The Impact of Advanced Video Coding

Video compression technologies will change the models we've grown to embrace since the introduction of 500-channel cable and direct-to-home satellite broadcast.
Publish date:

Video compression technologies will change the models we've grown to embrace since the introduction of 500-channel cable and direct-to-home satellite broadcast.

The benefits will include improvements to technologies for ordering movies and live events on-demand, replaying or delaying a program, providing for stunt features (still, rewind, pause) and even recording HD on one channel while viewing another and still retaining many of those stunt features.


The platform for receiving program content is going beyond the rental set-top box to desktop and mobile platforms. The systems to support the new platforms are made possible through improvements in compression technologies and delivery systems. The deployment of these new capabilities is being accelerated by well-known industry heavyweights, albeit not ones we previously might have expected to see operating in the media and entertainment space.

There will be new architectures from which to prepare, aggregate and distribute media content, made possible in part by the reduction of the encoding bit-rate for moving images, and by using a well-established transport mechanism deployed via larger bandwidth pipes. Transport mechanisms will include fiber-to-the-home (FTTH) or fiber-to-the-node (FTTN) and will use Internet Protocol (IP) for carrying compressed digital media packets in SD and HD.

The combination of these concepts creates the mechanism for a new paradigm in digital media.

Prime drivers behind these advanced coding capabilities include AVC and VC-1. Both new coding technologies are already changing the face of the digital medium, each offering its own set of values and opportunities into the professional and consumer marketplaces.

AVC, a next-generation extension of the MPEG-standard toolsets, targets a much broader set of opportunities than MPEG-2, introduced in 1994. Not only will AVC provide advancements for digital television, it is clearing the road for interactive multimedia and graphics in the media marketplace.

AVC's precursor, MPEG-4, was introduced in 1999 and demonstrated a 20- to 25-percent improvement over high-end MPEG-2 encoding. MPEG-4 led to the adoption in 2002 of H.264/MPEG-4, Part 10 Advanced Video Coding (AVC), a much more aggressive coding technology.


By virtue of the continual increases in computer power, new compression capabilities have become more efficient and less costly to deploy. AVC is conceptually dedicated to maintaining backward compatibility with MPEG-2, thus leveraging the heavy investments in hardware already in satellite, cable and broadcast facilities. With new features in AVC, the limitations of MPEG-2 have now been realized; for example, the creation and deployment of higher resolution motion imaging.

For comparison purposes, HD for home use still requires between 14 and 19 Mbps in MPEG-2, generally delivered via cable, satellite or ATSC terrestrial broadcast. Already, the capabilities of AVC encoders have shown as much as a 60-percent reduction from the same MPEG-2 bit-rates, bringing that same HD signal down to around 6 or 7 Mbps; a far more acceptable rate for all ranges of the cable, satellite and telco service providers, as well as a viable solution for digital cinema.

However, AVC was not the only new development on the digital horizon. September 2002 brought another dimension to digital video compression at IBC in Amsterdam. Microsoft delivered Windows Media 9 (WM9), a serious contender for the AVC domain. The coding technology inside WM9, which became the basis for VC-1, provided substantial improvements over MPEG-2.

For comparison, a standard-definition image (480/24p-by-8-bits) compressed between 4 and 6 Mbps in MPEG-2 will, with VC-1, require only 1.3 to 2 Mbps for the same content. For HD (720/24p-by-8-bits), the compression savings from MPEG-2 at 19 Mbps to VC-1 at 5 to 8 Mbps is as much as 3.8:1. We should point out, however, that these data rates are content-dependent.

The next step to widespread acceptance for VC-1 was at SMPTE's C24 engineering committee level. For more than a year, C24 has been reviewing the potential adoption of VC-1 as a standard (see "VC-1 Goes to Ballot," Jan. 19, 2005), even though the DVD Industry Forum has already endorsed both the AVC and VC-1 as approved standards for next-generation DVD formats. With more than 100 manufacturers having announced products based around the VC-1 encoding platform, these new technologies are currently paving the way for another next-generation deployment of both media over wired and unwired connections.


What this means for media servers is a potential quantum shift in storage, coding and distribution methodologies. One area that will change is in content aggregation and distribution. While there are obvious differences in cable and satellite delivery, both share common methodologies that have matured to the point that a tapeless environment is a given.

For the professional, high-reliability video servers, with their rather strict set of the MPEG-2 data encoding principles, represent a migration from the silicon card-based encoder/decoders that maintain broadcast-quality imaging. It is possible that emission-level encoding in AVC or VC-1 may become the norm for standard-definition imaging, leaving MPEG-2 and others for use in production or content assembly.

The perspective for content aggregation and delivery is already shifting, as demonstrated by EchoStar's DISH Network and SBC with its Project Lightspeed announcement. Both companies intend to use advanced coding. Whether for conventional satellite or IPTV (Internet Protocol TV), the ability to deploy other-than-MPEG-2 encoded content is very high on the agenda.

What value have we seen in these shifts in coding technologies? When considering the transport of linear programming, including stunt features on playback; integrating AVC and/or VC-1 makes practical economic sense, since the hardware required for recording, storage and playback is far more easily realized with conventional computer-based server products.


Streaming contiguous media files as data from a storage platform to a network, such as in IPTV, can be more easily achieved given the inherent buffering capabilities that can be built into the receiver/decoder platforms. Such buffering agents will allow for the smaller file structures of AVC or VC-1 to ride through discontinuities in a smooth fashion.

Adding the bidirectional signaling capabilities of IP delivery, the amount of error correction and the slewing of program streams may indeed create a more pleasurable viewing experience than today's MPEG-2-only delivery platforms.

From a distribution perspective, there will continue to be different models, depending on the infrastructure and the carrier. For example, for the feature-rich capabilities, cable already offers a fixed back channel that allows ordering and control of programming streams. For satellite, the requirement for a secondary, wired low-bandwidth back channel has limited the on-demand perspective. It should also be noted that low-bandwidth satellite backhauls are emerging. Provided the systems are not overloaded, this will allow for a wireless return path, enabling an elevated set of features for PPV or VOD.

When these various technologies, AVC, VC-1, FTTH/FTTN and IPTV are all set side-by-side, the emerging winner becomes the user, who will be provided a richer and more integrated media experience. How we experience media, and how the facility will need to adapt to the new paradigm, will continue to affect the industry in profound ways. It will be most interesting to see where ATSC finally will fit as the next-generation media experience unfolds.