[daip] DBCON => UVFIX

[Hanno Spreeuw]

Hi,

I think I may know what the problem is. These two observations both had 
a 333 MHz IF (with Stokes RR only) and a 323 MHz IF (with Stokes LL only).
What I did for both observations is to merge LL and RR, using DBCON. 
DBCON does not accept different Stokes values, so I manipulated the 
header (using PUTHEAD), so I turned the "LL IF "of both observations 
into a fake "RR IF".
So the file you downloaded is actually the DBCON of two files that were 
also made by DBCON.
Sorry I didn't mention this before, I totally forgot, I did this quite 
some time ago.
I suspect the problem has to do with the LL frequencies that are not 
accurately represented in the headers.
This final dataset is of pretty good quality, but I guess I hacked it to 
the extent that UVFIX cannot be expected to work properly any longer.

Cheers,
Hanno.

Hanno Spreeuw wrote:

> I just wanted to add that the angular separation would have been much 
> larger, of the order of a degree or so, if one set of visibilities 
> were converted to J2000 and the other were not.
> The angular separation is about ten beamwidths or so, but the 
> orientation (North, East, South, West) of the correctly located source 
> with respect to the "ghost" source is not the same for all sources. 
> The sum of the fluxes of these seem to amount to the correct flux.
> The file you downloaded is the result of a large number of selfcal 
> iterations on the original DBCON file. It gives a correct image in  
> B1950 coordinates.
> Hanno.
>
> Leonia Kogan schreef:
>
>> I was wrong at the previous message considering that DBCON does not 
>> provide the information about
>> subarray in "base".
>>
>> I UVCOPed  the huge HANNO's data to small data of 20 visibilities 
>> included both subarrays
>> and found that both
>> DBCON and UVFIX do the right job.
>> So the problem has not been detected yet :(
>>
>> I am sorry
>
>