[CCD-astrometry-photometry] RE: [Aavso-photometry] photometry variation with measurement aperture

[Radu Corlan]

On Mon, Mar 07, 2005 at 01:37:26PM -0500, arne wrote:
> Walt Cooney wrote:
> 
> >I did not include some of the other info you asked about.  The maximum 
> >pixel
> >values for the comp star and variable were around 5000-6000.  I have a 
> >16 bit
> >AP-7p so I was well short of half well depth but not rediculuously 
> >underexposed
> >either.  The net count which I presume to mean measurement aperture 
> >minus sky
> >ranged from 60,000 to 90,000 for a measurement aperture of 6 radii.
> >"Background" as MIRA calls the average sky background was around 1500.
> 
> Walt further replied that the background was "per pixel."  Since the
> peak pixel values for the comp and variable were 6000-ish, this means
> that sky becomes very important once you are more than a couple of
> pixels from the star center.  However, while a mismeasurement of sky
> would affect the results for a single image, this should be random
> over the entire image set.  Since each image has a systematic trend
> (brighter variable with smaller aperture), my guess is that the
> comparison star has a companion.  As you shrink the measuring
> aperture, the comp star companion is removed, making the comp star
> fainter and the variable effectively brighter.

Another cause i can think of is if field stars contaminate the sky 
annulus of one of the stars and the sky algorithm doesn't reject them 
properly, one of the star's sky will be systematically off (and that 
causes a aperture-dependent error). 

A "growth curve" (plot of flux vs aperture radius) would probably make 
the cause clear. If the curve has two "humps" before it levels off, the 
cause i a contaminating star in the center aperture. If the curve never
levels off but increases (or decreases) constantly after a while, the 
problem is likely with sky estimation. 

radu



>  Note, however, that Walt's light curve does improve as he moves to
> smaller apertures, indicative of less sky background and therefore
> better signal/noise. Arne
> _______________________________________________ Aavso-photometry
> mailing list Aavso-photometry at mira.aavso.org
> http://www.aavso.org/mailman/listinfo/aavso-photometry

-- 
Radu Corlan       rcorlan _at_ pcnet _dot_ ro  

   You can still escape the "Gates" of Hell!
                 Use Linux!