[daip] CALIB broken?

[Eric Greisen]

Andrew Biggs wrote:
> Hi Eric. The colours are different IFs/polarizations and the station IDs
> are C for "central" (a square km full of antennas) and E means "East" i.e.
> for the East arm with numbers increasing according to how far away they
> are from the centre. S therefore refers to the South arm.
> 
> Anyway, I've tracked down the source of the problem. The clue is that, in
> the plot I sent you, the first source looks fine. If I therefore run the
> 31DEC13 CALIB with just the second source (I usually run it with both
> simultaneously), the resulting solutions look absolutely fine. Does this
> ring any bells?

Given that I cannot find any difference between 31DEC13 and 31DEC12 
CALIB in either CALIB.FOR or any relevant subroutines, it is hard to 
come up with a clue.  Is your 31DEC12 up to date or when was the last 
time you ran a MNJ on it?  Can I see your inputs to CALIB when you try 
to do the 2 sources at once (and it fails)?  Also an IMHEADER of the 
input file and perhaps a SCAN output of LISTR?  It is as if some model 
is being applied to both sources but with a position error in the 2nd 
when you do both at once - perhaps the CALIB messages as well might 
indicate this.

Thanks

Eric