I am trying to create a BSDF proxy geometry for a 5 phase study using perforated shading screen. I have been following along with this tutorial for the five-phase method using venetian blinds as a practice study.
I have the following three questions. I would really appreciate if you could help me with them:
As I understand it, the “proxy geometry” refers to the 3D geometry of the shading layer of a fenestration system. It does not include the glazing surface, correct?
If the “thickness” in the BSDF material type is measured as the distance between the daylight matrix polygon on the exterior side of the window and the view matrix polygon at the interior side, then in the below image (There is a 6" thick wall with a 2" glazing flushed to the outer edge. There is a 0.25" thick opaque perforated shading screen outside the glazing) what will the BSDF thickness be? Will it be 6"?
For a point-in-time 5-phase study, if a shading system has multiple states at different hours of the day, do I need to generate BSDFs for each state corresponding to the hours being studied?
I will do my best with your questions and invite others to offer their opinions:
The “proxy geometry” should refer to the surfaces that stand in place of the actual geometry, but the documentation sometimes confuses this and calls the actual geometry used to model the system the “proxy geometry.” The semantics are less important than the concept, which is that the BSDF is attached to simplified geometry, usually one or two rectangles, and the detailed geometry is used for “direct” components (i.e., view and shadow rays).
The “thickness” parameter should enclose the entire system, whatever it consists of. If you don’t plan to alter the glazing independent from the interior geometry, then it makes sense to model these systems separately. If you always use the same glazing with a given geometric configuration, then it is more efficient to enclose them together and call this your system, with a single corresponding BSDF. The utility “pkgBSDF” can take care of the conversion details for you, and I strongly recommend it.
You need a different BSDF for each system state, unless the change is just in the glazing and not the system geometry. As I mentioned, then you would want to model the two separately.
Thank you Greg, that helps a lot!
I will look into pkgBSDF.
Quick follow-up to your response to #2:
I guess I’m still wondering what I should consider as the ‘system’ here.
I have a simple fixed glazing surface that I don’t plan to alter. And a perforated screen outside, that will have multiple states whose daylight impact I plan to study.
So, since I’m not altering the glazing, you are saying I should model the perforated screen separately and use the screen thickness for the BSDF ‘thickness’ parameter?
But then, if I use the same glazing, you are also saying it will be efficient to enclose the glazing with the screen using pkgBSDF?
Sorry, I’m not sure I’m understanding these two points correctly.
Regarding #2, I would include the glazing as part of the system for a more efficient rendering calculation. The only reason to exclude the glazing would be because you were going to vary just that, since you don’t need a BSDF to model most glazing types. By combining it with your perforated material, you avoid computing interreflections between the glazing and the shading system during rendering.
