BRDF algorithm like in mental rays mia_material

[Wobi]

Hi there,

Currently there are the three different BRDF models: Phong, Blinn and Ward in the VrayMtl. However, there is an advanced algorithm which is used in the mia_material (Arch&Design) shader from metal ray. As far as I understand it, this algorithm is kind of an advanced model for rendering reflections which is more realistic. So I suggest to build this advanced BRDF-Model into the VrayMtl.

[instinct]

Judging from what i saw it is not really a completely different brdf, but rather controlling reflectivity in a very awkward way. You should be able to yield the same result by using a gradient set to linear and use an outputmap's curve to adjust the curve to your liking. With a lot more control.
That is if we are talking about this:

http://mentalraytips.blogspot.com/2007/10/making-better-metal-with-miamaterial.html

[Wobi]

Nope. I'm pretty sure that this is a complete different algorithm from those that are build in the VrayMtl. It is the way reflections are depending on the angle. I'm not an scientist. But what I can tell is that it is impossible to perfectly match the reflections of the mia_material with a vraymtl. I recently had a job where I had to rebuild a scene which was made with vray into mental ray. And the main problem was to visually match the reflections...

[PIXELBOX_SRO]

isnt that all abou the fresnel curve and its shape?

[instinct]

That was my understanding. If not you could still use a curve for glossyness. Matching MR perfectly is something different than "more realistic" tho and i doubt that will
ever really work tbh.

Kind Regards,
Thorsten

[Wobi]

Maybe you have a look at this:

http://www.graphics.cornell.edu/pubs/2004/WLT04a.pdf

So, in the conclusion you can see that there are other models apart or Blinn, Phong and Ward which maybe even better/more realistic in most situations...

[vlado]

First, the Arch&Design specular model is based on the Ward reflection model; it is possible to find a formula that converts the glossiness values from Arch&Design materials for VRayMtl and vice-versa.

Second, for the reflection falloff, by default the Arch&Design material uses Schlick's approximation for the Fresnel term; V-Ray actually includes a hidden texture that matches that reflection falloff, which we use when we render the Arch&Design materials. This approximation allows for more artistic control, but it is not in any way more "correct".

You can expose the hidden texture by creating a MaxScript wrapper around it - extract the attached .zip file into the [maxroot]\scripts\startup folder and restart 3ds Max; you will get a new texture called MiaFalloff that you can plug into the reflection color slot of a VRayMtl material.

Best regards,
Vlado

[Wobi]

First, the Arch&Design specular model is based on the Ward reflection model; it is possible to find a formula that converts the glossiness values from Arch&Design materials for VRayMtl and vice-versa.

Second, for the reflection falloff, by default the Arch&Design material uses Schlick's approximation for the Fresnel term; V-Ray actually includes a hidden texture that matches that reflection falloff, which we use when we render the Arch&Design materials. This approximation allows for more artistic control, but it is not in any way more "correct".

You can expose the hidden texture by creating a MaxScript wrapper around it - extract the attached .zip file into the [maxroot]\scripts\startup folder and restart 3ds Max; you will get a new texture called MiaFalloff that you can plug into the reflection color slot of a VRayMtl material.

Best regards,
Vlado

My fault. Thx Vlado for making things clear. The texture map ist pretty nice to have, wonder why it is hidden though?

now, what about that formula to conert the glossines values from Arch&Design to Vray and vice-versa? Could you give this to us, since it would make things far more easy or me. I kind of made my own "transition curve" by rendering out a 100 different glossiness values from mia and vray and then compared them visually. But this does not seem very accurate in most situations, especially in the lower glossiness values...

[RobPhoboS]

Vlado, is there a similar thing we can un-hide in Maya to get access to Schlick's reflection falloff ?