applying a BSDF material

Hi All,

    I'm trying to define a curved perforated lamella. Seven or eight of these lamellae (500 mm width) are stacked to form a large Venetian blind.

    I used genBSDF to generate a BSDF for the perforated sheet (2 mm thick) and the resulting xml looks quite reasonable in the BSDFviewer.

    My model for the sheet lies in the negative z-space: getbbox D3T6.rad gives

         xmin      xmax      ymin      ymax      zmin      zmax

    -2.75546e-16       120 -1.83697e-16   103.923     -2.01     -0.01

(My model is in mm.)

    How to apply the BSDF material to the curved lamella is not entirely clear to me.

    I've looked at McNeil_BSDFsandPhases-1.pdf pages 44 - 48 and following that example I ended up with:

void BSDF perforated_lamella_front

    6 3.0 D3T6.xml 0 1 0 .

    0

    0

    void BSDF perforated_lamella_back

    6 -3.0 D3T6.xml 0 1 0 .

    0

    0

    My model then consists of numerous polygons like

    perforated_lamella_front polygon perforated_lamella_front.50

    0

    0

    12

          239.80015564      2110.05224609    -0.267752468586

         221.736953735      2110.05224609      4.51743745804

         221.736953735    0.0521981567144      4.51743745804

          239.80015564    0.0521981567144    -0.267752468586

    Looking at GW2_New_BSDF.pdf I read: "Extra effort needed to attach BSDF to sphere" and suspect that this extra effort might also be required for my curved lamella, but I've no idea how.

    My rendering shows no sign of geometry where I expect at least some blur.

Where am I going wrong?

    Regards,

Wouter Beck

Hi Wouter,

Did you create a Klems or a tensor tree representation of your perforated material? Are the perforations on a rectangular or hexagonal grid?

In either case, I suspect you want your "up" vector as specified by the 3rd-5th string arguments of your "BSDF" primitive to be the Z-axis (0 0 1) rather than the Y-axis (0 1 0). The latter is more appropriate when the surface is aligned in the X-Z or Y-Z plane. Although your lamellas curve, so long as they do not curve to a vertical orientation, and +Z up vector should work for you.

The main problem, however, is that you should set your thickness to 0 and don't give front & back surfaces. Front and back surfaces are only appropriate when you are creating "proxy geometry" for the BSDF. Here, you are intending the BSDF to substitute for the real geometry, not just serve as a proxy for it.

You could of course create a full model with perforated lamellas and run it through genBSDF without this extra step. That might even be more accurate, though I appreciate the challenges with creating perforated, curved geometry using commonly available tools.

Cheers,
-Greg

···

From: ascendilex | Wouter Beck <[email protected]>
Date: September 19, 2016 3:18:02 PM PDT
Hi All,

I'm trying to define a curved perforated lamella. Seven or eight of these lamellae (500 mm width) are stacked to form a large Venetian blind.
I used genBSDF to generate a BSDF for the perforated sheet (2 mm thick) and the resulting xml looks quite reasonable in the BSDFviewer.
My model for the sheet lies in the negative z-space: getbbox D3T6.rad gives

     xmin xmax ymin ymax zmin zmax
-2.75546e-16 120 -1.83697e-16 103.923 -2.01 -0.01

(My model is in mm.)

How to apply the BSDF material to the curved lamella is not entirely clear to me.
I've looked at McNeil_BSDFsandPhases-1.pdf pages 44 - 48 and following that example I ended up with:

void BSDF perforated_lamella_front
6 3.0 D3T6.xml 0 1 0 .
0
0

void BSDF perforated_lamella_back
6 -3.0 D3T6.xml 0 1 0 .
0
0

My model then consists of numerous polygons like

perforated_lamella_front polygon perforated_lamella_front.50
0
0
12
      239.80015564 2110.05224609 -0.267752468586
     221.736953735 2110.05224609 4.51743745804
     221.736953735 0.0521981567144 4.51743745804
      239.80015564 0.0521981567144 -0.267752468586

Looking at GW2_New_BSDF.pdf I read: "Extra effort needed to attach BSDF to sphere" and suspect that this extra effort might also be required for my curved lamella, but I've no idea how.

My rendering shows no sign of geometry where I expect at least some blur.

Where am I going wrong?

Regards,

Wouter Beck