Andrey noted that NSI uses wrong algorithm for standing wave averaging.
There was general agreement that we want to process the raw data ourselves.
ACTION ITEM: Josh will find out if we can get the raw data out of the NSI software.
Andrey recommends that we increased padding of data before FFT until the change in resulting efficiency value is < 0.2%.
There was general agreement that 1% accuracy in measured efficiencies was a good value.
Richard: I think using X,Y,Z offsets to determine beam squint is fine, with the alternative being Bernard's approach in the farfield. Andrey and Richard fit for minimum phase error. Bernard tried that, but if there are several turns of phase across the aperture, then the algorithm gets lost. In the farfield, you can simply find the peak or barycenter. Richard: But you must define the procedure.