[daip] CALIB problem

[Bryan Butler]

i'm trying to calibrate some 4-band data, in order to remove
CasA from the data.  i have a model of CasA from namir kassim
which i would like to use.  but, i get the following error
when i try:

 CALIB2: SETGDS: CENTERED  0 OFFSET  1 IN  1 MODELS 3DIMAG INCONSISTENT
 CALIB2: Divide data by model - first compute model by summing
 CALIB2: ZMI2: REQUEST FOR BYTES 198657 THRU = 237568 BEYOND EOF = 235520
 CALIB2: ZMIO: OPER=WRIT LUN=20 BLKNO=         195 8-BIT-BYTES= 38912
 CALIB2: ZERROR: IN ZMI2   ERRNO = 22 (Invalid argument)
 CALIB2: APIO: MDISK ERROR  3 FLUSHING BUFFER
 CALIB2: CCSGRD: ERROR  13 CLOSING GRID FILE

this seems to be traced to the routine SETGDS where a check is
done for whether this is 3D imaging or not:

C                                       3D image?
         IF ((ABS(CATCR(KRCRP)-CATCLN(KINAX)/2.0).LT.0.01) .AND.
     *      (ABS(CATCR(KRCRP+1)-1.0-CATCLN(KINAX+1)/2.0).LT.0.01))
     *      NCEN = NCEN + 1

which seems to enforce that the image that you are using is
'centered' (the reference pixel is half the total number of
pixels).  this is not the case for our CasA model image, which
is 115x107 pixels and has pixel 66,68 as its center.

it works fine when doing the same thing with a CygA model, but
the image which we have for CygA *is* centered (256x256 image
with center at 128,129).

is this a "bug" or a "feature"?

	-bryan