Eratta Data Set 20

[Tom Droege <tdroege@veriomail.com>]

Sigh!  The only way to take data is with simultaneous data analysis.  That 
way one eventually gets the labeling right.

First Jean-Claude Pelle pointed out that the some of the follow runs have 
'object' frames labeled 'dark'.  This was fixed for runs starting after 011103.

Michael Richmond noticed that DS 20 runs 6,7, and 8 have the V and I data 
reversed.  This is true for the reasons I gave in a previous post.  I have 
checked all the other disks in Data Set 20 and they look OK.  But I did not 
check every image.  Not that you can identify the V images by a dark spot 
at 1032x, 1724y on the raw image.

I in looking up the data disks to check them I noticed that the log book 
did not match the fits date-times.
Turning on the computer I noticed the clock set to December 20.  I 
panicked.  Time to finish my Christmas shopping.  A check with the GPs 
showed that the computer clock was either 17 or 18 days fast.  I have been 
dutifully setting it to within 1 second by GPS but apparently not notice 
the date.

Disks  		Taken		Fits		JDay
1,2		011116		011201		not important
3,4,5		011017		011018		2199		This is correct
6,7,8		011109		011127		2239		+18 Days
9,10		011117		011204		2246		+17 Days

The "taken" date is the date of the evening started.  Usually the UT is the 
next day so the entry for disks 3,4,5 appears to be correct.  Those of you 
that can compute Julian Day can check if the program is doing it 
correctly.  The next two runs were off  by either 17 or 18 days.  Something 
is funny in the code, or the Computer is losing the date when I turn it on 
and off.  The Fits date is what is shown in the fits header.  The JDay is 
the file label.  Fortunately I seem to be logging enough information to 
straighten this all out.  Note for these runs I have been setting the time 
to within a second or so by GPS.

Tom Droege