[evlatests] u,v,w, and UVFIX

[Ed Fomalont]

Hi Gustaaf et al,

    Thanks for looking into UVFIX.   The pulsar astrometry people first 
noticed the differential aberration effect when images taken at 
different times of the year showed a stretching of the coordinate system 
of about 1 part in 10,000--just noticeable on sources 10' from the field 
center at 1.4 GHz.  The uvw calculated from the on-line system, from 
John's program and with UVFIX (nodr) are consistent and the u,v,w are in 
the coordinate from of the earth center.  UVFIX (dr) converts the u,v,w 
to the solar barycenter frame.   By John's program, I guess you mean 
CALC.  I think it was agreed many years ago to not change the u,v,w 
calculation on line.  Running UVFIX (if nothing else is wrong!) is only 
needed for very precise full-beam astrometry.  It is recommended for 
full-beam VLA observations at frequencies less than 2 GHz.

    I'm not sure I understand the last little table in your note.  What 
is most important is the separation of the point source from the phase 
center.  Near the phase center I don't see why UVFIX would have affected 
the image amplitude very much since the change of uvw is very small 
until the source is many arcmin from the field center.  If there is a 
significant change in the image, something else is happening as well. 
In the past UVFIX had problems dealing with many channels, IF's etc, 
however I think this is all okay.  Check with Eric and Leonia, though, 
about this.

    Cheers,  Ed