[AG-TECH] broadcast quality video

[Ivan R. Judson]

These are very cool technologies. I'd like to know if anyone is
currently working on implementations of these codecs or has pointers to
implementations? There are brave souls lurking about who might consider
adding them to vic or other applications --

http://www.mplayerhq.hu/homepage/ has many supported codecs. It'd be
interesting to consider some of them for AG use.

Fwiw, it'd be an incredibly handy reference to have some kind of codec
comparison chart that shows the salient features of each codec we're
considering which might help drive us to clearer focus on some codecs
rather than others. Anyone want to start one?

--Ivan

..........
Ivan R. Judson .~. http://www.mcs.anl.gov/~judson
Futures Laboratory .~.  630 252 0920
Argonne National Laboratory .~. 630 252 6424 Fax
 

> -----Original Message-----
> From: owner-ag-tech at mcs.anl.gov 
> [mailto:owner-ag-tech at mcs.anl.gov] On Behalf Of Markus Buchhorn
> Sent: Monday, November 04, 2002 5:08 PM
> To: ag-tech at mcs.anl.gov
> Subject: RE: [AG-TECH] broadcast quality video
> 
> 
> At 09:25 AM 5/11/2002 +1100, Christoph Willing wrote:
> >Just supporting Chris Osland's assertion about the H261 
> being more of 
> >an issue than the cameras themselves. We have 2 BetacamSP cameras 
> >available but, after passing through vic, they look no 
> better than $200 
> >security cameras.
> 
> since h.261 is a CIF (and QCIF) codec only, that's not surprising.
> 
> >What does work is RPTtv, described at the AG retreat earlier 
> this year 
> >(http://bmrc.berkeley.edu/~delco/rtptv/). We have 2 labs 
> (each with an 
> >AG node) separated by about 5km but connected with gigabit 
> ethernet. We 
> >have full PAL video streaming from one AG node to the other and the 
> >quality is just astonishing compared with normal vic 
> streams. Quite a 
> >bit more bandwidth as well of course.
> 
> For sure - we used the old Fore streamrunners with MJPEG/ATM 
> between various sites, and it is superb, at that bandwidth. 
> There's no inter-frame compression, hence the bigger pipe need.
> 
> The compromise between these two is h.263 (the most recent 
> version - 2000?) which in one form or another is also used in 
> mpeg4 and divx. It supports (almost) arbitrary frame sizes 
> (8x8 multiple), and is very bandwidth efficient, in fact more 
> so than h.261. I don't know if any version of vic supports 
> it? I've used a package developed at Keio University which is 
> mpeg4 based, and it is very very nice - at very reasonable 
> bandwidth (less then 1Mb)
> 
> An alternate codec is mjpeg2000 - i.e. based on jpeg-2000 
> which is a wavelet codec, rather than the older jpeg which is 
> DCT based. mjpg2k has similar bandwidth needs to h.263 
> (perhaps not quite as low), and is certainly much narrower 
> than "normal" mjpg. It also does no inter-frame compression, 
> which has its advantages in the face of packet loss. Morgan 
> Multimedia has some plugins for it.
> 
> Cheers,
>         Markus
> 
> 
> Markus Buchhorn, ANU Internet Futures Project,        | Ph: 
> +61 2 61258810
> Markus.Buchhorn at anu.edu.au, mail: Bldg #108 - CS&IT   |Fax: 
> +61 2 61259805
> Australian National University, Canberra 0200, Aust.  
> |Mobile: 0417 281429
>