in order to compensate for your diffuse value of 0.8.

Reflectivity makes things more complex. I'd recommend setting your reflectivity to 0.2, while this is not actually correct, it will give you results that are probably fine.

Think of reflectivity as being the actual reflection portion of your shader, and your specular component is the reflection of a point light. So basically you're taking your real world reflection and splitting it into two different things. Unfortunately, I believe that Maya's shaders are additive in nature, so if you set "Specular Roll Off" to 0.2 and reflectivity to 0.2, where the two exist at the same time, you're actually getting a reflectance value of 0.4. However, if you set the reflectivity to 0.1 and the "Specular Roll Off" to 0.1, you get a proper result where both exist, but an incorrect result anywhere where you only get one of the two reflection types. So you can never make the standard maya blinn shader totally follow the laws of conservation of energy, but if you set both to 0.2, it should be close enough for rock and roll.

- Neil

[noizFACTORY]

Thanks Neil!

That does clear things for me. And "if it looks good, its right" rule still holds good anyway. Also with render passes, life becomes much easier.

-Sachin

[mister3d]

Neil, thanks a lot, my fears now all gone. :thumbsup: I'm ready now to jump into the creative part! You have explained to me some points I didn't know, so I feel I can create now many materials without fears.
About the energy preservation, I mainly use vray, and it has a built-in energy preservation. :applause: I'm not sure about mental ray, but I guess it's also has it built-in. In vray you even have an option to choose from monochromic energy preservation and rgb. I use monochromic if I need to make my specular colored differently from diffuse(for pearls). because it's not working another way.
Ok, now I'm going to collect a lot of references for different materials. Thanks again!

[soulburn3d]

About the energy preservation, I mainly use vray, and it has a built-in energy preservation. :applause: I'm not sure about mental ray, but I guess it's also has it built-in.

Brazil also uses an energy conservation model (Actually, I think it's more like vray also uses an energy conservation model, since I believe Brazil had it first :) ).

I wouldn't automatically assume that a renderer uses energy conservation though, max's standard material doesn't, maya's standard material doesn't. You may want to ask Jeff Patton, the local max mentalray expert, he may know if Mentalray's Arch Design material uses energy conservation or not.

- Neil

[playmesumch00ns]

Haha I think mental ray had it before all of them (DGS).

I wouldn't worry too much about energy conservation if you're using a blinn!

[soulburn3d]

Haha I think mental ray had it before all of them (DGS).

Well, on the 3dsmax side, ghost/brazil was out before mentalray for max was I believe. Although maybe you're talking about mentalray for softimage and xsi, in which case, it's quite possible mentalray had it first. Not that it matters in the grand scheme of things, as long as everyone has it now :) Good to know Mentalray does conservation of energy.

I wouldn't worry too much about energy conservation if you're using a blinn!

I disagree, if using Blinn, it's even more important to consider conservation of energy, because the shading model is so simple, to get a realistic result, you need to have a strong understanding of what makes a material look right, since the shader isn't doing any of that work for you.

- Neil

[playmesumch00ns]

What I mean is that given that most blinn implementations don't conserve energy when you change the roughness parameter, and the fresnel hack in there is so filthy, worrying about conservation of energy between surface layers seems a little like overkill :)

[soulburn3d]

What I mean is that given that most blinn implementations don't conserve energy when you change the roughness parameter, and the fresnel hack in there is so filthy, worrying about conservation of energy between surface layers seems a little like overkill :)

I understand your perspective, but still, if I understand the conservation of energy laws, I can apply them to any shading model to get more realistic results, even if I have to apply them visually because mathematically they don't quite work.

- Neil

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.