id	summary	reporter	owner	description	type	status	priority	milestone	component	version	severity	resolution	keywords	cc
1297	x_psf_sig and y_psf_sig	eschlafly@…	eugene	As of the July 2009 MD data release, x_psf_sig and y_psf_sig look badly calculated for faint objects.  A histogram of x_psf_sig vs psf_inst_mag, for instance, reveals a discontinuity in x_psf_sig, where suddenly errors get a hundred times worse.  I mentioned this a while ago and think a fix had already happened or was in the works.  However, I got bitten by it today and figured I'd make a ticket...	defect	new	normal		psphot		normal