There are a few reasons why the reported nominal criterion from the tolerancer may be different from that of the value reported elsewhere in OpticStudio:


 

1. When computing the nominal criterion in the tolerance analysis, OpticStudio considers any compensators which are defined. This means that OpticStudio optimizes for the best criterion value using the defined compensators as variables. The Merit Function or any other analysis feature in OpticStudio on the other hand, is evaluated without any compensator adjustment.
   
   
    



2. Sampling methods and sampling sizes may be different. In constructing physically significant Merit Functions for a selected criterion, OpticStudio uses Gaussian Quadrature pupil sampling. Some analysis features in OpticStudio use Rectangular Array sampling, which may result in a slightly different value than that determined using the GQ method. Theoretically, as the sampling size of both methods increases, the calculated results will approach equality. Similarly, even if the same sampling method is used, sampling size can have an effect on the outcome of a calculated value. Make sure that the sampling size is sufficient in either calculation before making a comparison of the results.
   
   
    



3. By default, OpticStudio reports a single criterion value which is the root-mean-square (RMS) across multiple wavelengths, fields and/or configurations. The results may be separated by field and configuration, but make sure that when comparing a value from a specific analysis feature vs. that of the nominal criterion, the data corresponds to a single field or the RMS across all defined fields and/or configurations.