Physically based materials - metal

[anchovy]

Vlado-

A quick question for you. Even though vray doesn't do internal computations using spectral data, is there a way it could do a metal material based on that BRDF/data and converted to RGB space. I haven't seen any great renders for jewelry etc for gold copper silver and the like. If im missing some docs or examples on this please let me know, I was just hunting around tonight. Say something like this...

http://www.kevinbeason.com/worklog/2...d-reflectance/

Thanks

-Michael

[vlado]

You should be able to directly plug those RGB values into the reflection slot of a VRayMtl material and get what you need. Make sure the diffuse color is black, Fresnel reflection is checked, and the IOR set to the proper IOR for the respective material.

Best regards,
Vlado

[anchovy]

Vlado-

I get that ideally you want to use the ior data to drive the reflectance but most metals ior seems to average out at a bit over 1.0 not enough to drive the reflection the way you would expect it. Certainly you can artificially up the ior or turn off the Fresnel. Any thoughts? Also does the wavelength dependency on the ior contribute substantially to the correct look or is this approximation fairly close? Im thinking the cook Torrance model would help more than something that was spectral correct.

-Michael

[joconnell]

I've always gone with the idea that any conductive material like a metal doesn't really exhibit much of the fresnel effect compared to the falloff curve of glass / plastic and so on so I generally leave fresnel off and don't use any falloff curves in the reflection value slot. There's a paper here seems to support this to an extent - http://www.graphics.cornell.edu/~wes...c/fresnel.html saying for example "For conductive materials (i.e. metals), the effect of Fresnel reflectance is subtle; for aluminum, reflectance remains above 86% for all angles"

I generally find a lot of folks that have issues with the look of shiny materials are more having issues with the scenes around the metals and their lighting - since any shiny object is so heavily influenced by the scene around it, a physically based material can always look pretty crappy if it's put into a flat, uninteresting environment from a reflections point of view. Fair enough having a physically based material removes one factor out of the equation (e.g. is it my material being wrong or is it my scene / lighting) but it mightn't be the solve all solution.

[lllab]

ior for usual metals are between 20-100, if you use one of that iors it looks automatic like metal. for best result see to get the actual ior value, but 22 is close for the usual types, i use for most.
sometimes it is worth to use more than one specular (with blend mat in max vray) with 2 different glossyness values.

cheers
stefan

[MoonDoggie]

If you can tell the difference between IOR 5.0 and IOR 15.0 on a render, then you have a better eye than me. I keep fresnel off for metals.

[JeffG2]

ping ping ping