[evlatests] Wrong bandwidths and frequencies in continuum

[Rick Perley]

    I noted a couple days ago that a P-band continuum observation, taken 
with 3 MHz BW, was listed by AIPS as being at 50 MHz BW.  The data 
appear fine -- only the header is wrong. 
    In other tests done over the past few days, we find that the 
BW/frequency is in almost all cases incorrect.  (The word 'almost' is 
used because there have been numerous combinations of BW and frequency 
in the tests, and most, but not all, give incorrect headers). 
    More interestingly perhaps, is that none of the tests taken before 
August 1 (and after the Modcomp switchover) provided incorrect headers 
for continuum observations. 

   Here is a log of recent continuum observations, and the resulting 
headers:

    July 30:  P-band continuum at 3 MHz followed a line observation at 
12 MHz.  Headers are *correct*. 

    July 31:  L and C bands were observed cyclically at 50, 25, and 12.5 
MHz in continuum.  Headers are *correct*. 

    August 1:  P-band continuum at 3 MHz followed a continuum 
observation at 50 MHz.  Headers are *incorrect*.  (BW was 50 MHz, and 
the SSLO is also not correct for the BD side). 
   
    August 3:  P-band continuum followed a 50 MHz continuum observation 
at L-band.  Headers are *incorrect* (and the same as above). 

    August 6:  P-band continuum:  File began with 2 scans at 6 MHz, 
followed by one at 3, and then one at 1.5 MHz.  Headers think all 
observations were at 6 MHz. 

    August 6 (later):  The entire file was at 3 MHz.  Headers show two 
'FreqIDs' -- a single (invisible) record at 50 MHz BW (and the same 
wrong frequencies as on August 1 and 3), followed by the corrected BW 
and frequencies. 

    The only other (probably irrelevant) fact is that all the 
observations made prior to July 30, and the second run from last night, 
used stop times in the observe file, while all the others  using 
'durations'.