[daip] FITLD broken?

[Eric Greisen]

Adam Deller wrote:
> Hi,
> 
> There is something very strange going on with FITLD for VLBA data in
> AIPS - it is scrambling the SU table for some files (but not all).  I
> have checked out this particular file (you can see it at
> /home/marathon2/data/1023+0038/bd141a/BD141A_UG_HPOL4IF8MHZ16CH_1.6515GHZ00)
> using fv and see a sensible looking SU table.  However, as soon as I
> load it into AIPS with FITLD the SU table reads:
> 
>        1    5.0000000000D+05    1.4176255806D+02     3.9039125519D+01
>   2.0000000000D+03    1.0792773795D+02    4.5395855014D+01
>        2    5.0000000000D+05    1.5612327792D+02    -8.8208229444D-01
>   2.0000000000D+03    1.3016422419D+02    7.9956464878D+00
>        3    5.0000000000D+05    1.5591154583D+02     4.0463333333D-01
>   2.0000000000D+03    1.2979993496D+02    9.2488222431D+00
>        4    5.0000000000D+05    1.5271527786D+02    -2.0054371833D+00
>   2.0000000000D+03    1.2689149360D+02    6.4704735040D+00
> 
> where the columns are channel bandwidth, RA epoch (deg), Dec epoch
> (deg), Epoch (years), RA apparent (deg) and Dec apparent (deg).  You
> can see that the apparent source positions have changed by many
> degrees - something is amiss.  I tried running FITLD from 31DEC09 but
> got the same result.  This is data from the hardware correlator from
> late 2008.
> 
> Cheers,
> Adam
> 

Indeed - with partly processed files from FITAB the SU table was 
encountered before the observing data made it to the header.  Moved the 
call to the apparent position fixing routine to near the end where it is 
known in both FITLD and UVLOD.  N.b.  CASA forced me to add this to the 
tasks since they provide apparent positions for the birth of Christ.

Tomorrow's MNJ will fix this.

Eric Greisen