Re: changes to TN 0075

[Stupendous Man <richmond@a188-l009.rit.edu>]

  Okay, I've incorporated the latest version.

  Great work, by the way!

  I'd guess that the wierd pattern of standard deviations in
the crowded region is due to source confusion: when there are 
close pairs of similar brightness, the software may break
them into TWO sources in some images, but leave them as ONE
source in others.  That will cause (one of) the two stars
to appear brighter than usual in some images.  How much 
brighter?  If the two stars are exactly the same magnitude,
the merged source will appear 0.75 magnitudes brighter.

  In your plots of standard deviation vs. magnitude, we would
see an increased standard deviation for these stars which are
sometimes merged, sometimes not.  If exactly half of the
observations of two identically bright stars were merged, the
stdev would be about 0.4 magnitudes.  If one quarter of the 
observations were merged, the stdev would be about 0.3 magnitudes.

  Now, if the two stars are NOT identically brighter, then
the increased brightness of a merged pair would be less than 
0.75 mag ... and the stdev of a number of sometimes-merged,
sometimes-not would likewise be smaller than 0.4 mag.

  The ridge line of the "large stdev" locus is about 0.15 mag.
My guess is that it's due to occasional merging of stars with
somewhat fainter companions.

                                         Michael Richmond