Music piracy effort faces key test

Two years have passed since the group of computer, consumer electronics and record companies agreed to work together toward broad anti-piracy technology standards. A first round of standards released a year ago has seen almost no adoption by participating companies, and a second, more ambitious song protection plan has been long delayed.

These setbacks have contributed substantially to the dearth of unambiguously legal music online. The big record labels have refrained from releasing much music on the Net until they feel confident they can protect their copyrights. As a result, the landscape continues to be littered with trial projects and start-ups failing for lack of access to the most popular music.

Few believe that Secure Digital Music Initiative (SDMI) will be a perfect salve for the online music industry's woes. Regardless, the energy that has been poured into the project for two years needs to resolve quickly into products; otherwise, critics say, the industry must move on.

Some SDMI members, who are paying fees of $20,000 to participate, are asking for signs that the end of the process is near. If those signs aren't forthcoming at this week's Los Angeles meeting, some say, the cracks developing in the foundation are likely to widen.

"We're certainly willing to go through tests, but there has to be an end goal," said Scott Moscowitz, chief executive of Blue Spike, whose music watermarking process is one of the finalists in the SDMI testing process. "But we can't go forever working in a vacuum."

A few companies, including MP3 chipmaker Micronas Semiconductors, have pulled out of the development effort. Analysts expect more to follow suit. Increasingly, people across the online music world are asking if SDMI is relevant.

"Record labels and device makers have made their own deals with other (content protection) providers," said Aram Sinnreich, a music analyst with Jupiter Research. "I think it's a serious question whether SDMI needs to exist anymore."

Mission creep?
The SDMI project kicked off in early 1999, and was created for the record, computer and consumer electronics companies to find a way to block unauthorized copying and widespread distribution of digital music files. The group was intended to set generic standards for this protection, so different technologies used by the various participants would work together as easily as CD players from different manufacturers work together today.

The group settled on the unconventional means of "watermarking," rather than the types of content lock-up mechanisms used elsewhere in the software industry. Under this model, an inaudible bit of data would be included inside songs, which would tell a music software device whether it could be copied or played.

But the process has been contentious since its beginning because the music, computer and consumer electronics companies don't always have the same interests. Testing the technology has taken longer than anyone expected, and the online music business has evolved to look different than it did in early 1999.

Some participants bemoan a process that has developed something like "mission creep," a term used to criticize military missions that expand beyond their original bounds.

"Two years ago when SDMI formed, we thought we had some finite things to deal with," said Scott Burnett, business development executive for IBM Global Media and Entertainment. "But so many new things have come along."

Many of those changes have added work to the technical side of the process, including the addition of devices such as cell phones and handheld computers. But other developments have been more demoralizing to participants, some analysts say.

Record labels have begun to strike content-protection deals with companies outside the SDMI framework. Napster also emerged as an unforeseen threat to the record companies' dominance. But by the end of last year, one of the big record companies had come to terms with Napster and was promising to keep it open in some form.

SDMI has accomplished at least one of its goals, even if modest. Early in its life, the group settled on a basic "Phase 1" watermark, which serves largely as a placeholder to tell devices when they should upgrade to new software.

Support for that standard was added to just one device--Sony's Music Clip player--only to be removed after producing negative customer reviews. Almost no new music is created and distributed with the watermark.

The SDMI has been working on a more ambitious standard, originally due in spring 2000. Technologies adopting these specifications would prevent a song from being passed from hand to hand as now happens through services such as Napster or Gnutella.

The group is reviewing the test results of several watermarking technologies that could be used for this purpose, deciding which of the remaining finalists are hardest for hackers to break through, have the least impact on a song's sound, and take the fewest computer resources to support.

Perhaps most controversial are the results of the public challenge to hackers issued late last year. According to SDMI, two technologies emerged unscathed. But outside groups, including a group of researchers and students at Princeton University, Rice University and Xerox PARC as well as, most recently, a pair of French programmers, say they were able to break through the technologies.

But the long delays and advances in network technology have called the latest efforts into question. These protection plans are based on determining when a file has been compressed--as happens when a song from a CD is put into the MP3 format.

Insiders from the technology and record companies admit that this will be far less useful once high-speed Net connections become mainstream, as is beginning to happen. Fast connections mean that music files could be sent immediately online, instead of squeezed into formats such as MP3. Under that scenario, the newest anti-piracy plans under development would need another lengthy round of development, these critics say.

The record companies say they're still committed to the process, although insiders, rather than acknowledging any specific technical advances, tend to cite the advantage of creating a forum for all the involved parties to talk.

"The idea was to get all of the same players that have an interest in digital music distribution at the same table," said Frank Creighton, director of anti-piracy efforts for the Recording Industry Association of America. "The fact that this is taking them so much time is not alarming. In fact, I think it's prudent. This is a huge undertaking."

But others say time is important, as devices increasingly start rolling off assembly lines offering MP3 support. Those devices, along with ordinary software programs, can play music created from every CD that's ever been pressed without any copyright protections.

This fact alone is enough to undermine any final results SDMI reaches, many analysts say.

"There is an existing base of millions of (CDs) out there, and they're always going to have to support those," Jupiter's Sinnreich said. "I think SDMI is largely moot."