[daip] New Client Reply - [!LJF-732007]: ACCOR vs. DiFX-correlated data

Andrzej Marecki do-not-reply at nrao.edu
Fri Mar 12 04:00:03 EST 2010


New Client Reply: ACCOR vs. DiFX-correlated data

OK, but...

Deller et al. (2007, PASA 119, 318) conducted a comparison study between VLBA and DiFX correlators. In Section 3.5.2 of that paper, they wrote:

"The VLBA correlator data were read into AIPS using FITLD with the parameter DIGICOR = 1. The DIGICOR parameter is used to apply certain scalings to the visibility amplitudes for data from the VLBA correlator. Furthermore, to obtain the most accurate scaling of the visibility amplitudes, the task ACCOR was used to correct for imperfect sampler thresholds, deriving corrections of 0.5% to the antenna‐based amplitudes. These ACCOR corrections were applied to the data, which were subsequently written to disk in FITS format.

The software correlator data were read directly into AIPS and then written to disk in the same FITS format as the VLBA correlator data. No corrections to amplitude or phase of the software‐correlated data were made in AIPS."

To me, it is rather clear they did not use ACCOR for DiFX-processed data (but please correct me if I'm wrong). Yet, the result of the comparison was surprisingly good - see Table 4.

Another underpinning of my notion/prejudice (you name it) that ACCOR is somehow closely related to the VLBA *hardware* corellator is a clear notice in the EVN Data Analysis Guide
(http://www.evlbi.org/user_guide/guide/userguide.pdf, page 14):

"Do not run ACCOR! ACCOR should only be run on data that were produced at the VLBA correlator in Socorro."

The above sentence and particularly the word "only" in it makes it clear (to me) that ACCOR should *not* be used for *any* other correlator except the VLBA hardware correlator. Fine, so I never apply ACCOR to JIVE-coorelated data but I'm confused now w.r.t. the DiFX-correlator data.

To summarize: please, please send me just a one-word - "yes" or "no" - answer to my initial question: to use or not to use ACCOR on the data produced by NRAO DiFX correlator?

Ticket Details
===================
Ticket ID: LJF-732007
Department: Data Processing (AIPS)
Priority: Default
Status: Open
Link:  https://help.nrao.edu/staff/index.php?_m=tickets&_a=viewticket&ticketid=107




More information about the Daip mailing list