[AG-TECH] AG on wireless

[Jay Beavers]

Actually, running at these bitrates is not something limited to Windows
Media compression.  H.261 can run these bitrates too, but with the
quality hit.

We're in limited private distribution of our binaries today and have
been running tests with the software between our lab and some
universities for some time, as lab only testing is pretty worthless for
these types of applications ;-)  We presented the first results of our
'over the Internet2' testing at the 2001 AG Retreat, so we've been out
there for a while now ;-)  We're waiting until we're comfortable with
the code quality and interoperable with lots of video equipment (such as
the Hauppauge) before we go to an open release.

It's definitely true that the work done in WMVideo and WMAudio
compression will not be cross platform.  There are limited ports to
Windows CE and Macintosh, but the licensing / IP of these codecs are
pretty tightly held and considered to be a key Windows differentiator.
This is why we in MS Research are funding work on transcoding solutions
to h.261, because we don't want to imply we're looking at a port of
Windows Media to a cross platform codebase nor any form of open source
release of the algorithms but we want to be interoperable with our AG
work.

 - jcb

-----Original Message-----
From: Tony Rimovsky [mailto:tony at ncsa.uiuc.edu] 
Sent: Friday, March 22, 2002 10:49 AM
To: Jay Beavers
Cc: Allison Clark; ag-tech at mcs.anl.gov
Subject: Re: [AG-TECH] AG on wireless

But in practical application *today* on the AG, windows media
compression is not in use outside of your lab, correct?  Will there be
a way to use this compression from a linux based video or audio
source?  I'm not trying to stoke OS flames here, just curious.  I
haven't looked into licensing/source/etc for this at all.




On Fri, Mar 22, 2002 at 09:57:16AM -0800, Jay Beavers wrote:
> I would disagree with the assessment that AG scenarios are too heavy
for
> wireless networks.
> 
> We comfortably fit four way conferencing, with one of the four streams
> being 640x480, in 3 Mb/s today using Windows Media compression.
> Assuming we limited ourselves to 6 Mb/s of bandwidth, we could add
> another 7 320x240 streams at ~400 kb/s for audio, video, and data
> bringing us up to 11-way conferencing.
> 
> Certainly, using more bandwidth per stream (such as MJPEG compression)
> or trying to do 20-30 streams isn't feasible over 802.11b, but for the
> sweet spot of 4-7 room conferencing, wireless is completely feasible.
> 
> -----Original Message-----
> From: Allison Clark [mailto:aclark at ncsa.uiuc.edu] 
> Sent: Friday, March 22, 2002 8:47 AM
> To: ag-tech at mcs.anl.gov
> Subject: Re: [AG-TECH] AG on wireless
> 
> Ag-tech always has such lively discussions. My question stemmed from
the
> 
> Advanced Networking with Minority Serving Institutions (AN MSI)
program 
> that has a little money to fund some ag node testbeds at MSIs. The
> question 
> of wireless came up for one of the schools. I wanted to check with the

> experts as I don't want the testbed to be too far out there - it
> actually 
> needs to work. Bottom line -sounds like it will suck up too much
> bandwidth. 
> Once again thanks to everyone.
> 
> 
> 
> Allison Clark, Ph.D.
> Assistant Director, Digital Equity Initiatives
> NCSA/The Alliance
> 152 CAB
> 605 E. Springfield Ave.
> Champaign IL 61820
> Phone: 217-244-0768
> Cell:   217-493-8935