[evlatests] More remarkable results

[Rick Perley]

    After bandpass calibration, gain calibration, I splitted out 950 
channels to look at detailed amplitude and phase behavior.

    Only two baselines (of the 45 in the database) are misbehaving:

    1) Baseline 18 x 19 is today's chosen one for the 4-second phase 
oscillation.  The amplitude is a few degrees.  For the first time, I see 
also an amplitude oscillation -- same period.  (Probably visible because 
this is 3C84, with lots of flux). 

    2) Baseline 23 x 25 is *hugely noisy*.  Both the amplitude and 
phases have *eight times* the scatter of any other baseline (except 18 x 
19 -- see above). 

    Other than these, the data are very nice indeed. 

    The mini-halo of 3c84 is easily seen in the visibilities.