Announcement

Collapse
No announcement yet.

abrupt reflection falloff with Fresnel on Vray Mtrl

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

  • abrupt reflection falloff with Fresnel on Vray Mtrl

    Hi there,
    I have a problem with a pretty simple scene. A cube with a glossy VRay Material (Fresnel is on) is lit by some VRay Rectangular Lights from underneath.
    The problem is that the highlights on the bottom of the cube suddenly come to a stop so a hard edge occurs. The position of the edge depends on the IOR Value of the Material. When I disable the fresnel effect the hard edge disappears.
    As far as I understand the reflection should continually attenuate as the facing ratio gets bigger but an abrupt stop is not normal, is it?

    I already changed the geometry and the material, the problem persists.
    In case thats important : The scene I am working in has quite small measurements the vray Rect Lights are scaled by the factor 0f 0.15

    Thanks,
    Matthias



    Click image for larger version

Name:	fresnel.jpg
Views:	1
Size:	23.9 KB
ID:	873123

  • #2
    Try playing with the cutoff parameter in your vray material in the options section - if the contribution from the reflection gets quite low, vray just kills it off as part of it's early termination scheme to speed things up. The default of .01 is probably too low in your case.

    Comment


    • #3
      It might also happen sometimes if you are looking at the back side of the surface.

      Best regards,
      Vlado
      I only act like I know everything, Rogers.

      Comment


      • #4
        Thanks guys. Anyhow - changing the cutoff threshold does not change anything and the normals of my cube are pointing outwards, towards the camera (if that's what you mean, Vlado) .
        For now I am using Spherical Lights instead of Rectangular ones and the issues does not occur any more. Anyways, I am still wondering what's going wrong there.

        Comment


        • #5
          Is the cube a 6 face primitive? Did you try rounding the edges a bit? I've had problems with primitives because the of their infinite edges which dont exist in the real word.

          Comment


          • #6
            The actual model was a harddrive with a high polycount - I also tried a simple Cube, doesn't change anything.

            Comment


            • #7
              Are the edges of the cube smooth? Could it be that the smooth normals are too different from the actual geometry?

              Best regards,
              Vlado
              I only act like I know everything, Rogers.

              Comment

              Working...
              X