The Digital Music Revolution

The Compressed Sound

The MP3 standard began not as a product but as a dissertation. Karlheinz Brandenburg started research at the University of Erlangen-Nuremberg in the early 1980s on how the human ear processes music. The core question was psychoacoustic: which information in an audio file does a human actually hear, and which can be dropped without anyone noticing?

Human hearing follows two masking phenomena:

• Frequency masking: a loud tone at a certain frequency covers quiet tones at neighboring frequencies. These can be deleted from the file without perceptible quality loss.
• Temporal masking: shortly before and after a loud sound event, weaker tones cannot be perceived; this information is redundant too.

Brandenburg used these findings to develop an algorithm that encodes only what the ear actually registers. The result was a compression ratio of about 11:1 with barely perceptible quality loss: a CD-quality audio file of 50 megabytes became an MP3 of 4–5 megabytes.

The standard was published in 1993 as part of the MPEG-1 standard (ISO/IEC 11172-3), developed at the Fraunhofer Institute for Integrated Circuits (IIS) in Erlangen, where Brandenburg worked from 1993. Alongside Brandenburg, the team included the American researcher James D. Johnston (AT&T Bell Labs), who made key contributions to the psychoacoustic modeling.

Fraunhofer patented MP3 and began collecting license fees in 1995. The codec spread rapidly anyway, not because of the industry, but despite it.

Napster and the Democratization of Theft

In the summer of 1999, Shawn Fanning, 18 years old, sat in a dorm room at Northeastern University in Boston and wrote software that exchanged music files directly between private computers, hard drive to hard drive. His co-founder Sean Parker helped with promotion.

Napster went online in June 1999. The growth rate was unprecedented: in January 2001 the service counted over 26 million users; at its peak, millions of MP3 files were exchanged daily without a cent flowing to artists or labels. The mechanism was technically elegant: a central directory listed which user had which files; the actual file transfer ran directly between machines.

The music industry’s reaction was sharp. In early December 1999, the RIAA (Recording Industry Association of America) filed suit. On April 13, 2000, came the most spectacular individual lawsuit: the band Metallica had discovered that a demo of their unreleased song I Disappear was being played on radio stations via Napster. The band manually submitted 335,000 usernames that had allegedly traded copyrighted material and demanded their suspension. Napster suspended them.

In March 2001, Judge Marilyn Hall Patel issued a preliminary injunction: Napster had to implement a filter within 72 hours or shut down. In July 2001, Napster ended free sharing; bankruptcy followed in 2002. But the files (and users’ expectation of free music) remained.

Steve Jobs and the 99-Cent Deal

When Steve Jobs presented the iTunes Music Store on April 28, 2003, he had solved something the entire music industry had failed at for four years: he had brought all five major labels to one table and talked them into saying yes.

Jobs’ starting position was unusual: Apple was not a music company. But Apple had the iPod (2001), and with it a device for which legal music files would be a killer feature. Jobs courted the heads of Universal, Sony Music, Warner, BMG, and EMI personally and demonstrated an interface so simple it disarmed piracy: one click, one song, 99 cents.

The terms were unusual for the labels, but acceptable:

• 70% of the proceeds went to labels and publishers
• 30% stayed with Apple, a thin margin Jobs calculated as break-even
• DRM protection (Apple’s FairPlay) was built in, but more moderate than the labels wanted
• Individual songs could be bought; the album bundle was gone

The result astonished everyone involved: in the first week, 1 million songs were sold; Apple had internally planned six months for that milestone. In the first year, users bought over 70 million songs.

iTunes proved that people pay for digital music when buying is more convenient than stealing. The model also settled a previously open power question: Apple, not the labels, controlled the customer interface.

Daniel Ek and the Streaming Paradigm

In October 2008, Daniel Ek, together with Martin Lorentzon, launched Spotify, initially invitation-only in Europe. Ek’s basic observation was sobering: despite iTunes, despite the Napster lawsuit, despite DRM systems, piracy was still rampant. The reason was not moral indifference; it was a usability problem.

Ek stated the dilemma precisely: “You can never legislate away from piracy. The only way to solve the problem was to create a service that was better than piracy and at the same time compensates the music industry.”

Spotify solved the usability problem through three design decisions:

1. Instant access: no downloading, no file system, no waiting. Music started playing in seconds.
2. A universal catalog: one of the largest catalogs, with no per-purchase decision required.
3. A free entry point: an ad-funded free tier lowered the barrier; a premium subscription at €9.99 removed the ads.

The model scaled: Spotify had five million paying subscribers in 2012, over 220 million in 2023. Streaming’s share of global music revenue exceeded 67% for the first time in 2023.

Dead End: DRM

Digital Rights Management was the music industry’s answer to piracy: audio files were wrapped in encryption that restricted playback to authorized devices. The implementation was technically elaborate; the results were catastrophic.

Microsoft’s PlaysForSure (2004) was supposed to create interoperability: music from any compatible store should play on any compatible device. In practice the system was fragmented: PlaysForSure tracks did not play on the iPod, and when Microsoft introduced its own Zune player in 2006, it was not PlaysForSure-compatible either.

Sony’s XCP rootkit (2005) became a PR disaster: Sony BMG had shipped CDs with First4Internet’s Extended Copy Protection (XCP), which, when played in a PC, secretly installed software that modified the operating system to prevent copying, without informing the user. Security researcher Mark Russinovich discovered the rootkit on October 31, 2005; Sony recalled the XCP CDs (52 titles, over two million of them sold) and replaced them with versions without the software.

Apple’s FairPlay was the most successful DRM system, but even Jobs recognized the limits. In an open letter, “Thoughts on Music” (February 2007), he called on the labels to abandon DRM: the only people constrained by DRM were the paying customers, not the pirates, who traded DRM-free copies anyway. In 2009, iTunes switched to DRM-free AAC files.

DRM failed on a structural contradiction: it is a tool deployed against one’s own customer. Every technical barrier that a legitimate user experiences as a hurdle, a pirate experiences as an advantage, because the pirated copy does not have that hurdle. The only thing DRM effectively protected was the record companies’ feeling of control.

📚 Sources

• Wikipedia: MP3
• Wikipedia: Karlheinz Brandenburg
• Internet Hall of Fame: Karlheinz Brandenburg
• Wikipedia: Metallica v. Napster, Inc.
• Wikipedia: Spotify
• History.com: The death spiral of Napster begins
• Rolling Stone: Steve Jobs’ Music Vision Transformed an Industry
• Rolling Stone: iTunes’ 10th Anniversary
• Britannica Money: Daniel Ek
• Kerrang!: Metallica vs. Napster