Even before we could send recorded sound reliably over the Internet and a file-sharing revolution set the music industry in an uproar, the large size of digital audio files was causing headaches. A minute of raw stereo sound at CD quality — or 16-bit, 44.1 KHz sampling — occupies 10.5 MB on a hard drive. Obviously, files of song-length recordings at that size can quickly get unwieldy.
Compression proved to be the answer. In 1987, technicians at the Fraunhofer Institut Integrierte Schaltungen in West Germany began working on a sophisticated compression algorithm, one that could achieve compression ratios of 10:1 and better while preserving high-quality sound.
The original test bed for this technology was a box of electronics the size of a refrigerator, but by 1992 the Fraunhofer team, working with Deutsche Thomson Brandt, had reduced it to software and it became part of the MPEG standard. Officially known as MPEG-1 Layer III — also endorsed by the ISO — the new kid on the block soon became known as MP3, and the rest, as they say, is history.
Players, both software and hardware, were created to play MP3s. Combined with Napster’s peer-to-peer file-sharing network, the compact new codec — or compression-decompression algorithm — helped overturn the traditional relationship between music publishers and consumers, a confrontation now being played out in the courts.
Fraunhofer and Thomson still own the copyrights to MP3 and charge developers to use it in their authoring packages and players. A refinement, MP3Pro, has recently hit the streets, and the new MPEG-4 standard incorporates Advanced Audio Coding (AAC) developed by Dolby and used by Apple for its iTunes pay-per-track online music store.
Meanwhile, other developers have not been idle. Microsoft incorporated a proprietary audio codec into its Windows Media Audio (WMA) format, and RealNetworks used its own codec in its RealPlayer systems. Like MP3, these formats must be paid for by developers wishing to use them. Both the authoring applications used to create music in these formats and the players themselves are usually charged fees.
Those fees typically vary with the number of units to be produced and, in some cases, the nature of the use. The numbers are serious: 100 million MP3 players, for example, are expected to be in use worldwide by the end of this year, and a billion music tracks are said to be downloaded every month.
Meanwhile, Out in the Open
Open-source developers have been busy, too. As Christopher Montgomery, founder and lead programmer at Xiph.org, an open-source development project, puts it, “Open source is not a fad any more than the Internet is; it is a necessary force driving innovation and the Internet forward while protecting the interests of individuals, artists, developers and consumers.”
Several open-source audio encoding-decoding formats have been created, and they are supported by many authoring and player packages. There’s LAME. There’s Monkey’s Audio and a voice-oriented codec called Speex.
There’s also Exact Audio Copy, software written by a student at the University of Dortmund in Germany. Finally, and perhaps most significantly, there’s the oddly named Ogg Vorbis, which could be the strongest challenger yet to the proprietary music formats.
“We are seeing Ogg Vorbis adopted by MP3 player manufacturers and chip manufacturers who, as the market grows and as the lower end becomes increasingly commoditized, are looking for ways to boost their feature sets,” IDC senior analyst Susan Kevorkian told TechNewsWorld.
“One way for them to do that is to support a greater number of codecs,” said Kevorkian. “The crux of the matter is how many people are using it and how much music is available encoded in that format.”
Who’s the Fairest of Them All?
You might think the choice of compression format would be easy: Run some tests and see which sounds best. But it’s not so simple. First, modern compressors are based on psychoacoustic effects — that is, tests that rely on human hearing. Comparison tests — and there have been many — are therefore largely subjective. Almost everyone comes up with a different set of bests and worsts.
In addition to psychoacoustic issues, there are Internet-specific issues: To be practical, a codec must support streaming. There’s also the choice between lossy and lossless algorithms. The former eliminates supposedly nonessential data to achieve a greater compression ratio, but at a cost in quality and in the time required to encode and (more critically) decode.
Data transfer rate plays a role, too: The holy grail of audio compression is the best quality combined with the lowest bit-rate. Codecs can succeed or fail on that factor alone. Backward compatibility is important, too, as well as support from third-party manufacturers. Even content matters: Classical music demands more from a codec than rock. Finally, there’s the question of digital rights management (DRM) to prevent unauthorized copying of audio content.
DRM the Issue
DRM might be the critical issue. “The question is, are the music labels actively supporting [Ogg]?” Kevorkian said. “Because if digital rights management isn’t a part of it, then it’s certainly not the direction that the music industry’s going. In fact, [open-source developers] are proponents of unprotected codecs.”
However, said Kevorkian, the major record labels are experimenting more actively with CD content protection technologies that include both a protected set of Red Book audio files — the standard audio CD file format — plus Windows Media Audio files, which are protected so that you’d be able to transfer them to your PC’s hard drive but not distribute them online.
“Also,” she said, “as the music industry supports paid music services more actively, we expect the majority of paid music services to support Windows Media Audio because the codec and the DRM are closely integrated, so every music service provider doesn’t have to think about combining the two like Apple did.”
Whether or not open-source audio standards will threaten the dominance of MP3 and WMA remains to be seen. Open-source audio developers — opposed by aggressive commercial giants, ideologically committed to nonprofit but needing development capital — seem to face unbeatable odds. Perhaps they’ll take heart by looking at open source’s recent progress in the corporate world.