Both codes get the total flux from circular-aperture photometry using user
determined photometry radii and sky background estimation from the median
value of a surrounding annulus.

bmac1 estimates the peak flux by fitting a 2-d gaussian to the PSF core.
bmac2 estimates the peak flux by using IDL's cubic convolution interpolator
congrid(image,/cubic)
(which is a cubic approximation to sinc/fft interpolation.)

In either case, the resulting peak is compared to a lookup table of peaks
measured with the same algorithm on rebinned simulated monochromatic data,
where the simulated data is generated with a pixellated hard-edged pupil.

(In normal use my strehl code averages the two results together for no
particular reason, but I broke them out here.)

Bruce