[daip] Re: AIPS-DBCON

[Eric Greisen]

Shinji Horiuchi writes:
 > Hi Eric,
 > 
 > I asked you about AIPS-DBCON almost three months ago.
 > We solved the problem with polarization internally in the program 
 > outside of AIPS.
 > We are now worrying about merging two IFs data.
 > Is there any simple way to form a two IF fits file put of two single IF files?
 > It seems DBCON doesn't create such data.

The problem is the same as the polarization one.  AIPS assumes that
the 2 or 4 polarizations and the N IFs are all observed and correlated
at once and appear in an array (3 complex, 2 or 4 pol, N IFs) at each
antenna pair/time sample.  DBCON also does not merge two IFs.  It will
put together 2 files that were observed at different frequencies,
simply concatenating them.  It assumes that the u,v,w's are all in
wavelengths at the observed frequency and so this is okay.  It would
however be a mistake to recompute the u,v,w with UVFIX or to try to do
frequency-dependent things such as the primary-beam correction in
IMAGR.

I suggested using FUDGE to build your own program to do the
polarization merger - it could also do an IF merger with some effort.

Eric Greisen