I am trying to perform NSC tolerance analysis on a freeform illumination system that creates a discrete illumination pattern. I want to use the spatial uniformity of each discrete part of the pattern as the criterion for tolerancing. I am using an NSDD operand (#Pix=-4 and Data=1) targeting the standard deviation of non-zero pixels to 0.

However, I have encountered an issue where dark pixels do not all have a value of 0. To address this, I am seeking a method to introduce a threshold that filters out these non-zero, yet small values on the detector. The objective is to exclude these values from the NSDD calculation. How can I effectively implement such a threshold?

From what I understood, the Minimum Absolute/Relative Ray Intensity setting sets a threshold on the rays to be traced. But I would like to set a threshold on the detector and not on the raytracing.