[evlatests] WIDAR bronze dataset - C band dual pol.

[George Moellenbrock]

Folks,

Re ACCOR, Eric has replied to a wider group than the original discussion.
Michael requested something that corrects the cross-correlations using lag
zero (i.e., spectral average) of the autocorrelations, since this is not
done "on-line" yet, and may be the origin of some of our imaging
troubles.  Yes, ACCOR is it.

Some background:  ACCOR, as I understand it (from considerations
associated with VSOP ~10 years ago), was introduced to compensate
for non-ideal level populations in 4-level VLBA sampling.  I.e., this seems
to be a means of compensating for the antenuator fixedness of (most)
current observations. Leonia's VLBA Sci Memo #9 I believe describes
how it is justified to use the AC measurements (sqrt) as corrections for
the cross correlations.  An outstanding question is whether or not Leonia's
analysis obtained only in a relatively narrow linear regime of smallish
sampler population discrepancies.  Applicability to the 8-level case and
the possibly wider range of sampler population discrepancy is not yet
understood.  But it is worth a try.  In any case, it is exactly the lag zero
AC correction Michael requested, and works as Eric describes (SN tables
and whatnot).  FYI, CASA does not have an ACCOR equivalent, as yet.

Nonetheless, this is still an _antenna-based_ correction, and so the gain
calibration will pick it up if it is not compensated a priori, so long as the
data are _self-calibrated_ on a sufficiently short timescale.  I.e., I
don't think
this illuminates any fundamental problem with the data.

Cheers,
George

> Actaully George is correct - aips task ACCOR does what Michael is
> asking.  It averages the channels of an AC record per IF and per
> polarization and writes the result suitably to an SN table.  That can
> then be applied with CLCAL and following DOCAL TRUE.  It can be plotted
> with SNPLT and edited with SNEDT.  The help file for ACCOR is misleading
> but the code is just what one wants.
>
> Eric
>
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>
>