[daip] correcting for moving sources

[Leonia Kogan]

Bryan,

I thought a little bit about the difference between CLCOR  and direct 
(your task -correction of the phase)
correction of the visibilities and arrived to the conclusion that they 
should go to the same result.
And as we discussed it is not necessary to speak about the variable in 
time correction of the source position.
The comparison is valid for the constant correction of the source 
position. You know that we made a deep
comparison of CLCOR constant correction of the source position and shift 
at UVFIX which practically identical
to your task of visibility phase correction. We got the very good 
coincidence of the results

But still what is the difference:

I think the difference can be at the computation time for the following 
reason:

1. Number of antennas is smaller than the number of the baselines at the 
same time (4 times for VLBA)
    CLCOR better.

2. Time interval at the CL table is bigger (typically several times) 
than the time step at the visibilities:
    CLCOR better.

3. If there is more than 1 frequency channels (IFs) then UVFIX should 
repeat the recalculation for each
    channel; CLCOR calculates only one parameter: delay for this purpose.
    Again CLCOR better.

4. For better application of the CL table phase rate in time (rate) is 
calculated also.

So if the computing time is essential than CLCOR will have advantage in 
comparison with UVFIX-your task.

My estimation does not include the application of the CL table during 
calibration but ... I do not know

I put this question at the minute of tomorrow AIPS meeting. Come by

Leonia

Leonia