[opendtv] Re: Twang's Tuesday Tribune (Mark's Monday Memo)
- From: Craig Birkmaier <craig@xxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Tue, 11 May 2004 10:21:02 -0400
At 10:16 PM -0400 5/10/04, Kon Wilms wrote:
>Oh please. People have been crowing about MPEG4 object layers for what,
>almost 6 years. I have yet to see anything widespread materialize that
>is anything but a glorified one-off show-floor demo.
>
This is not surprising given the realities of the television business
today and the history of the MPEG standards process. MPEG-1 and
MPEG-4 can be viewed like bookends around the MPEG-2 standard.
MPEG-1 was developed largely without the support of broadcast
interests. Philips was involved because of its interest in recording
video on CDs. The standard was strongly influenced by academic
interests and several large telecommunications companies. At the time
it was developed, the major contribution of MPEG-1 was to prove that
the core concepts - the DCT transform with Huffman coding and motion
compensated prediction - could do the job, achieving relatively high
compression ratios. But MPEG-1 was all FRAME based, and almost all
existing video formats used interlaced fields.
When the ISO began the MPEG-2 process the plan was to develop a
standard to handle Standard Definition video. MPEG-3 was designated
as the process to develop a standard for High Definition video. The
process to develop a U.S. standard for broadcasting HDTV turned
digital about the same time that the MPEG-1 standard was being
finished. The proposed system being developed by Sarnoff/Thomson,
used an enhanced version of MPEG-1 (they called it MPEG-1.5); some of
the concepts (intellectual property) used in this system made their
way into the MPEG-2 standard.
The real importance of MPEG-2 is that the ISO process was strongly
embraced by the leading video equipment manufacturers around the
world. The tools that made it into the MPEG-2 standard involved a
pile of new patents, many of which deal with the encoding of
interlaced video. Most of these new patents are just minor tweaks on
techniques that were or would soon move into the public domain. As
the standard was created it became obvious that the MPEG-2 tools
could handle both Standard Definition and High Definition video - SD
is covered by Main Level, HD is covered by High Level. The need to
develop an MPEG-3 standard became unnecessary.
Work began on MPEG-4 soon after the MPEG-2 standard was finished. But
the video equipment companies that poured so much money, time and
effort into the development of MPEG-2 pulled back; they primarily
participated in MPEG-2 to make certain that there would not be an end
run around MPEG-2. The driving force behind MPEG-4 became companies
that needed much more aggressive compression techniques that are
possible with MPEG-2, which ONLY deals with the encoding of complete
image rasters. At one end of the spectrum, telcoms and other
companies interested in very low bit rate encoding put their support
into MPEG-4; these included France Telcom, Franunhoffer and others
who were looking at wireless telephones as a delivery platform for
video and other forms of digital media content.
At the other end of the spectrum, companies involved in "multimedia"
and the delivery of new forms of digital media content via the Web
and packaged media put their support into MPEG-4; these included
companies like Apple, Microsoft, Intel, etc. Thus MPEG-4 developed a
split personality with the ability to handle very low bit rate and
very high quality digital media. But there was almost no commitment
to using MPEG-4 by the companies that commercialized MPEG-2. To date,
the only major video product to come out of the initial MPEG-4 effort
is the Sony HDCam SR recording system with is based on MPEG-4 Part 2
video studio profile. The Adcanced Audio Coding portion of the MPEg-4
standard is being deployed in a number of areas; it is the basis for
audio coding of the Apple iTunes music service.
But the visual composition tools have not been exploited yet. There
are several reasons for this. The first is complexity; there are many
tools that must be supported in order to properly develop an end to
end system, and the player complexity is much greater than that
needed for a dumb MPEG-2 video decoder. The second is more important.
Let's call it it Techno-political.
This is the world that we discuss on a daily basis on this list. A
world controlled by a handful of big media conglomerates and video
equipment vendors who have little interest in changing the business
model that has put them on the top of the heap. A perfect example is
the DVD consortium, with its very limited tools for authoring of DVD
discs. The real goal here was to deliver movies, not a full spectrum
of digital media titles that could take advantage of the tools in
MPEG-4.
As Kon knows only too well, there has been tremendous resistance
among "his" customers to put any resources into a set top box, other
than those that are absolutely essential. The goal has been to drive
down the cost of these boxes, not to evolve their capability. This is
now changing, thanks in large part to the needs of the DBS services,
who need the ability to send information efficiently to millions of
STBs, where it can be composed to deliver localized and customized
services.
So the issue becomes one of what standards to support for these
services. While MPEG-4 has been lying dormant, many other contenders
have emerged. For example, FLASH has become an important standard for
media composition in the web space. QuickTime offers many of the same
tools that are in MPEG-4, and they are used extensively. In fact, the
MPEG-4 file format is an extension of the QuickTime file format.
Now that the resources needed to implement the tools in MPEG-4 are
becoming commonplace and affordable, it is likely that we will see
the entire concept of localized media composition start to take
off...and once again, terrestrial broadcasters will be left behind,
having chosen standards that constrain their ability to compete.
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: