RE: Focus indication

[<aah@nofs.navy.mil>]

To some extent it depends on the need for the fwhm calculation.
If it is just to focus the system, you don't need to be careful,
you just need reasonably good relative values.  The simple
scheme Michael suggested works fine for that.  The radial
profile Chris mentioned is often used for _interactive_ focus measures,
as the fhwm is obvious, the curve width tells you the symmetry of
the focus, and any outlying stars show up as bumps in the
wings.  The more complete gaussian fits mentioned by Andrew
and Doug are overkill for quicklook processing IMHO.
  My impression from Tom is that he wanted a measurement
with which he could tell when something changed; that is, engineering
data.  I'd keep it simple.  Tools such as iraf, PCVISTA and AIP have
more accurate methods such as curve fitting that you can
utilize when analyzing the final dataset.
Arne