from:	 Jonathan Weintroub 
to:	 "Young, Ken" 
cc:	 Mark Gurwell ,
         Eric Keto ,
         Nimesh Patel ,
         Chunhua Qi ,
         Lars Kristensen ,
         David Wilner ,
         Qizhou Zhang ,
         Sridharan Tirupati Kumara ,
         Jun-Hui Zhao ,
         Alice Argon ,
         "Primiani, Rurik" 
date:	 Thu, Dec 5, 2013 at 12:27 PM
subject:	 Re: SWARM Correlator Test Track - When did the music stop?
mailed-by:	 cfa.harvard.edu

Hi Taco, et al.,

Likewise my thanks for the various informative emails (e.g. earlier note on closure phase) which I follow with interest albeit without comment.

I'd like to make a discussion of the status of SWARM verification data a substantial 
part of tomorrow's system call.  To that end, would you please place transcripts of 
relevant emails and attachments on the Twiki. I think you have already posted to this location:

https://www.cfa.harvard.edu/twiki5/view/SMAwideband/ScienceVerification

After you post I'll link to this when I send out the agenda.

Feel free to edit or reduce the messages or attachments, but please don't be reluctant 
to put preliminary or rough notes up.  The intent of the Twiki is a repository of 
intermediate results and rough notes . . . as an alternative to email to limited 
distributions, while not feeling that you are spamming folks.  For example, 
Bob Wilson is back, and I hope he'll participate, I'd like Laura to see the data too, 
Hawaii folks, etc.

The SWARM data looks encouraging, but there are a number of quite puzzling anomalies.  
As when we met last week to discuss this, I don't think there are clear hypotheses 
as to root cause, though possible phase tracking errors are suggested.  I'd at least 
like to make a short form list of these anomalies, some of which may trace to the 
same root cause(s).

1.  SWARM spectra are in general noisier than comparable legacy correlator.  
    Most easily seen in phase continuum.

2.  The degree to which this is the case (ratio of phase variances SWARM/legacy) 
    varies from baseline to baseline

3.  The amplitude of a spectral line in SWARM varies from baseline to baseline 
    to a greater degree to legacy (about 10% in legacy I think)

4.  Within the same baseline averaged spectrum the amplitude of the two peaks of 
    the MWC349 double line seen in SWARM are not consistent relative to those in legacy.  
    I think primarily because of 3. it is felt that SWARM is the spectrum in error.

5.  SWARM closure phase on point source is non-zero, legacy is zero as expected.

6.  SWARM closure phase is substantially noisier than legacy.

Please amend or augment this list (applies to whole distro), and consider posting some 
version of this  to focus discussion tomorrow.  My thinking is to brainstorm hypotheses 
and tests thereof at the system meeting.

Thanks again, and I hope you are ok with wider distribution and discussion tomorrow.

Best,

Jonathan