[Aavso-photometry] What error to report in WebObs

[BailyHill at aol.com]

Hello Brad;

Past discussion had touched on many of the points that you mention in your 
email.  There is no best way.  SN is clearly not adequate.

What I have been doing lately, is that I take 10-20 images of each field.  
These are full exposures--typically 2 minutes each--depends on the field and how 
faint the object is and how bright the comp star is.  I then reduce the data 
as a mini time series, and calculate the average and the std dev of the 10-20 
measurements.  This Std Dev represents my measurement process, including most 
all of the errors, except the error in the standard value of the comp star.  
Since we use the average of some number of runs, this average for the comp star 
should be pretty good.  

Since I submit the comp value, should a better estimate of the comp star 
become available, then my pt can be corrected in the future.  In theory, the std 
devs of the comp stars should be better than the pt that we customarily do.  
This methodology also give me confidence that I am not/or am catching a flare.  
I report all 10-20 values to the data base.  I also watch the SN calculator in 
Maxim, and I adjust the exposures, aperture radius and the chosen comp star 
to maximize the SN for both the unknown and the comp.  This process seems to 
give me the best, consistant pt.  This process was coupled with a lot of work to 
improve my darks, bias' and flats--but that is another topic.

Hope this helps--comments welcome

Gary