[daip] New Staff Reply - [!XSK-573080]: Calibrating weights doesn't seem to work....

[Eric Greisen]

New Staff Reply: Calibrating weights doesn't seem to work....

There are a lot of tasks with DOCAL so I am loathe to
put this is.  It is described in Appendix E of the CookBook.
The actual system temperatures do vary between SPWs and
between antennas enough to make weights vary by factors > 2.

TYAPL is the preferred approach at higher frequencies although
the SY table has problems at lower frequencies due to to RFI.

BDFIn does not offer any averaging or scan selection.  The CASA
route allows time and frequency averaging.  Since no info about
the original frequency spacing is retained, it is impossible
to correct amplitudes for delay errors by that route.  The
time averaging works, but the SY data are then lost to AIPS.

REWAY is in some ways more reliable in that it looks at the actual
achieved rms rather than the one expected from Tsys.  The weights
found with APARM(1)=0 are noisy and would add noise to the imaging.
Usually some sort of rolling buffer is used (APARM(1)=9 or 19 or ??)
and then checked with VPLOT or ANBPL.

REWAY works with the real part and the imaginary part (avoids the
Ricean biases) and so would be messed up by a significant delay error.
Of course your imaging would be even more messed up by bad delays.

Usually FRING and BPASS take such things out.

Eric Greisen


Ticket Details
===================
Ticket ID: XSK-573080
Department: AIPS Data Processing
Priority: Default
Status: Open
Link:  https://help.nrao.edu/staff/index.php?_m=tickets&_a=viewticket&ticketid=3270