iraf.net

NUMBER:	463
MODULE:	splot
SYSTEM:	V2.11-V2.11.3a
DATE:	Tue Feb 15 15:59:39 MST 2000
FROM:	valdes

BUG:	The 'h' key method for measuring equivalent widths will produce
	an arithmetic exception (such as divide by zero) when the pixel
	coordinate becomes large (say > 4000).  The exact error and limit
	value depends on the host system.  The workaround is to extract the
	desired piece of the spectrum into a shorter piece.  If you know
	the pixel coordinates you could use IMCOPY with an image section.
	If you know the wavelength region then use SCOPY.

	    on> imcopy longspec[5210:5310] temp
		or
	    on> scopy longspec temp w1=6140 w2=6143 rebin-

	    on> splot temp
	    

	Alternatively one can use one of the other methods for measuring
	equivalent widths such as 'e', 'k', or 'd'.

STATUS:	The algorithm is revised in the next release.

NUMBER:	521
MODULE:	mscred.ccdproc
SYSTEM:	V4.8: November 4, 2002
DATE:	Fri Mar 14 14:59:49 MST 2003
FROM:	valdes

BUG:	When using the 'merge' option in conjunction with 'bpmasks' output
	and 'fixpix' the merged masks are not properly created.  They will
	have names 'tmp...', will not contain the right pixels, nor will
	the BPM pointers in the data images be correct.  The workarounds
	are either not merge the amplifiers or do the merging first and
	then do the fixpix and saturation/bleed trail masking after the
	extensions are merged.

STATUS:	The latest version of MSCRED has this bug fixed.