[AG-TECH] Question about Codec
Brian Corrie
brian.corrie at newmic.com
Wed Aug 6 15:26:53 CDT 2003
Hello all,
> This is being tackled currently, by our (very quietly being worked on)
> network services work. However, I know Brian Corrie has some
> related work
> happening at NewMIC.
Yes, we are working on some solutions in this area...
> Our work is very early, but we're hoping to have something to
> tinker with
> soon. It's based around making the capabilities that the venue client
> uploads to the venue much more rich. The venue is already
> doing a match
> based on capabilities, but we have someone working on a more
> complex generic
> matching engine that can be plugged in. It will have
> configurable Rule Sets
> so you could upload a rule set that specifies your
> preferences (lower bw,
> lower latency, audio priority, etc) and get a match based on that.
>
> There's a lot of work to do, but we're making good progress
> on this and it's
> looking *extremely* promising so far :-)
One of the things I promised to do for the GGF ACE WG is to start a
discussion on how one can describe services for just such an application.
Unfortunately, work and the critically important holidays I just got back
from got in the way 8-). Hopefully we can start up that discussion from here.
> > -----Original Message-----
> > From: owner-ag-tech at mcs.anl.gov
> > [mailto:owner-ag-tech at mcs.anl.gov] On Behalf Of Osland, CD (Chris)
> > Sent: Wednesday, August 06, 2003 10:46 AM
> > To: ag-tech at mcs.anl.gov
> > Subject: RE: [AG-TECH] Question about Codec
> >
> >
> > Can I pick up on the idea of
> > "> i think allowing for varying levels of bandwidth capability
> > > of different sites
> > > to allow more people to participate is a different challenge
> > > that should be
> > > addressed independently."
> >
> > As far as I am aware, this would be an enormous challenge
> > [but I'd love to find out it's already been tackled!]
Yes, this is very difficult, but it can be solved in a number of ways. I see
the road to a solution for this problem to be through a combination of new
network protocols (such as those described by Bob Olson actually being
adopted, deployed and used 8-) as well as making use of advanced
collaboration services. In AG speak I think these would be network services.
We have built (still early stages yet) a filter/transcoder service that acts
like a QuickBridge except that instead of just passing along all video/audio
feeds it allows the client to select m of n video feeds. Such a service would
typically exist at the edge of a low bandwidth network (wireless, DSL, Cable)
and clients would connect when on a bandwidth challenged network.
The system is also a transcoder, and can convert video streams to other
codecs (presumably lower quality) for low bandwidth connections but we have
not yet done much experimentation with that aspect yet. In particular we
haven't explored the impact of the added latency of the transcoder on the
collaboration as of yet...
> > Let me demonstrate through a scenario currently playing out
> > in the UK. A high percentage of the academic community have
> > connection at serious bandwidth to SuperJanet 4 - say 650
> > Mbaud or more. Quite a lot of the smaller research
> > institutes may only have 2 or 8 Mbit and are currently
> > looking for the money to upgrade to 34 Mbit or more.
> > Bandwidth costs in the UK!
> >
> > At a completely parochial level, I do 40% of my work located
> > at RAL (Gbit or more connection) and 60% at my father's house
> > (128 Kbit/sec ISDN 2B+D). Even there I would love to join in
> > on AG events, at least contributing my video stream, and
> > maybe seeing a chairman, if there is one. At present I could
> > only do this either by killing off video completely, or by
> > going via dial-up audio only into the AP10 at (say) RAL.
> >
> > AFAIK, there is no way to select which video streams should
> > go to any site if they are multicast, thereby saving inbound
> > bandwidth.
The video filter/transcoder would allow this, but it does require someone to
manage it. This is the same situation as the QuickBridge. In an ideal world,
there would be a set of these that would be available as services in the AG
world that one could do service discovery on find the apporpriate "service
near you".
By the way, we quite often use full AG capability over 802.11b wireless
connections and if signal strength is good it can provide full access to a
relatively decent sized meeting (several multi-video sites) without causing
packet loss (although it does tend to be relatively unfriendly to other users
and vice versa). In addition, I often use my home Cable connection (which is
similar to DSL, although my cable connection seems to be a good one) for AG
work. One has to tweak things carefully with the cooperation of the remote
sites to make it good quality. The transcoder described above removes these
problems at the cost of having the user choose which video streams to view.
Ideally, this would be automatic, based on who is speaking or some similar
heuristic (similar to what the H323 systems do).
Cheers,
Brian
> > The requirement as I see it is for a given site to say that
> > it is not interested in receiving particular streams (rather
> > than the opposite which would be for a site to be able to
> > restrict the sites it is sending to).
> >
> > I don't know whether anyone has had thoughts on this; I do
> > realise that the main point of the AG is that everyone sees
> > everyone else and this is a facility that is in the opposite
> > direction!
> >
> > Cheers
> >
> > Chris
> >
> >
> >
> > ____________________________________________________________________
> > Chris Osland Office tel: +44 (0) 1235 446565
> > Digital Media and Access Grid Medialab tel: +44 (0) 1235 446459
> > BIT Department Access Grid room tel: +44 (0) 1235 445666
> > e-mail: C.D.Osland at rl.ac.uk Fax: +44 (0) 1235 445597
> >
> > CLRC Rutherford Appleton Laboratory (Bldg. R18)
> > Chilton, DIDCOT, Oxon OX11 0QX, UK
> >
> > [The contents of this email are confidential and
> > are for the use of the intended recipient only.
> > If you are not the intended recipient do not take
> > any action on it or show it to anyone else,
> > but return this email to the sender and delete your copy of it.]
> >
> >
> >
> >
> >
> > > -----Original Message-----
> > > From: Gurcharan.S.Khanna at Dartmouth.EDU
> > > [mailto:Gurcharan.S.Khanna at Dartmouth.EDU]
> > > Sent: 06 August 2003 16:16
> > > To: bdr at internet2.edu; judson at mcs.anl.gov
> > > Cc: dwkim at cosmos.kaist.ac.kr; ag-tech at mcs.anl.gov
> > > Subject: Re: [AG-TECH] Question about Codec
> > >
> > >
> > > hi,
> > >
> > > i'd like to second what bob would like to do with DVTS. using
> > > DV is really
> > > great because there is no codec involved, no compression
> > needs to take
> > > place. the DV stream is simply packetized and sent out and
> > > the receiving
> > > system just has to render the DV stream which doesn't seem to take
> > > much cpu power at all. literally, all one needs is a firewire
> > > DV cam and a
> > > pc with firewire in and free software. it can't get much
> > > easier and cheaper
> > > than that. [i've been running a multicast DV stream reliably
> > > for the past 6
> > > months or so. and now hardware cards are available as well]
> > >
> > > now all we need is the multicast user interface and we have
> > DV quality
> > > video and audio in an AG setting. of course, some will say
> > that 30Mbps
> > > is too big a penalty to pay for this simplicity. but big
> > > bandwidth should be a
> > > non-issue for us; at least it should be a challenge and not a
> > > thing to avoid.
> > > i think allowing for varying levels of bandwidth capability
> > > of different sites
> > > to allow more people to participate is a different challenge
> > > that should be
> > > addressed independently.
> > >
> > > just my opinion,
> > >
> > > -gurcharan
> > > ===========================
> > > Gurcharan S. Khanna, Associate Director
> > > Research Computing, Dartmouth College
> > > gkhanna at dartmouth.edu
> > > office: 603-646-1644
> > > http://research.dartmouth.edu/ http://www.dartmouth.edu/~gkhanna
> > >
> > > --- Bob Riddle wrote:
> > > I guess I might as well confess I'm hoping to do some DVTS
> > > work before
> > > the end of the year ...
> > >
> > > Ivan R. Judson wrote:
> > >
> > > >Additionally, we are looking into higher quality codecs,
> > both Larry's
> > > >OpenMASH and some other more informal work hopefully
> > > involving Colin and Co
> > > >at ISI. OM has a new H.263 codec that works very nicely, but
> > > the current
> > > >agvic has a bug that causes it to crash when it sees an OM
> > > H.263 stream, OM
> > > >also has as bob pointed out MJPEG support.
> > > >
> > > >The other work involves potentially taking recently
> > > available prosumer
> > > >cameras that produces a MPEG2-TS stream via firewire at HD
> > > resolutions (so
> > > >the specs say :-), grabbing the data and streaming via vic
> > > shouldn't be
> > > >insurmountable; decoding might take a bit of horsepower, but
> > > at least it's
> > > >MPEG2 (decoders for which are seemingly embedded in most
> > > graphics chips
> > > >now).
> > > >
> > > >Another place to consider looking it Linux Media Labs MPEG-4
> > > boards, with
> > > >drivers for mplayer. We don't have this working, but would
> > > be happy to help
> > > >you make it work.
> > > >
> > > >Cheers,
> > > >
> > > >--Ivan
> > > >
> > > >
> > > >
> > > >>-----Original Message-----
> > > >>From: owner-ag-tech at mcs.anl.gov
> > > >>[mailto:owner-ag-tech at mcs.anl.gov] On Behalf Of Bob Riddle
> > > >>Sent: Wednesday, August 06, 2003 8:11 AM
> > > >>To: dwkim
> > > >>Cc: ag-tech at mcs.anl.gov
> > > >>Subject: Re: [AG-TECH] Question about Codec
> > > >>
> > > >>
> > > >>If you use the openmash vic, you can use RtpTV to generate
> > > >>mjpeg streams that the openmash vic will interpret. You'll
> > > >>need very capable cpus on the AG display machine.
> > > >>
> > > >>dwkim wrote:
> > > >>
> > > >>
> > > >>
> > > >>>Hi all:
> > > >>>
> > > >>>We are wondering that there is any other codec not h.261 for
> > > >>>example, Can We use DVTS(digital video transport system)
> > > >>>
> > > >>>
> > > >>or HDTV in
> > > >>
> > > >>
> > > >>>AccessGrid ?
> > > >>>
> > > >>>Thanks for any information
> > > >>>
> > > >>>Daewon kim
> > > >>>
> > > >>>
> > > >>>
> > > >>>
> > > >>>
> > > >>>
> > > >>--
> > > >>Bob Riddle (bdr at internet2.edu) Technologist,Internet2
> > > >>3025 Boardwalk, Suite 100 Ann Arbor, Michigan 48108
> > > >>Business Phone: 734.913.4257 Fax Number: 734.913.4255
> > > >>
> > > >>"Opportunity is missed by most people because it is dressed
> > > >>in overalls and looks like work" Thomas Edison
> > > >>
> > > >>
> > > >>
> > > >>
> > > >>
> > > >
> > > >
> > > >
> > > >
> > >
> > > --
> > > Bob Riddle (bdr at internet2.edu) Technologist,Internet2
> > > 3025 Boardwalk, Suite 100 Ann Arbor, Michigan 48108
> > > Business Phone: 734.913.4257 Fax Number: 734.913.4255
> > >
> > > "Opportunity is missed by most people because it is dressed
> > > in overalls and
> > > looks like work" Thomas Edison
> > >
> > >
> > >
> > > --- end of quote ---
> > >
> >
>
>
More information about the ag-tech
mailing list