I noticed that the Peak Extraction (PE) was not activated when I used it in the Daylight-Coefficient (DC) method with one ambience bounce. Could you please confirm or correct my observations (below)? Many thanks!
PE appropriately models specular transmission of a fenestration fenestration that has a view component, only when the sun is modeled as a solar disc rather than a direct sky patch.
PE combined with the DC method with one ambience bounce cannot appropriately model the solar disc.
If I correctly understand your understanding, your understanding is correct.
Peak Extraction happens if all the following are true:
The “aBSDF” material type is applied.
The XML file used by aBSDF has a sufficiently pronounced peak in the “through” (view) direction for the point of concern.
The point of concern is in the direct beam of one or more light sources, or is a view ray looking directly at the surface.
The DC method generally fails with #3, since the sky patches are typically represented as part of a distant glow surface, so are neither a light source nor a direct view.
You are welcome. I should add that while peak extraction may not be triggered in the DC method, the results will be no less accurate using the “aBSDF” type versus “BSDF”, and should give the same answer.