[Aavso-photometry] FITS test results so far

[Arne Henden]

Michael Koppelman wrote:

 > It is certainly enlightening to see how these different software packages 
handle this. Clearly in most cases if images are combined the TIME-OBS keyword 
needs to be managed manually.
 >
 > I almost never combine images and then do photometry. I always do photometry 
and then average the photometry if I find I need to beat the errors down. This 
allows me to do a weighted average if the uncertainties are not equal. It also 
takes out the extra math involved with registering and rotating images if there 
is some slight change in the field over time. It's hard for me to think of a 
reason why one would stack images before doing photometry. Can anyone help me?
 >
You will find that most photometry packages start having systematic
errors when working close to sky/low signal/noise objects.  It often
helps to stack first to increase the S/N, then do the photometry.
Arne