Beware of DS23

[Tom Droege <tdroege2@earthlink.net>]

I think I have caught up on the first round of data processing.  I have run 
about 320 CDs through the Michael Richmond pipeline.  There are another 
couple of hundred but they all have various problems with the 
date/time/position in the fits header.  You all may remember the time I 
discovered that I was faithfully setting my clock to one second but it was 
18 days off.

OK, this brings me to the subject.  Files earlier than Mhra2251503.cal are 
suspect as to date/time.  All the times are good to the nearest second, but 
there are multi-day and hour errors in earlier data.  My notes are good 
enough that much of the data can be straightened out if there is some 
reason to reduce data, but it will be a painful process.

So beware, if you are trying to match up time curves from before and after 
Dec. 8, 2001.

I am also confused by the day.  After Dec. 3, 2001 I have been 
consistent.  The date and time have been set into the computer correctly 
and were good to 1 second at the start of the run.  But parts of the data 
seem to indicate that the date may be +/- up to 2 days off.

As a sanity check, right now my wrist watch reading Chicago DST time reads 
18:51.  The computer which is set to GMT Casablanca , Monrovia is reading 
11:51 PM  So 5 hours difference.  I think this is right for DST?
If this is correct, then the time of exposure has been good to a few 
seconds since Dec. 3, 2001.

Let me give an example of the confusion:

I took a data run on the evening of June 1, 2002, local time.

The date the shutter first opened is June 2, 2002  04:07:04 in the fits 
header.  This is about right.  That is I started the run late in the 
evening on the first.

The file name was recorded as hvra2426671.fits

Now the JD calendar that I printed from the AAVSO gives the JD for June 2 
2002 as 2428

Can someone that has been through this check to see if it is correct and if 
not, what needs to be changed?

Note to Rob, I think part of this is your doing?  I know I do a JD 
calculation that I use to figure out what RA to put in the fits header for 
CRVAL2.  I just put in a constant to get it to come out right to give the 
correct RA for the field center.  Where do you get the number you use to 
make up the file name?

Hmmm!  Seems to me that if we have the pointing correct (Michael's pipeline 
finding star positions makes sure of that) and the time of exposure start 
is OK, then we have correctly tagged the data point.  Am I right?  The file 
label may be wrong, but that does not enter into anything of importance as 
long as the log book shows when the data was really taken.  (It does)

Tom Droege