[opendtv] Re: A detail in the history of video standards

At 12:27 AM -0700 8/31/05, John Willkie wrote:
>In many places, MPEG was viewed largely as being off-shore.  I even hear
>people talking about it being a "French" outfit.  To which I ask "Leonardo
>Chiagrigniole (cq) founded a French outfit?"

It is very important to be clear about which MPEG standard we are 
talking about.

MPEG-1 was put together by a subset of the companies that put 
together the MPEG-2 standard. At the time it was developed, MPEG-1 
was viewed as a multimedia standard, enabling the playback off video 
on specially configured PCs and MPEG-1 enabled CD players. Philips 
played a major role in MPEG-1 as they were very interested in 
expanding the CD market. Sony and the rest of the Japanese CE 
companies were not major players in MPEG-1. OTher major players in 
ALL of the MPEG standards included Alcatel (The French Phone 
company); Fraunhoffer Institute (Germany); Columbia University (US); 
MIT (US); General Instruments which became Motorola (U.S.); and IBM 
(U.S.); and Sarnoff Labs (funded by US and international companies). 
Bell labs played a role early on until it was decimated.

MPEG-2 was a VERY DIFFERENT situation. By the early '90s GI had 
demonstrated the feasibility of encoding both SD and HD, proving that 
video compression was going to become an important technology for 
mainstream video production and broadcast operations. At this point 
the major CE companies jumped in with both feet and dominated the 
process. The fact that Leonardo was educated in Japan and speaks the 
language should not be overlooked - he had the ability to communicate 
at a high level with all of the participants. The MPEG-2 effort was 
well underway before the influx of participants and money form the 
video centric companies. There was a STRONG movement to make MPEG-2 
progressive only, because of the added efficiency of processing only 
frame based material and the added quality possible with the 
elimination of interlace.

The video companies had just completed "their" transition to digital, 
creating the entire "uncompressed" digital infrastructure around the 
international (and interlaced) ITU-R BT.601 standard. They perceived 
MPEG-2 as a major threat, and moved rapidly to co-opt the standard to 
their needs - i.e. the protection of interlaced SD.

Of particular note, much of the IP that was being considered for 
MPEG-2 was entering the public domain. A major component of the 
MPEG-2 standards effort was the re-invention of this IP to protect 
the potential royalties for another 17-20 years. Virtually all of the 
IP in MPEG-2 related to the coding of interlace was patented in the 
1993-1995 time frame.

The MPEG-3 standard was skipped, as it was designated for the 
encoding of HD. But the MPEG-2 standard incorporated this, so MPEG-3 
never happened. Leonardo and his minions then set about the task of 
developing MPEG-4, but they did not enjoy the same high levels of 
support and funding from the companies that dominate the MPEG-2 IP 
pool. The MPEG-4 standard was also viewed as a computer centric 
multimedia standard. MPEG-4 Part 2 video was evolutionary at best, 
providing only small (about 10%) improvements in compression 
efficiency over MPEG-2. The one area where there has been some 
interest from the video centric companies is in the extensions 
related to encoding high quality video - Sony's HDCam-SR is based on 
MPEG-4 part 2 video.

When I started to participate in the MPEG standards process in the 
late '90s, things were languishing. There was little industry support 
to do anything with the MPEG-4 standard, especially the very useful 
tools for real time object based composition. I started an effort to 
create a new profile for progressive video formats that eliminated 
the need for most of the new IP used in the MPEG-2 standard. I was 
getting some traction until the funding I was relying upon to 
participate in the MPEG process was cut off. Shortly there after 
Leonardo agreed to a codec shoot-out to determine if a new video 
compression algorithm should be added to MPEG-4. This led to the 
formation of Joint Video Team with the ITU, which led to the creation 
of MPEG-4 part 10 and ITU H.264, aka AVC. Once again, the companies 
holding most of the IP in MPEG-2 jumped in with both feet; it should 
come as no surprise that they were able to add most of the interlaced 
centric IP that they had created for MPEG-2 into the AVC standard.

Small wonder that Leonardo has been richly reward for his efforts as 
the Chair for the MPEG standard efforts.

Perhaps in ten years we may be able to put this mess behind us and 
finally eliminate the interlace footprint from SD and HD.

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: