Re: stellar subtraction method -- _really_ hard

[aah@nofs.navy.mil]

Michael R. mentioned that the Alard & Lupton method is difficult, and
I agree.  One of the drawbacks is that you basically have to have the
raw pixels for all images of a given field at your disposal, in order
to determine the reference frame and the subsequent subtracted frames
based on that reference.  They give a processing time of 1min per frame
(1024x1024, so not much different from a mark III frame) with a 200MHz
PC, which is good, but the input matrix is pretty complex.  The
results they give sure look nice!
  What I'm happy to see is that several groups are now thinking about the
problem of automated stellar photometry in really crowded fields.  I don't
think there is a consensus as to which method is the best, or even if there
is a generic 'best'.  I think that an examination of these methods for
implementation by TASS is a little premature until we can do reliable
photometry of isolated stars, but I would not turn anyone away from doing
such a study.
Arne