Aperture and annulus selection: does target's need to be same as the comp's?

For simple photometry, you need to keep the measuring apertures the same.  You can change the sky annulus as needed.  This is because a star has a gaussian-like profile that doesn't end at fwhm - it keeps going.  As you increase aperture size, you get the curve of growth plot that is shown on VPHOT.  So if you use a small aperture on a comp, it will have a smaller fraction of its light included within the measuring aperture, and so will appear fainter, which would push the target star brighter.  I'd show the plots except attachments are not permitted.

There are some tricks that can be played if really necessary.  You can calculate that curve of growth, and then apply an "aperture correction" so that the percentage of encircled flux is the same.  You can use the smaller aperture for both target and comp.  You can use an even larger aperture that includes comp and contaminating star, and use it for the target as well.  You can choose another comp.  You can do psf-fitting.

Arne

     There is another "middle way" not mentioned by Arne.  That is to use a smaller measuring aperture for the crowded target, but "always and forever" measure the target that way.  Yes, the zero-point will be offset from the ordinary set-up for the reasons Arne explains, but one can determine this offset by measuring the comps in both set-ups (many times on many nights to get it right fro the particular system).  You or others working with larger image-scale (or when comparing with the literature) might be able to separate things better, so it is useful to have this offset.  At least in the case where one is just looking for the relative changes, this is acceptable.

\Brian