I must confess that I don't watch Survivor or other "reality" TV shows. Nothing religious, it's just that they hold about as much interest for me as the Home Shopping Network or cricket reruns on the BBC.
Of course, it helps that whenever I need a dose of drama I can tune into the streaming media marketplace. Over the past few years, we've experienced Congressional hearings, antitrust claims, and the latest Star Wars sequel—who needs reality TV?
This brings us to the latest drama, regarding license fees for MPEG-4's "(Visual) Patent Portfolio." Briefly, MPEG-4 is a streaming codec and transport standard from the folks who brought us MPEG-1 and MPEG-2. Through a string of technologies, it's designed to allow companies like CNN to post one audio/visual presentation and have it magically play optimally on everything from your HDTV set to your wireless phone. Of course, the sweet spot for MPEG-4 is the streaming marketplace currently dominated by RealNetworks, Microsoft, and to a lesser degree, Apple.
For the record, I've believed in standards since my "I fought the law and the law won" days at Iterated Systems, where our fractal-based technology lost out to JPEG and MPEG-1. Standards mean universal support across many markets and devices, and prevent one company from acting arbitrarily regarding their technology. I also like the theory of MPEG-4, the ability for publishers to post one file on their Web sites, as opposed to the politically correct three formats that many now post, and have any media player play back the file.
It's just that MPEG-4 comes with more rough edges than a packet full of sandpaper, the latest being the "usage" royalties proposed by MPEG-LA, the licensing body for the MPEG-4 "visual" patent holders (audio patents are licensed by Dolby).
MPEG-LA proposed a four-part royalty program. There's a royalty of $0.25 per encoder and decoder, compared to $2.50 for each MPEG-2 encoder or decoder. More importantly, there's a cap of $1,000,000 for both encoders and decoders, so Microsoft, Real, or Apple can pay $2,000,000 and ship as many encoders or decoders as they like.
There's also a $0.02/hour fee for packaged goods shipped with MPEG-4, compared to $.035 for MPEG-2-based DVDs. No problem here as well.
The sticking point is a running-time fee of $0.02/hour "for every stream, download or other use of MPEG-4 video data in connection with which a service provider or content owner receives remuneration." If folks pay to see your video, MPEG-LA wants to hear clinking in their piggy bank.
Though MPEG-LA declines to call this a usage tax, this is what it's been coined. And a host of companies, including Apple, think that a usage fee just won't fly. In fact, Apple held back on shipping the MPEG-4-compliant QuickTime 6 until more acceptable terms could be reached.
Apple's case was significantly bolstered when Dolby, licensor for MPEG-4 audio, announced a program without usage fees, for streaming or packaged media. Instead, Dolby chose to put a cap on PC-based software codecs, but not on hardware-based decoders like those used in cell phones.
Granted, some definition issues will apply, like whether an AAC decoder for your Palm is a PC-based software decoder or not. But these pale in comparison to the problems that MPEG-LA's proposed usage fee brings to bear.
For example, as Rob Koenen, president of the MPEG-4 Industry Forum (M4IF) points out, when does a Web site receive "remuneration" for the viewing? Without doubt, if the compensation is direct, the issue is clear.
But what happens when you click over to ESPN.com to enjoy reruns of Duke being bounced from the NCAA tournament? There is no direct remuneration, but this would create a royalty obligation. How would you compute the royalty for my RealOne subscriptions that I could watch 24 hours a day, but don't? Deciding when the usage royalty applies and how to compute it will quickly get convoluted, and may need to change for every business model.
The usage fee also has severe enforcement problems. According to their Web site, MPEG-LA has 397 MPEG-2 licensees. But the number of sites currently streaming video surely climbs to the hundreds of thousands, and, unlike with DVD and MPEG-2, you can't go to Blockbuster or Amazon. com to see who they are. You have to dig them out.
Finally, the thorniest problem is that MPEG-4 video quality, in all published reviews, still trails both Real and Microsoft, which makes it hard to charge either the end-user or publisher. And despite the lovely vision of standards, both Microsoft and Real can beam video to cell phones today, delivering the MPEG-4 vision without the usage fee.
Speaking to the MPEG-LA folks, I was struck by their thoughtful awareness of all these issues. They were far from the monolithic, "our way or the highway" group that I expected, and are honestly striving to create a commercially reasonable model. Clearly, the business issues can be resolved, either by adapting the two-tier Dolby model or some more enforceable model that ties the royalty obligation more directly to revenue.
From a quality perspective, however, MPEG-LA is tied to the technology delivered by its patent holders. Until MPEG-4 quality is at least equal to other commercially available alternatives, the best business model in the world may not be good enough.