Dead on Arrival?
By Carlo Longino, Thu Jul 17 10:30:00 GMT 2003
MPEG-4 was envisioned as the be-all and end-all for streaming media, especially for mobile devices. But a licensing tiff threatens its adoption.
Mobile streaming video has long been held as the touchstone of advanced mobile data. Video calls, streaming movies to handsets, TV news on PDAs - it's been a holy grail of sorts for carriers and content providers, the final piece to justify the hype of high-speed wireless networks. But many current efforts resemble the early days of streaming media on the Web - tiny, jerky images that could barely be considered video. Compression codecs simply haven't been sufficient to deal with the low throughput, high latency, and other drawbacks of mobile networks.
But MPEG-4 was supposed to change all that. It was meant first to unite streaming media on one standard, regardless of the platform (such as mobile, DTV, or Internet), while featuring different profiles that catered to each. It's been ratified by a number of standards bodies, including the 3GPP which designates MPEG-4 as its video codec, along with H.263, an older videoconferencing codec.
MPEG-4 support has already been included in such products like Apple's QuickTime on the Web and RealPlayer and PacketVideo's player in the mobile space. Japanese carriers NTT DoCoMo and KDDI use the standard for their i-motion and ezmovie mobile video services. But these carriers, and other Japanese mobile content providers, are balking at the standard's licensing fees.
The MPEG-4 patents are held by about 20 companies and feature a host of names familiar in the mobile industry - France Telecom, Fujitsu, Matsushita, Microsoft, Mitsubishi, Philips, Samsung, Sanyo, Sharp, Sony, Telenor, Toshiba, and JVC - and are administered by the Denver-based MPEG LA, the same firm that handles the earlier MPEG-2 standard. MPEG LA first announced its licensing terms last year, setting rates of 25 cents per encoder and decoder, though Apple later influenced those fees to be limited to USD 1 million per licensee per year. But what has the Japanese companies - as well as others - concerned is that MPEG LA will charge content providers 2 cents per hour of streamed content that's encoded in MPEG-4.
So for every hour of video that each subscriber watches, content providers must pony up two cents. It doesn't sound like much, but with Japanese broadcasters hoping to shift to digital transmission using the standard in 2005, with those digital signals available on mobile devices, it could turn into a hefty bill.
What's clear here is that MPEG LA is looking for a bigger piece of the pie than they received with MPEG-2, when they just charged per encoder and decoder. So instead of just making money each time an encoder or player is bought or downloaded, they'll earn each time it is used (though the usage fees only apply to revenue-generating services). But it's also possible this licensing strategy could drive providers away from MPEG-4, not only reducing the amount of royalties collected, but also undermining the strategy of having a unified standard.
One such possibility for mobile content and service providers is H.263 - which unlike MPEG-4, is free. Though H.263 may not provide as high-quality video as MPEG-4 in some cases, companies will likely question if the marginal improvement in performance is worth the additional cost. H.263 is already supported in a number of mobile players and is used in favor of MPEG-4 by companies such as Nokia.
But two other possibilities may emerge from entrenched Web streaming-media strongholds, Microsoft and RealNetworks. Most would agree that the companies' latest codecs outperform MPEG-4, and both companies are prepared to take on MPEG LA with their licensing and fee structures. Microsoft, for instance, instigated something of a streaming-standard price war earlier in the year when it announced it would license its Windows Media Video 9 codec for products other than Windows PCs, charging license fees of 10 cents per decoder and 20 cents per encoder, or 25 cents for both, with no usage fees. This came after RealNetworks opened some of its source code to allow developers to program their own applications to play Real-format files.
But another alternative emerging is the H.264 standard, also known as MPEG-4 Part 10 AVC (advanced video coding), developed jointly by a group from the ITU and the Motion Picture Experts Group. H.264 was intended to be the successor to MPEG-4, but may render it dead on arrival, thanks to a much improved performance. But there's also a growing push to make a "baseline" H.264 profile royalty-free to spur adoption, which could turn many content providers in its favor.
The current royalty schedule seems to doom MPEG-4. Imagine a company like AOL with 30 million subscribers that broadcasts streaming media to them - 2 cents per hour per user could quickly add up. And with rival codecs available at lower encoder/decoder rates and without a user fee - not to mention equal or better performance - why should they, or any other provider, choose MPEG-4?
Carlo Longino is a freelance writer based in Austin, Texas. His previous experience includes work for The Wall Street Journal, Dow Jones Newswires, and Hoover's Online.