Announcement

Collapse
No announcement yet.

Metallic material IOR

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Metallic material IOR

    Hi,

    I ask me for what can be used the IOR parameter of the metallic material? At the help page I find not enough information about the IOR. How doe's the IOR works at this material type? Should at IOR 25 the material not a red metal?

    Best -
    Micha

    www.simulacrum.de ... visualization for designer and architects

  • #2
    Originally posted by Micha View Post
    Hi,

    I ask me for what can be used the IOR parameter of the metallic material? At the help page I find not enough information about the IOR. How doe's the IOR works at this material type? Should at IOR 25 the material not a red metal?

    Best -
    Micha
    Hi Miicha, read this to get started.

    https://www.chaosgroup.com/blog/understanding-metalness

    Comment


    • #3
      Hi @Grigio18,

      this side is interesting, also the IOR table at the end. I thought the IOR of water is 1.3, of plastic around 1.5 and diamonds around 2. Now I see Gold has an IOR of 1.35. At classic shaders a very high IOR is needed for metals, like 15..25. I'm confused. Do you know why at the BRDF materials high IOR values are needed and here low values?
      www.simulacrum.de ... visualization for designer and architects

      Comment


      • #4
        Originally posted by Micha View Post
        Hi @Grigio18,

        this side is interesting, also the IOR table at the end. I thought the IOR of water is 1.3, of plastic around 1.5 and diamonds around 2. Now I see Gold has an IOR of 1.35. At classic shaders a very high IOR is needed for metals, like 15..25. I'm confused. Do you know why at the BRDF materials high IOR values are needed and here low values?
        AFAIK, the Ior value we have in brdf materials works perfectly only for dielectrics.
        For metals we used to increase IOR just to raise the min reflection value, but there are some side effects and the results are not physically correct.
        The point is that to represent metals reflections we need not only the IOR value but also the K value (extinction coefficient).

        In max there is a plug in to correctly represent metals IOR: https://www.sigershop.eu/free-3ds-ma...omplexfresnel/

        The metalness workflow is probably a little more accurate describing metal materials but it's there just to make our life easier working with textures from substance designer and in general any other software that uses that kind of texture to generate a shader.

        Comment


        • #5
          The metalness workflow is probably a little more accurate describing metal materials but it's there just to make our life easier working with textures from substance designer and in general any other software that uses that kind of texture to generate a shader.
          This is not accurate. The metalness workflow is not purely for ease of use; we have actually implemented the proper microfacet complex Fresnel effect, which is more accurate than the Siger texture. In this table here I am comparing the V-Ray metalness ("V-Ray error" in the table) vs the ComplexFresnel texture ("Ole error" in the table) vs the physically accurate complex Fresnel reflectance:

          https://github.com/vkoylazov/metalne...al_presets.csv

          As you can see, the V-Ray error is smaller, which means that it is closer to the actual physical effect.

          Here is also a graph to show the reflectance based on angle; the solid line is the V-Ray metalness, the long dashed line is the actual physically accurate reflectance, and the dotted line is the ComplexFresnel texture. As you can see the ComplexFresnel graph is way different from what it actually needs to be:


          In short, do not use the ComplexFresnel texture, use the V-Ray metalness.

          Best regards,
          Vlado
          Last edited by vlado; 06-08-2019, 03:09 PM.
          I only act like I know everything, Rogers.

          Comment


          • #6
            Thanks vlado, good to know!
            I was wrong with the adjective, I should have written not a little but a lot more accurate!

            Comment

            Working...
            X