Startup claims video compression breakthrough

A Silicon Valley startup claims it has found a new approach to video compression that blows away the commonly used MPEG-2 standard and will allow transmission of broadcast-quality video over DSL (Digital Subscriber Line) connections.

Pulsent Corp. said on Monday that after four years of work it is ready to come out of stealth mode and show off technology that it claims provides a 400 per cent improvement in bandwidth and storage efficiency over MPEG-2. The technique would allow TV-quality video to be transmitted over 1.1M bps (bits per second) connections, according to a Pulsent statement.

Enhanced video compression could drive new services such as video on demand and allow DSL providers to compete with cable television providers.

The Milpitas, California-based startup will preview its invention at the National Association of Broadcasters (NAB) event in Las Vegas next month. It faces tough competition even if it can deliver on its claims. Many of the companies it will have to persuade have already invested in other MPEG-2 successors, such as MPEG-4 and H.26L, which is expected to become part of the MPEG-4 standard.

Pulsent might have the edge today, but MPEG-4 is getting close to achieving the same compression ratio in combination with bitrate, said Danilo Tromp, a systems architect working on MPEG-4 products at Hilversum, Netherlands-based NOB Holding NV, the Dutch broadcast facilities.

“Pulsent’s claims sound good, if in fact they can achieve a broadcast quality stream at 1.1M bps. However, MPEG-4 comes close today and I think that with H.26L it will reach that in the near future,” he said, adding that broadcast quality in Europe is defined as a 720 x 576 pixel picture streaming at 25 frames per second.

Etienne Fert, video and communication research group leader at Philips Research in Paris, a part of electronics giant Koninklijke Philips Electronics NV, agreed.

“MPEG-4 with H.26L can meet such a bitrate. At 1M bit it will be at least VHS quality, moving up to DVD (digital versatile disc) quality at 1.5M bps,” he said. “I don’t think Pulsent’s announcement is a breakthrough compared to existing technologies.”

Key to Pulsent’s technology is an object-based approach to video compression, as opposed to the block-based approach taken by MPEG-2, according to the company. Pulsent’s way is to identify and model structural elements in a frame, instead of what it calls modeling of “arbitrary blocks.” Pulsent claims its objects can be far more accurately modeled than arbitrary blocks.

MPEG-4, the first version of which was released in 1999, also takes an object-based approach to video compression and is now making its way into software and hardware products. Backers of MPEG-4 include Philips, Microsoft Corp., Apple Computer Inc., Cisco Systems Inc. and Sony Corp.

A clear disadvantage over MPEG-4, according to Tromp, is that Pulsent’s technology is proprietary. Pulsent says it has over 200 patents filed and in process.

“MPEG-4 is an open standard, that is what we prefer,”he said. “However, we would like to see some material in broadcast quality and are interested in Pulsent’s technique.”

Tromp’s words were echoed by Paul o’Donovan, senior analyst at Dataquest Inc., a unit of Gartner Inc.

“In general, the big problem Pulsent has got is that their technology is not standardized. It can’t be adopted by anybody wanting assurance that this is going to be around for the next 10 years. MPEG-2 and MPEG-4 are standards and that is what the industry wants,” he said.

Besides just software, Pulsent also plans to sell an ASIC (application-specific integrated circuit) designed to be hardwired into devices such as set-top boxes, televisions and DSL modems. The chip will be able to do real-time video encoding and decoding, the company said.

“Nonsense,” commented o’Donovan. “This would mean that every device created so far would have to be thrown away and replaced by a device with this chip. That is just simply not going to happen.”

Pulsent seems to see itself taking a major role in a new market for video on demand over DSL. The company states on its Web site that DSL pipes today are too narrow for the delivery of quality video in the MPEG format and sees its technology as the solution. O’Donovan does not see a new market at all.

“The problem with TV over DSL is not the bandwidth, but it is the business model. There are plenty of technological solutions, but there are no business solutions out there. If it was so good, if people could make money out of it, it would long have been done,” he said.

Pulsent could just join the MPEG-4 Industry Forum, as its technology “seems to be awfully close to MPEG-4 in some aspects,” o’Donovan said. But, he continued, they probably won’t because “they want to make money.”