Hi,

While trying to reduce the data from expt. w401b (4-5 Feb 2001),
I encountered a problem when FITLD'ing the data into AIPS.
Some job outputs on the FITS archive tape cannot be concatenated
with the rest, because for some reason they apparently have two 
FREQID's. These are, however, identical, as FITLD recognizes it.
 
But, as the data are probably associated with FREQID=2, and the 
uv-file with which the new one is to be concatenated has only one,
the data are not read in at all:

-------------------------------------------------------
Incoming FREQID #   1
 1634.00   1650.00
Incoming FREQID #   2
 1634.00   1650.00
!!!!! WARNING: FREQIDS   1 AND   2 ARE IDENTICAL
Attaching data to W401B1      .LOAD  .   1 Vol:   2
CL table interval set at  0.50 minutes
           !!!!!!!!!!!
Digital corrections being performed
Total power corrected even if no zero-padding
           !!!!!!!!!!!
Current file has    24080 visibilities
File ref. freq:     1634.00 MHz
UV table spanned time:   0/21:39:13 -  0/22:14:19
Rejected  24080 records due to frequency selection
     0 vis. written
-------------------------------------------------------


If I start FITLD'ing with such a file with two FREQID's,
the message I get:

-------------------------------------------------------
Incoming FREQID #   1
 1634.00   1650.00
Incoming FREQID #   2
 1634.00   1650.00
!!!!! WARNING: FREQIDS   1 AND   2 ARE IDENTICAL
CL table interval set at  0.50 minutes

(...)

Current file has    24047 visibilities
File ref. freq:     1634.00 MHz
Found J0238+16         at   0/20:23:30 src #   2
                                         fqid #   1 was    2
     10000 vis. written

(etc...)
-------------------------------------------------------
 
So I can actually FITLD these files one by one, and the resulting
AIPS uv-files will look "normal", having (correctly) only one FREQUID.
But unfortunately I cannot concatenate them later on, because
some antennas appear later, and thus the AN tables of these files
are not identical, and DBCON fails to do it correctly.

Is there an easy (or maybe not-so-easy) fix to this problem?
It may also be worth investigating how FITLD could correctly handle
such an unlikely (but apparently real) situation.

Correlator job nos. were 3320-3334. Files from jobs 3322, 3323,
3327 and 3328 are affected (have two, otherwise identical FREQID's).

The AIPS version I use is 31DEC01.

Thank you very much for your help in advance.

Cheers,
Sandor Frey