[Aavso-photometry] Averaging images for optimal photometry

[gianlucaros at gmail.com]

I have read that combining (averaging) multiple images of the same
variable imporves SNR. As this operation is carried out automatically
by many softwares which use different algorithm I wonder whether there
is a standard procedure in averaging images, or a preferred software,
for photometry. My concern is that many softwares actually rotate and
deformate images to get the stacking. Is there any risk for photometry
or averaging just increases SNR and thus is preferable? Concerning CCD
errors is the formula 2.5*Log (1+1/SNR) acceptable? Reading the AAVSO
manual there is a table (4.6.1) which states the absolute error as a
function of SNR which gives different results. Can you explain the
difference between the two methods and which is the best?
Thank you
Gianluca (RGN)