RealTime IT News

Dirac from the BBC

With ambitious plans to shake up the video-compression space, the research arm of the British Broadcasting Corporation (BBC), which is widely hailed for its research work in the areas of digital television, digital radio and multimedia technology, developed a prototype video coding algorithm based on wavelet technology .

Seeking open source help to turn it into a commercial product, the BBC's R&D unit released the Dirac general-purpose video codec to developers on SourceForge, banking on its capability of being "competitive with other state-of-the-art codecs."

The codec, named after British math genius Paul Dirac, combines the use of wavelets, motion compensation and arithmetic coding, and it can achieve a two-fold reduction in bit rate over the widely deployed MPEG-2 format.

Wavelets are mathematical functions used to compress digital images and the BBC boasts that the new approach to high-quality video compression is different from that used in the main proprietary or standard video compression systems, typically from tech heavyweights like Microsoft , Apple and RealNetworks .

Although a note posted on the BBC R&D Web site claims the codec can also be optimized for Internet streaming resolutions, Dirac remains simply an early-stage research tool in need of major development.

The BBC admitted as much when it released the source code to the open source community. "A lot remains to be done to convert our promising algorithm and experimental implementation into practical useable code. This includes optimization so that it can decode in real time. Algorithmic enhancements are needed to improve the compression performance still further," the company said.

"The resulting codec needs to be integrated with other parts of a compression system, such as players and interfaces using standard IO formats. We welcome help and support in creating an open and freely available compression system based on this technology."

Even as the BBC is pushing to use an open source codec to tilt the scales away from the established formats offered by Microsoft, RealNetworks and Apple, digital media analysts are expressing doubts.

Yankee Group analyst Mike Goodman believes a new video compression codec would find it really tough to win adoption, particularly among content producers.

"It is pretty hard to establish a new codec in today's market, even if you go the open source route," said Goodman in an interview with internetnews.com.

"Unless there's something about this codec that offers significant improvement over other available codecs, why will anyone migrate? You already have [RealNetworks'] Helix out there as an open source video codec, so this is not something new or unique. If it's just an incremental improvement, companies aren't going to make another investment in a new format. It will be a very tough road to hoe for the BBC.

"The other issue," continued Goodman, "is whether devices will support this codec. If nothing supports it, what use is it?"

Jupiter Research analyst Joe Wilcox believes there may be room for competition in the area of video compression, but, like Goodman, warned that technology offered by existing companies, particularly Microsoft, was already enjoying widespread adoption at the enterprise and consumer levels.

Wilcox, who tracks the digital media marketplace for the Microsoft Monitor Weblog, said the BBC's Dirac would have to offer significant improvements over existing formats if it is to find a niche.

"There is still room for competition because of how segmented the market is," said Wilcox. "Video streaming has gone beyond the Web and onto cell phones and wireless devices. And there may still be opportunities for newcomers, because this is still a very young market."

Still, by releasing the code on SourceForge, the BBC might be able to drum up enough interest from volunteer developers to make it work, Wilcox said.

"The BBC name doesn't hurt. A BBC effort that is truly open source could be very appealing, especially in Europe where there are strong feelings about not letting an American company become too dominant."

Yankee Group's Goodman agrees.

"The fact that this came from the BBC gives it some legitimacy. A lot of these open source efforts fizzle out, so in terms of a potential lifespan, Dirac has a fairly robust backer."

Still, Goodman thinks it will be quite a stretch to cut into Microsoft's dominance.

"Microsoft is giving away a lot of the technology for content owners to encode their offerings. The content guy will tell you they're already paying nothing. Unless this product is technically superior and blows everything away, the content guys aren't going to bother with it.

"It's a catch-22 when you're dealing with Microsoft," Goodman continued. "On one hand, you're dealing with the 80-pound gorilla. And, on the other hand, you have to give Microsoft credit for doing a very nice job developing their codecs and DRM."

Despite the doubts, the BBC is plodding ahead.

"The philosophy behind the Dirac codec is 'keep it simple,'" BBC R&D said. "This is an ambitious aim since video codecs, particularly those with state-of-the-art performance, tend to be fearsomely complex. [We] would like to collaborate with the open source community, academics and others to produce an open codec. It is therefore important for us to keep the principles and design as simple as possible and to provide copious documentation."