[daip] AIPS CLCOR problem

[Eric Greisen]

What version of CLCOR are you running - if 31DEC04, when was the last
MNJ run?

Actually - since you have J2000.0 coordinates, that is less important
(there were real problems with B1950).  ANTP is supposed to shift the
apparent coordinates while ANTC shifts the observed coordinates.  That
will be part of the problem.  The other part is more insidious - the
apparent coordinates from the VLA from FILLM have until a few days
ago been those computed by the VLA and contain diurnal aberation (up
to 0.3 arc sec) for some time in the observation.  Starting with
31DEC05 and in the nearly frozen 31DEC04, FILLM will not use the VLA
supplied apparent coordinates, but compute them with the same routines
that will be later used by CLCOR.  This now - finally - gets us on a
consistent basis.  It may not be perfectly correct but it will be
consistent which is more important.  What you are seeing - you get 1
sec on later tries but not the first is the discrepancy between how
the VLA computes apparent coordinates (changes with each scan) and how
a post-processing system will deal with them (a single number).  If
this error is significant, update your 31DEC04 AIPS - or install
31DEC05 - and re-run FILLM.  You can then copy the gain tables
(pre-CLCOR) and flag tables from your current data set.

Eric Greisen