[daip] CALIB, avoiding EVLA-EVLA baselines

[Eric Greisen]

Elias Brinks writes:
 > Hi Eric,
 > 
 > Our project (LITTLE THINGS) is severely hit by the aliasing problem.  
 > We're
 > using either  0.78 or 1.56MHz bandwidth.
 > 
 > I'm trying to calibrate some siilar data Juergen Ott recently got, to  
 > see is
 > we can calibrate/rescue the project using tour UVLSF fix. At least,  
 > we have
 > line-free channels at either end of the bandpass, so hopefully a fit  
 > will be
 > able to get rid of most of the aliased continuum.
 > 
 > But, we also have to calibrate the amplitude/phase, of course. My  
 > plan was to
 > run CALIB, excluding the EVLA-EVLA baselines. This should still give me
 > enough baselines to calculate the Antenna based solutions (via the  
 > VLA-EVLA
 > and VLA-VLA route). Now, how do I achieve this in CALIB? I read and  
 > re-read
 > the part on ANTENNAS, DOFIT and ANTUSE, but got confused.

   I recommend the verbs VLA and EVLA and making a UVFLG entry to flag
out the vla-evla baselines.  Then unflag them in UVFLG after CALIB has
worked.

 > 
 > In fact, I was expecting something like a BASELINES adverb where I  
 > could simply
 > specify which ones to exclude in the baseline-based solution part,  
 > and then
 > have ANTENNAS=0 (or DOFIT=0) to get ANTENNA based solutions for all.
 > 
 > Going this route would allow me to use the regular CHAN 0. Anyway, in  
 > an 0.78MHz
 > bandwidth I wouldn't have any channels not affected by the aliasing,  
 > so this would
 > seem to me the only sensible approach (and yes, this will break down  
 > the moment
 > we get even more EVLA antennas...by which time I hope the aliasing  
 > has been
 > dealt with in hardware).

It will not be dealt with in hardware.  The cost estimate was
prohibitive.

Eric Greisen