Wednesday Night Data

[Tom Droege <tdroege@veriomail.com>]

Data Wednesday night consisted of two runs:

1) Flat.  This took 25 data frames moving forward in RA about 1 degree 
after each frame.  The first, second and every 8th frame was dark.   This 
gave 21 object frames and 4 dark frames.  Each was of 100 second 
exposure.  The darks were median combined (their mean levels were equal to 
within the noise) for a master dark, and the 21 object frames were median 
combined for a master flat.

2) Follow.  The sky was followed with 100 second exposures.  One every 189 
seconds.  There were 50 exposures taken of the same field.  Tracking error 
was a little over 200 pixels N-S and 40 pixels E-W.  This means that 90% or 
so of the stars should appear on all 50 frames.  This run covered about 2.6 
hours.  I applied the master dark and master flat to one of these 
images.  The result looked very flat.  I was about to make a quantitative 
measurement of "very flat" when we had a power outage and I lost the dark 
and flat which I had not yet saved. ;^)

I co-added two of these frames and the noise level decreased by the 
expected square root 2.  One then found about 40% more stars using the same 
settings.  Looks like one frame has 3000 or so stars at 5 sigma.  One might 
want to co-add 4 or so frames together and thus get more sensitive 
measurements and increased interval.

The stars are as round as I have seen, and consistent in shape with the 
observed drift.

If anyone cares, the frames are centered at 20h10m, 04:25N

The fits headers are close enough on these runs that one can easily 
identify star fields.

I plan to run in this mode through the winter.  This data should be useful 
for the discovery of short period variable stars.

Does anyone want Wednesday's data to start developing a pipeline?  This 
would be Data Set 19.  Unless the experts are pulling my leg, I would 
expect that something new will be found on each data set as above.  To ask 
for Data Set 19, send me a mail message with your name and address.

Tom Droege.