iraf.net

I use msctpeak often for tweaking the WCS on images--both using mosaics and for single images. Usually this works just fine. But occasionally I get images where the original WCS is pretty close to correct, yet when I mark and center some stars and do a fit, the residuals are enormous--hundreds of arcsec. I've seen this behavior on images from different telescopes, often when the true residuals should be 1 arcsec or less (the circles marking the stars come up bang on the stars in the image). (I just posted another unrelated message about msctpeak, the present post relates to a long-successful installation running IRAF 2.14 under Mac OS10.6)I figure there must be something amiss with the image header, since I use essentially the same parameters in msctpeak when it succeeds or fails. Here is a complete image header from an image where the fit fails, as described above:t2034[2048,4094][real]: PKS0646+06 Ha
No bad pixels, min=0., max=0. (old)
Line storage mode, physdim [2048,4094], length of user area 2673 s.u.
Created Sun 18:24:40 24-Feb-2013, Last modified Sun 18:24:39 24-Feb-2013
Pixel file "t2034_bu.fits" [ok]
EXTEND = F / File may contain extensions
ORIGIN = 'NOAO-IRAF FITS Image Kernel July 2003' / FITS file originator
DATE = '2013-02-24T23:24:40' / Date FITS file was generated
IRAF-TLM= '2013-02-24T23:24:39' / Time of last modification
PRESCAN = 20
POSTSCAN= 32
ISTART = 10
JSTART = 50
EXPTIME = 1200.000
EXPTYPE = 'object'
OBSERVAT= 'KPNO'
TELESCOP= 'Burrell Schmidt'
OBJECT = 'PKS0646+06 Ha'
OBSERVER= 'Winkler, Thariani, Slingerlend'
DATE-OBS= '26/11/98'
UTSTART = '11:07:12'
ST = 'sidereal time'
RA = '06 48 41'
DEC = '06 26 33'
EPOCH = 2000.0
AIRMASS = 99.90
INSTRUME= 'NFCCD'
DETECTOR= 'SITe 2x4K'
FILTER = '6'
GAIN = 1.58
RDNOISE = 10.1
WCSDIM = 2
LTM1_1 = -1.
LTM2_2 = 1.
WAT0_001= 'system=image'
WAT1_001= 'wtype=tan axtype=ra'
WAT2_001= 'wtype=tan axtype=dec'
TRIM = 'Feb 19 16:17 Trim data section is [28:2075,3:4096]'
OVERSCAN= 'Feb 19 16:17 Overscan section is [2082:2100,3:4096] with mean=991.85'
ZEROCOR = 'Feb 19 16:17 Zero level correction image is Zero_all'
CCDSEC = '[28:2075,3:4096]'
LTV1 = 2076.
LTV2 = -2.
CCDPROC = 'Feb 19 17:35 CCD processing done'
FLATCOR = 'Feb 19 17:35 Flat field image is Dflat_ha.fits with scale=26740.32'
RADECSYS= 'FK5 '
EQUINOX = 2000.
CTYPE1 = 'RA---TAN'
CTYPE2 = 'DEC--TAN'
CRVAL1 = 102.3375
CRVAL2 = 6.5725
CRPIX1 = 743.
CRPIX2 = 2417.
CD1_1 = -4.1111111111111E-4
CD2_1 = 1.5075936490322E-11
CD2_2 = 4.11111111111111E-4And here are a few lines from the star catalog:
# RA_HRS DEC_DEG MAG
6.76403222 6.4256848 9.077
6.799376687 6.2338303 9.185
6.799376687 6.2338303 9.185
6.850791493 6.9235634 9.19
6.850791493 6.9235634 9.19
6.772017747 5.727332 9.209
6.772017747 5.727332 9.209
6.803207633 5.8300209 9.251
6.803207633 5.8300209 9.251 Finally, for completeness, my msctpeak parameters: images = "t2034" List of WCS calibrated Mosaic images
coordinates = "bright_hrs_2.txt" List of ra(hr), dec(deg), optional id
database = "coords.dat" Database for astrometric fit
(extname = "") Extensions
(epoch = 2000.) Coordinate epoch
(update = yes) Update image header WCS following fit?
(autocenter = no) Center catalog coords when entering task?
(boxsize = 7) Centering box fullwidth\n
(projection = "tan") Sky projection geometry
(fitgeometry = "rscale") Fitting geometry
(function = "polynomial") Surface type
(xxorder = 2) Order of xi fit in x
(xyorder = 2) Order of xi fit in y
(xxterms = "half") Xi fit cross terms type\n
(yxorder = 2) Order of eta fit in x
(yyorder = 2) Order of eta fit in y
(yxterms = "half") Eta fit cross terms type?\n
(reject = INDEF) Rejection limit in sigma units\n
(interactive = yes) Enter interactive image cursor loop?
(frame = 1) Display frame number
(marker = "circle") Marker type
(omarker = "plus") Overlay marker type
(goodcolor = "blue") Color of good marker
(badcolor = "red") Color of bad marker
(fdimages = "")
(fdcoords = "")
(mode = "q")