[daip] ACCOR question

[Leonia Kogan]

Michael,

Couple comments:

The whole idea of the ACCOR ( correct cross correlation using autocorrelation)
works only for VLBA. 

> I've
>usually tossed these when it only shows up for a few minutes, but what
>if ACCOR consistently derives a corretion of 2 for an antenna for
>a couple of hours (but 1 the rest of the time).  SHould I beleive it?

Yes, if it is VLBA antenna, although the theory says that the
correction of the cross using auto works properly only for small
deviation of the mean of the auto spectra of 1.0. I'd say <1.5

The bigger deviation says about serior problem of the digitizer at the
relevant BBC. 

Are this 2bit(four level) data?
What antenna, frequency, IFnumber, bandwidth?

Leonia

----- Begin Included Message -----

>From michael at polaris.phys.yorku.ca Tue Jul 29 12:29 MDT 2003
Date: Tue, 29 Jul 2003 14:32:25 -0400 (EDT)
From: Michael Bietenholz <michael at polaris.phys.yorku.ca>
To: lkogan at nrao.edu
Subject: ACCOR question
MIME-Version: 1.0
X-MailScanner: Found to be clean
X-MailScanner-SpamCheck: not spam, SpamAssassin (score=0.2, required 7,
	SPAM_PHRASE_00_01, USER_AGENT_PINE)
X-Lines: 9
Status: RO


Usually the amplitude corrections made by ACCOR are close to 1.0
(within 10%) once in a while it comes larger values (ie >1.5).  I've
usually tossed these when it only shows up for a few minutes, but what
if ACCOR consistently derives a corretion of 2 for an antenna for
a couple of hours (but 1 the rest of the time).  SHould I beleive it?

       thanks,     michael bietenholz



----- End Included Message -----