[opendtv] Re: New Sony COO bullish on Blu-ray
- From: Craig Birkmaier <craig@xxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Thu, 18 May 2006 08:26:48 -0400
At 1:55 PM -0700 5/17/06, Kilroy Hughes wrote:
>I agree there's no threat of critical mass home networking of CE
>devices, etc. soon based on the assumption you're talking about "Media
>Networks", but I draw a different conclusion based on IP "Data
>Networks".
Kilroy makes some good points in this post, however, his conclusion
is still much the same. networking CE and related (cable/DBS) devices
is going nowhere, while there is critical mass growing for the
sharing of content over IP networks.
>IP data networks are in around a quarter of US households, and they are
>a majority of the households with enough CE toys to network. These data
>networks are marginal for compressed realtime SD video delivery, and
>worse for compressed HD delivery. Getting a bunch of DLNA devices
>sharing a wireless network, transcoding to standard MPEG-2 and audio
>codecs, image formats, bitrates, transport muxes, RTP/RTSP protocols,
>packet prioritization, doing discovery and negotiation, etc. I would
>kindly call a "long term project".
A good analysis. These networks can get the job done, most of the
time, but they are not up to the task of simultaneous streaming of
compressed bits to multiple devices in the home. By the way, the same
is true for the growing market for media networks in digital media
workgroups. in these professional markets, companies like Avid and
Omeneon are selling workgroup solutions that use up to four gigabit
E-net channels per workstation to help balance the flow of bits to
and from the servers, and to get around blocking problems.
I disagree with Kilroy to some extent about the "long term" aspects
of making all of this stuff work together in the home. No doubt this
is true for MS, however, virtually all of this stuff is already in
place and working for OS-X.
>
>That theory of the universe is the "push model", where server devices
>generate realtime video streams in a standardized wire format and client
>devices are dumb video decoders. Network performance demands are very
>high (low latency, no packet retransmission, etc.).
Let's just call this the universal view from the Content Protection Racket...
See CPR Redefined:
http://broadcastengineering.com/mag/broadcasting_cpr_redefined/
>An alternate theory of the universe is the "pull model", where files are
>read over the network using either a network file system or the Internet
>file system HTTP:GET(URL; byteRange). If there are PVRs with gigabytes
>of TV shows, optical disc with movies, music collections, photo
>collections, etc. sitting on hard disks, a client device that
>understands those file types can find them and read them over the data
>network and decode them. If you bring a laptop or PPC phone within range
>of your WiFi, it's simple to play the TV shows, songs, and photos
>sitting on your home computer(s) as long as they're shared on the data
>network. CE devices (like HD DVD-V players, game consoles, dedicated
>media terminals) that have the decoders and can join a network can
>similarly read and play remote files (LAN, or WAN if HTTP). DRM
>permitting, this works today with data networks and has a simple
>incremental growth path. Client devices have to be smarter, to be full
>"players" for some media format(s), not just Program Stream decoders.
>There's little distinction with MP3 audio files or JPG photos, but a
>media format like DVD-Video makes a big difference.
Yes, this is the current reality. All possible today with OS-X and
Windows. It will only get better with Vista. As Kilroy points out,
however, existing data networks are only viable until multiple people
in a household start to access files at the same time. And the
ability to connect to program caches on CE, cable and DBS devices is
still mostly a myth.
it was interesting to see some of the capabilities that Dish Networks
hopes to implement in their next generation STBs. But once again, the
issue will be whether these products work in an open network
environment, or only with each other.
>
>In order to play a DVD-Video with menus, subtitles, choice of multiple
>audio tracks, different languages, aspect ratios, parental rating
>versions, etc.; you need a fairly sophisticated DVD-V player engine that
>can randomly access interleaved portions of Program Streams sprinkled
>around the disc and demux certain PIDs, execute display commands, etc.
>to create a decodable linear video stream; then that needs to be
>composited with other graphical overlays in decompressed YUV, scaled,
>cropped, etc. and changed interactively in response to user interaction.
>In push mode, to accomplish full DVD-V functionality, the server would
>have to be the DVD player, generate the composited and interactive video
>image, then re-encode it to the standard wire format (elementary stream
>encode, Program Stream mux, realtime protocol and encapsulation). In
>pull mode, the client device would be the DVD player, and the disc or
>disc image would just be a remote disc or disk drive on the data
>network. The same principle applies to interactive TV, games, Flash,
>Web pages, etc.
Yup. The push model makes little sense, unless you are a back lot
lawyer in Hollywood, desperately trying to protect your empire...
>
>Pull mode requires minimal network performance and only TCP/IP and file
>system level compatibility. The media types don't need to be understood
>by all source and sink devices; only a particular media type on a
>particular consumption device. The rest of the system will store and
>copy media files they don't understand without any need for
>compatibility.
Yup. This makes the Media PC or a Mac Mini look like the ideal client
device. But the reality is that virtually any device with some brains
can do the same thing. Just ask Richard H. about the similarities
between the next generation STBs he is working on and PCs optimized
as home entertainment appliances.
>
>Where you need a totally compatible wire format is between every
>consumption device and the "glass" and AVR. Uncompressed digital audio
>and video are the best solution for that short one-way trip.
>
Correct. In the past decade we have evolved from analog component to
DVI with HDCP, to HDMI. I'm certain that there are folks working on
something to obsolete HDMI at this very moment.
Regards
Craig
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts: