[daip] CLCOR and the ATLBA

[Sean Dougherty]

Eric,
I sure you took input from Chris and others on what was required for the
ATLBA. I include a fragment from a recent email from Adam Deller  and 
his last statement suggests that even he is not 100% sure.  I've cc'd
both Adam and Chris Phillips in the hope that they may comment.

Adam's email of 25 june 2007:

"Having dug a little deeper, i believe there is some hard-coding in the 
clcor routine which is incorrect.  When you run it, do you get something 
to the effect "WARNING: This is ATLBA data - ignoring clcorprm(4)", 
followed by a similar line about clclorprm(7)?  I do.  When I went 
through the CLCOR.FOR source code, I found it hard-coded ATLBA data to 
right handed coordinates and local meridian - wrong on both counts, I 
believe."


Nevertheless, thanks for the prompt reply.  In the meantime, I can make
the necessary changes to CLCOR in my local version to see if I can get
the correct delay/rate/phase solutions.

Cheers,

Sean


Eric Greisen wrote:
> Sean Dougherty writes:
>  > Eric,
>  > 
>  > I have recently been trying to calibrate a test VLBI experiment with the
>  > ATLBA using the Swinburne software correlator. Unfortunately, there was
>  > an error in the position of one of the antennas used in the array and
>  > running CLCOR was required to fix the problem. The ATLBA has a
>  > left-handed coord system and uses Greenwich meridian. Using the
>  > supposedly correct values for CLCORPRM(4)=0 (for LH) and CLCORPRM(7)=0
>  > (for Greenwich) according to the help files did not produce the
>  > satisfactory solutions for delay, rate and phase as expected.
>  > 
>  > After exploring a number of potential solutions (none of which worked)
>  > in conjunction with Ed Fomalont (NRAO) and Adam Deller and Chris
>  > Phillips in Australia, a close examination of the aips code for CLOCOR
>  > shows that when the array is 'ATLBA' that it is NOT possible to use
>  > these values for the ATLBA. The code is written such that VLA, ATCA and
>  > ATLBA are hardwired to be RH and use local meridian (!&*@^#%$).
>  > 
>  > Any chance of modfying the CLCOR code so that it works correctly?
>  > 
>
> CLCOR does what the AT people (e.g. Chris Phillips) told me it should
> do.  Have they changed their minds?
>
> Since you are already familiar with the code - a specific suggestion
> for a code revision would be a good idea.  The number of places
> affected is not trivial:
>
> primate<929>$ AIGREP ATLBA $TST '*.FOR'
> /home/primate/AIPS/31DEC07/APL/SUB/NOTST/GETTIM.FOR
> /home/primate/AIPS/31DEC07/APL/PGM/NOTST/GETJY.FOR
> /home/primate/AIPS/31DEC07/APL/PGM/NOTST/ACCOR.FOR
> /home/primate/AIPS/31DEC07/APL/PGM/NOTST/PCCOR.FOR
> /home/primate/AIPS/31DEC07/APL/PGM/NOTST/CLCOR.FOR
> /home/primate/AIPS/31DEC07/APL/PGM/NOTST/SPLIT.FOR
> /home/primate/AIPS/31DEC07/APL/PGM/NOTST/SPLAT.FOR
> /home/primate/AIPS/31DEC07/APL/PGM/NOTST/UVCRS.FOR
> /home/primate/AIPS/31DEC07/Y/PGM/NOTST/LOCIT.FOR
> /home/primate/AIPS/31DEC07/Y/PGM/NOTST/CLPLT.FOR
> /home/primate/AIPS/31DEC07/Y/PGM/NOTST/VPLOT.FOR
> /home/primate/AIPS/31DEC07/Q/PGM/NOTST/CVEL.FOR
>
> Eric Greisen
>