Return to Home Page
Re: Software problem in time-series photometry > > The centering error on faint stars that Taichi-san mentions > > in AIP4WIN can be avoided altogether by doing photometry as > > C-Offset where the comparison xtar (presumably bright > > enough to be reliably centered on all images and the > > variable and check are not centered except on the first > > image of the set and the apertures are rigidly placed in > > the same location relative to the comparison star. > > Thanks, Lew! Is everyone using this option? (In precise photometry, > I regard this option as a default -- we use multiple stars to compensate > a small-order deviation from a constant offset. With an Alt-Az mount, > one also need to incorporate image rotation). Use this option whenever > applicable. If everyone uses this option and still gets unrealistic > depressions in the light curve or other unusual features, we should search > for a different explanation -- in any case, this would provide a critical > test for the software. I noticed that this does not give sufficient explanation of the observed phenomenon (depending on the algorithm of centering, though). If object, comparison and other stars are independently centered, by using something like a center of gravity, the net systematic effect is expected to be "brighter than real". This is because pixels with higher counts (caused by random fluctuation or photon noise, CCD instrumental noise) are preferrably selected as the center of gravity. This is a commonly observed phenomenon for photometric routines which first list the detected stars and aperture photometry, and the select the matched objects. The observed phenomenon in AIP4WIN is the reverse, which can not be easily explained with independent aperture determinations using independent centroid determinations. Regards, Taichi Kato
Return to the Daisaku Nogami
vsnet-adm@kusastro.kyoto-u.ac.jp
