[Aavso-photometry] Aavso-photometry Digest, Vol 57, Issue 3 best options for averaging

[Brad Walter]

 I measure the individual images and average the results. I prefer that to
averaging the images because it avoids processing that your software may do
that you may not intend. For example when aligning images before combining
them, your software may move stars fractional pixel values and use
mathematical approximations to distribute the light in that pixel among
adjacent pixels. In some software that can be avoided by using an option to
align using whole pixel values only. If you know your software doesn't do
anything "unexpected" then averaging images and averaging measurements from
images should be equally good. 

I would average data images rather than median combine because the
statistical formulae we commonly use to evaluate the affect of combining
measurements is based on averaging values rather than taking the median
value. There is almost always a difference between the average and the
median values with slightly more noise using median combining compared to
averaging. If you are combining a large number of images such as 20 or 30
darks or flats, then there is little difference between averaging and median
combining, except that median combining removes affects of cosmic ray hits.


Message: 1
Date: Sun, 10 Aug 2008 21:04:56 +1000
From: "Terry Bohlsen" <terry.bohlsen at bigpond.com>
Subject: [Aavso-photometry] Best options for averaging files.
To: <Aavso-photometry at mira.aavso.org>
Message-ID: <001701c8fad8$eca586d0$0800000a at flatscreen>
Content-Type: text/plain;	charset="iso-8859-1"

Dear All
I am new here and would first like to say hello. 
I have commenced taking images for photometry and have been taking reading
of long period variables and analysing them with AIP4WIN new mag measuring
tool.
I have measured my transformation coefficients on a number of runs as
suggested in the Cohen paper and am happy with them.
My question is, what is the best method of combining images taken
sequentially?
Currently I take 3-5 exposures through each filter. I then have the option
of median combining the images and then measuring the mags on the combined
image or I can measure the mags on the individual images and either
submitting the individual readings or to average the readings over the 3-5
images and submitting that.
Which is preferred?
Thanks in advance.

Terry Bohlsen
Armidale
NSW
Australia