<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7654.12">
<TITLE>RE: Keeping track of spectral windows at the EVLA: SDM vs. BDF</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<BR>
<BR>
<BR>
<BR>
<BR>
<P><FONT SIZE=2>Of course this begs the question of who comes up with swbbName and swIndex<BR>
to begin with. This seems properly left to the OPT: there is enough<BR>
information there, and these values can be determined independently for<BR>
every scan. It would be preferable to have a deterministic algorithm for<BR>
doing so. For swbbName this is trivial. For swIndex within a given<BR>
swbbName I propose:<BR>
<BR>
1- sort by the lower edge of the lowest frequency channel in each subband<BR>
(to match the ALMA ACA correlator)<BR>
2- for remaining matches, give the lower swIndex to the SpW with<BR>
the narrower channels<BR>
3- for remaining matches, give the lower swIndex to the SpW with the<BR>
fewer total channels<BR>
4- at this stage we have only exact matches left. If the OPT or whatever<BR>
is explicitly setting the StB filter to be used, the lower filter gets<BR>
the lower swIndex.<BR>
<BR>
<BR>
<BR>
Proposed new attributes convey important information; it seems to me that it is better<BR>
to define two new VCI attributes (swbbName and swIndex).<BR>
Using VCI "name" for this purpose will look as a cludge; at this early stage we can add things to schema without worrying about legacy code.<BR>
Adding a new attribute to VCI schema is not a big deal; most of the work is in implementing the functionality.<BR>
<BR>
<BR>
Comment regarding selection of swIndex step 4:<BR>
OPT assigns the Subband ID (SBID).<BR>
The CM uses the filter with the same ID as subband, if possible.<BR>
If not possible, CM uses the first available (i.e. unused) filter).<BR>
I propose that the swIndex in step 4 should be assignd according to the SBID (most of the time that would be the same as Filter ID).<BR>
<BR>
<BR>
Should I then add the following:<BR>
swbbName to element baseband, type:string (in fact enumeration)<BR>
swIndex to element subband, type:int.<BR>
<BR>
<BR>
Sonja<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
All of this of course is negotiable.<BR>
<BR>
A couple points:<BR>
<BR>
* In the BDFs, all SpW in "lower" swbbName's come before all SpW in "higher"<BR>
swbbName's, regardless of the actual frequency tunings. A/C comes<BR>
before B/D, full stop; similarly data from A1/C1_3BIT comes before<BR>
A/C_8BIT.<BR>
<BR>
* In the BDFs, within a given swbbName, the above scheme would put<BR>
non-overlapping SpW in frequency order (low to high).<BR>
<BR>
* None of this says ANYTHING about the ordering of the actual Spectral<BR>
Windows seen by the observer in post-processing, or even in the SDM.<BR>
That is a separate discussion, which I hope to initiate tomorrow or<BR>
Thursday.<BR>
<BR>
<BR>
<BR>
Comments?<BR>
<BR>
Michael<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>