X-Mailer: exmh version 2.0.2 2/24/98
To: Warwick.Wilson@atnf.CSIRO.AU
CC: lstavele, mkesteve
Subject: Re: REFBEAM, etc. 
In-reply-to: Your message of Wed 1999/03/17 09:22:46 +1000
             <9903162322.AA14574@acorn.atnf.csiro.au> 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Wed, 17 Mar 1999 11:23:26 +1100
From: Mark Calabretta <mcalabre@grus.atnf.CSIRO.AU>


On Wed 1999/03/17 09:22:46 +1000, Warwick.Wilson@atnf.CSIRO.AU wrote
in a message to: Mark.Calabretta@atnf.CSIRO.AU
and copied to: lstavele@atnf.CSIRO.AU, mkesteve@atnf.CSIRO.AU, wwilson@atnf.CSI
     RO.AU

>To accommodate the new "multiple pointings per scan" (MX) mode, I propose
>adding the current reference beam to the so-called sys-cal 
>( i.e. baseline = -1 ) record which is written to the rpfits file
>at the start of each integration.  It will be quantity number 17 in the
>sc_cal array where the antenna number = 0. i.e. in the same area as the 
>azimuth, elevation and focus parameters appear, immediately following
>the time of measurement of the cal signal.  Unfortunately, this
>location in current rpfits files is not defined.  It should be ignored
>if there is no SCANTYPE card in the header - see below. 

I wasn't aware that sc_cal was used in this way!  There's nothing about it in
the rpfits documentation.

>I also propose writing a new user defined card ( the first ) to the header 
>which will contain the scan_type parameter.  The first 10 characters
>of the card will be "SCANTYPE= ".  The value will appear in card(11:30)
>as a list of space separated two letter codes.  
>
>Currently valid codes are:
>
>	FS	ATCA style frequency switching - within a scan
>	MO	ATCA style mosaicing mode
>	PU	Pulsar binning mode
>	TC	ATCA tied array circular polarisation
>	TL	ATCA tied array linear polarisation
>	RF	Reference scan - for single dish spectral line work.
>	SF	Slow frequency switching - per integration cycle
>	FF 	Fast frequency switching - within an integration cycle
>	MX	Multibeam MX mode - multiple pointings within a scan
>
>Comments ?

Fine, except that SCANTYPE = "track" seems oxymoronic - "SCAN" is an
unfortunate choice of collective noun for integrations in single dish
observing.  Could you put an expanded description of the two-letter codes
in the comment section of the card for human readers.

Cheers, Mark