The broadcast industry will begin reaping the benefits of more efficient compression thanks to MPEG-4 AVC H.264.
With about half the bit rate requirements of its predecessor, MPEG-4 AVC H.264 promises in the short run to benefit broadcasters in straightforward ways like cutting SNG backhaul space segment charges. In the medium to long-term, it promises to usher in new acquisition, editing and distribution alternatives that may change the fundamentals of the broadcasting business.
At the core of many initial MPEG-4 AVC H.264 encoder products is a codec technology from LSI Logic in Milpitas, CA. High Definition Technology Update paid a recent visit to the company’s headquarters to discuss the potential for MPEG-4 AVC H.264 with Bob Saffari, LSI senior director of marketing and business development.
HDTU: Going back several years, the broadcast market offered LSI a way to gain experience and knowledge in terms of video quality. How important was that? And, how important will it be to HD?
Bob Saffari: It was extremely important. It was paramount to realize how quality impacts the decision-making process from the broadcast sector all the way to the consumer level. What we learned back in those days working with Hollywood and people in the industry helped us to produce the highest quality video for encoding, authoring and broadcast solutions. Once the DVD from studio professionals created the desire to do cinema quality content, consumers wanted their content to be as good as what came from other providers.
Applying the same principles and embedding special parameters in our software tools allowed us to enable consumer DVD recorders that could produce video quality similar to authoring platforms but for personal content creation and archiving.
With HD, the challenges are far greater. Much more data needs to be analyzed and processed for compression, multiple standards and newer formats must be supported, and above all video artifacts have nowhere to hide anymore. They are much more visible with larger screen HD displays.
For example, in the case of EVD in China (the HD Red Laser DVD player standard in China), with almost four to six times more data than SD TV resolution, we had to apply new techniques to get MPEG-2 HD content to fit on a red laser disk. EVD’s success heavily depends on the availability of high-quality HD content. This was critical to enable post-production houses in China to convert their local HD content.
In parallel, the places where HDTV is gaining momentum, such as in the U.S., Japan and Korea, video quality becomes even more of an issue. Given the trends for adoption of larger displays for HD (greater than 32in in 2005), it is like having a magnifying glass. The artifacts become more visible.
Therefore, understanding the limitations of the standards provides ample opportunities for creative solutions to be developed not only at the headend but also inside the DTV and consumer set-top boxes. With consumers watching sports events in high definition, HD has to shine now more than ever. It’s extremely important to create a solution that the consumer enjoys.
Last but not least, with HD we should be looking at supporting multiple compression formats, not only within the broadcast and content creation side but also within the consumer space. For example, for HD-DVD and Blu-ray DVD content creation, the challenge is to not only provide an MPEG-2 HD but also H.264 and VC1 HD encoding solutions. Consequently, multi-format encoders with far greater flexibilities are required.
Therefore, working with professional and broadcast players in the market will play a key role for us in addressing some of these tough technical and industry challenges.
HDTU: What is the future path for H.264 in terms of compression efficiency? Is there room for greater compression efficiency? Will it primarily be used for distribution and transmission applications, or will it find its way into acquisition and post?
BS: What’s exciting to see is how MPEG started. We are right at the beginning of leveraging this new, great algorithm. There is tremendous opportunity for algorithm development in reducing the bit rate while maintaining high-quality video. LSI is pioneering in it and will continue to do so. The innovation will continue in this area for years to come.
When it comes to transmission and acquisition, the answer is yes and yes. We do see an opportunity for such technology in both acquisition and the distribution market. However, acquisition will lag a little behind transmission.
Acquisition with H.264 has tremendous advantages for HD. It will be used at a professional level and come down to the consumer. Two requirements will be key: low-power consumption and preservation of high-quality video.
In post-production where the task is creation, the CPUs available today in a workstation are capable of MPEG-2 SD encoding, decoding, blending and essentially doing editing functions. But for the foreseeable future, special processors will be required for encoding for H.264 in SD and HD. The value of having high-quality encoders in post-production will remain. Software running on CPUs to perform HD H.264 encoding is still years away.
HDTU: As stations look to convert production to HD, one of the biggest areas of expense and concern is field acquisition – or ENG. What solutions from LSI will be driving efficient HD field acquisition?
BS: Although our previous generation multi-standard DV and MPEG-2 codecs were used in a few professional camcorders with a hard disk attached, with respect to newer tape-based formats (MPEG-2 HD standard) we currently do not have a solution.
However, I do believe HDV for the next few years will play a key role in reducing the cost of ENG and field acquisition when it comes to HD. The H.264 codec is more computationally intensive (especially in HD) and obviously is more power hungry; therefore, it will take a while before it gets adopted as a codec inside the HD camcorders and ENG.
On the other hand, for inter-studio distribution and contribution, the cost of bandwidth associated with HDV is too high (25Mb/s). I think you will see H.264 come to play as transcoding devices initially.
In terms of satellite newsgathering, H.264 will have a better chance and more value from that aspect to all of the constituent players in the market. Although it initially may be a little more expensive, on a price per channel basis H.264 will come down as more powerful and integrated products become available providing all of the same quality at half the bit rate.