A first look at DS24

[Ted Woodhouse <twoodh@junction.net>]

	Hi everybody!  I've been away from the list for about two years, 
recovering from a long illness.

	I've been looking at DS24.  I haven't come up with a solution to the 
problem, but these are
a few preliminary observations which may be helpful to others working on 
the images.

	One of the dark files is bad - hvra2597478.  The mean of the ADUs is way 
out of line
compared with the other darks.  If you display the image, it looks as if 
someone has roughly torn
a piece off the top of the image.  A sticky shutter?  All the other dark 
images look good.

	There is quite a noticeable variation of mean ADUs per image as a function 
of declination.
Taking the overall mean of the ADUs for the sets of 7 images at each 
declination you get the
following:
    Dec (degrees)       V                           I
    -2                       9066
    -1.33                   9037
    -0.67                   9020
     0                       8991
     0.67                  8964
     1.33                  8940
     2                       8931
Roughly, this works out to a gradient of about 68 ADUs per degree.  There 
is a lesser trend in RA,
of about 32 ADUs per degree of RA, increasing with increasing RA.  These 
figures are for a
2032- by 2030-pixel rectangle within each image, with all the edge pixels 
removed.

         The images were dark-subtracted using a median combine
of the dark frames Tom provided on the CDs, and flat-divided using a median 
flat produced from
all 49 images for each CD.  This doesn't appear to be a CCD problem, as 
more light must
be entering the camera at lower declinations.  Must be sky brightness or 
neighbors'
lights.  I'll dig into this further and see if it relates to Tom's problem.

	Can we get a BZERO = 32768. added to the FITS headers?  It would make life 
much easier
when using CFITSIO.

	Tom, what program are you using to produce star lists, once the images 
have been flat-fielded?

					Ted