[daip] [!4546]: aips - FRINGE odd behaviour on initial, baseline-based coarse search

[Michael Bietenholz]

Michael Bietenholz updated #4546
--------------------------------

FRINGE odd behaviour on initial, baseline-based coarse search
-------------------------------------------------------------

           Ticket ID: 4546
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/4546
           Full Name: Michael Bietenholz
               Email: mbieten at yorku.ca
             Creator: User
          Department: AIPS Data Processing
       Staff (Owner): -- Unassigned --
                Type: Issue
              Status: Open
            Priority: Default
                 SLA: NRAO E2E
      Template Group: Default
             Created: 20 February 2014 04:34 PM
             Updated: 20 February 2014 04:34 PM
                 Due: 25 February 2014 12:00 AM (4d 7h 26m)
      Resolution Due: 01 March 2014 12:00 AM (8d 7h 26m)



As I understand it, FRING first does a coarse FFT fringe-search on single baselines to get starting values, and then uses these starting values to do a least-squares fit involving all the baselines.    In that case, the results of the initial, single-baseline part of the search should not depend on which  other antennas might be included in the fit.   However, it seems like the results of the initial coarse search some do depend on whether or not other baselines are present.

For example, I ran FRING normally, using all baselines (refant = 7)

FRING1 11:17:25 Time=   0/ 08 32 31, Polarization = 1
FRING1 11:17:25 Coarse search spacing:   0.488E+00 ns;   0.243E+00 mHz
FRING1 11:17:25 B = Baseline  R = Rate (mHz)  D = Delay (nsec)
FRING1 11:17:25 B= 07 - 08 IF=  1 R=      0.24 D=       0.0 SNR= 348.3
FRING1 11:17:25 B= 02 - 07 IF=  1 R=     -0.49 D=      -0.5 SNR= 328.1
FRING1 11:17:25 B= 01 - 07 IF=  1 R=     -0.49 D=       0.0 SNR=   7.2
FRING1 11:17:25 B= 03 - 07 IF=  1 R=     -0.24 D=       0.5 SNR= 164.0
FRING1 11:17:25 B= 04 - 07 IF=  1 R=      0.00 D=       0.5 SNR=   4.5  <<<<<<<<

Now my expectation would be that if I run FRING on the same data, with all other parameters the same, but selecting only ANTENNAS=4,7, I would get the same result from the coarse fringe search on baseline 4-7.    This is not the case, here's what I got - notice that the
SNR is much lower this time

FRING1 11:11:33 Coarse search spacing:   0.488E+00 ns;   0.243E+00 mHz
FRING1 11:11:33 B = Baseline  R = Rate (mHz)  D = Delay (nsec)
FRING1 11:11:33 B= 04 - 07 IF=  1 R=      0.73 D=      17.1 SNR=   2.2  <<<<<<<

Am I misunderstanding how FRING works?  Do you know what is happening here?


------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff