[Aavso-photometry] Question about identifying ref stars and SNR

[arne]

Yenal Ogmen wrote:
> Hi,
> I have two questions needed to be answered:
> 
> 1) I have faced with a problem in identifying
> magnitudes of reference stars of the same magnitude in
> CCD box. For example RR Peg chart has two 113
> reference stars, and in the CCD Box there are 11.303
> and 11.321. How can I know which star's magnitude is
> which one?
>
The CCD box has been removed from the current VSP release
for many reasons including this confusion of identity issue.
The photometry table includes RA and DEC for all stars, so
you can manually figure out which star is which.  We intend
to include an optional field of radial distance from the
field center as a further aide.  You can get the photometry
table by using the latest version of VSP, or by clicking
the photometry table button on the entry form.

> 2) I am planning to reduce exposure time and number of
> stacked frames to have more time for observing more
> variables. So this will pull my SNR down. Therefore,
> what is the lowest value for SNR which photometry can
> be done accurately?
> 
That is a tough question.  My usual rule of thumb is that I like
to get at least SNR of 100 on every star I submit.  This rule
is tempered by the available time - if it takes too long to
get a SNR=100 observation, then I compromise and accept poorer
signal/noise.  Now, what is "too long"?  What is acceptable
"poorer signal/noise"?  These answers depend on the
project and what else I want to do that evening.  If I am
monitoring CVs looking for outbursts, then all I really need
is to ensure the star is near quiescence, which is usually very
faint.  If I integrate for hours to get that measure, then I can't
monitor other stars.  So I might accept a good detection of the
star in quiescence, perhaps SNR of 10, and then move on.  Or, I may
find that stacking images is a real pain, and so I only take single
images for photometry, in which case the exposure time limits the
SNR that I can achieve.  So there is no single best answer to
this question.  I would say that you should try to do at least
SNR=20 for any observation if at all possible, as in this manner
you will be providing photometry that is more precise than visual
observers.  I would also recommend at least SNR=50 for any low-amplitude
variable.  If either of these two guidelines takes too long with
your telescope on a given object, then drop that object from
your program - there are lots of variable stars in the sky, so
limit yourself to those you can do well.
Arne