Recovering the reductions from the "obj" stage
Midway through the reductions, we revised our procedures to those described
earlier. Unfortunately this meant recovering the partially reduced
objXXX.fits (ccdproc'd, pupil removed, and revised-WCSed) images from
tape, and re-reducing. Hahahaha. These are the steps necessary.
- Fix the bad pixels (fixem.cl) 10 min (total)/image
(I/O bound)
- cp /data0/massey/Calib_Oct/bpkpno*.pl.
- hedit obj*.fits[1] BPM bpkpno1.pl up+ ver- show+, etc.
- hedit obj*.fits[1] SATURATE 39000, etc
- ccdproc obj*.fits [generates new bad pixels masks with saturated stars marked]:
1.9 min/image
- msccmd "fixpix $input BPM" input=@all [actually fixes the bad pixels]
7.5 minutes/image
- Check the wcs for all images:
msccmatch @all M31F2.usno update- interact- fit- verbose+ accept-
3.8 min/image
- mscimage --->fobj 37min/image (CPU-bound)
- fobj--->ffobj using "imexpall.cl" which uses the bad pixel masks to
set interpolated data to 50000
imexpr "b!=0 ? 50000 : a" outimage inimage inmask
3 min/image
- Prepare to stack:
- imexamine one fobj from each filter, and record fwhm
- implot one fobj from each filter, and record rms
- philmatch with fwhm and sigbk specified
- stack and shift with stackall.cl (mscstack and philtrans.cl)