[daip] CALIB question (Classic Aips)

[Eric Greisen]

Samuel Conner writes:

 >   I am attempting to use the SMODEL parameter in CALIB (Classic
AIPS, 31DEC2004, a July build) to correct the pointing on a calibrator
scan in an archival observation (multisource data). I'm using a 2004
Goddard VLBI position, which differs from the historical scan position
by about 75 mas.

There were a variety of corrections made - you would be better to
update with the do_daily.hostname script to the now frozen version.
I doubt that that will affect this "problem".

 > 
 >   In earlier versions of CALIB, it was possible to set SMODEL = 1,
delta-ra, delta-dec,0  to correct a pointing in a multisource UV
datafile as well as single source files.

I am having trouble following this argument. SMODEL = 1,0 with no
shift is the same a declaring the source to be a point source
calibrator of flux 1 Jy.  That would, if the source was observed at a
wrong position, generate some phases in an SN table.  But these are
not phases to be applied to the target sources only to the calibrator.
But CLCAL will apply them to both unless it is run with great care.

If SMODEL is ignored and I think that that is the case in multi-source
files, then it will do the very same thing but using the flux in the
source table instead.

What you really want to do is run CLCOR on the calibrator to correct
its position.  There is a gotcha if these are old data in B1950
coordinates.  You should update your version of aips to the frozen
version, and then reload the data with FILLM before doing CLCOR.  This
might even be significant for J2000 positions although the error there
is much less devastating.

 > 
 >   It appears to me that this is no longer the case. The HELP file
seems to imply that SMODEL only works with single-source UV data, and
when I try to CALIB with SMODEL set or not set, the gain phases are
the same (this in an A array C-Band observation, where I would expect
substantial phase differences for a position offset this large).

     The help file says this correctly.

 > 
 >   The HIstory file does list the SMODEL parameters, though they
seem to have had no effect on the calibration.

     We did not zero SMODEL as we should to keep the HI routine from
recording it.
 > 
 > Two questions:
 > 
 > A)  Am I missing something or is SMODEL indeed disabled in
multi-source data?

     It is disabled - you could use a source model but not SMODEL.

 > 
 > B)  Would it be safe for me to adjust the coordinates in the SU
table in order to achieve the same effect? 

     What would that do?  It certainly would not change any phases. 
You need to do CLCOR but the update is highly recommended first.  Bugs
in exactly this area were addressed.

Eric Greisen

 > 
 > 
 > Thanks for your time,
 > 
 > Sam Conner
 >  
 > 
 > _______________________________________________
 > Daip mailing list
 > Daip at listmgr.cv.nrao.edu
 > http://listmgr.cv.nrao.edu/mailman/listinfo/daip