[evlatests] Line subarrays, round 2.

[Vivek Dhawan]

At Ken's prodding I sent in a pair of files in line mode today.

sub 1 = EVLA, changing BW every 4 min or so, including a reference
         pointing scan and a continuum scan in all 5 scan types, 2
         repeats. All line scans were in mode 4 (=RR,LL in IF 1,2)

sub 2 = VLA L band, 25MHz, mode 4, 4 scans of 15min each.


The data in sub 1 were exactly as expected. Sub 2 gave me an instant 
headache - neither the scans nor the modes were as expected. So I do
as perplexed Indians often do and went up the mountain to consult an 
astrologer, in this case Ken, and all was light: Sub 1 forces its 
correlator mode on sub 2.

I can verify that is so. The header labels are identical. The amplitudes
in sub 2 looked a little strange* so I am unsure if sub 1 sets the IF
filter bandwidths as well, or just the correlator mode and the data is
aliased or somehow not normalized. I have not looked at this last thing
in detail, because we either will live with the sub2 mode being the same
as sub 1, and avoid it in observe files, or the 2 subarrays will be made
truly independent and we test it more carefully then.

* different by upto a factor 4 from baseline to another (on 3C286),
also one IF to another on the same baseline. sub 1 showed no such
oddities.

Vivek.