[Aavso-photometry] Sources of accuracy and precision in photometric measurements

[Radu Corlan]

> Hello Radu, from the VS front tonight,
> 
> For those interested, this is ./noao/digiphot/apphot/phot/apmeasure.x
> 
>                     fctn = max (0.0, min (1.0, aperts[k] - r))
>                     sums[k] = sums[k] + fctn * pixval
>                     areas[k] = areas[k] + fctn
> 
> For inner pixels, not on the boundary, fctn=1; for outside
> pixels, fctn=0; for border pixels it is between 0 and 1 .. an
> approximation as noted.

Yes, this is what some people call the "irregular polygon" 
approximation. 

BTW, a good way to test whether an aperture approximation is a dominant 
source of error in any case is to resample the image 2x or 3x using
nearest-neighbour (so that each pixel is replicated in a 
2x2 or 3x3 box) and of course double or triple the aperture radiuses. 
This way the intrinsic error due to sampling remains the same, but the 
error due to aperture approximation goes down dramatically. If one 
compares the scatter of data in the two cases (or the magnitude of 
errors in the case of a synthetic image) it will be easy to see if the 
aperture approximation is significant: if it is, the errors will go down 
on resampling.

Radu

> 
> Justin
> _______________________________________________
> 
> Aavso-photometry mailing list
> Aavso-photometry at mira.aavso.org
> http://www.aavso.org/mailman/listinfo/aavso-photometry

-- 
Radu Corlan      

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