Until finally I found myself in the strangest place of all, a small town in western North Carolina that seemed as far from the global confluence of technology and music as could be. This was Shelby, a landscape of clapboard Baptist churches and faceless corporate franchises, where one man, acting in almost total isolation, had over a period of eight years cemented his reputation as the most fearsome digital pirate of all. Many of the files I had pirated—perhaps even a majority of them—had originated with him. He was the Patient Zero of Internet music piracy, but almost no one knew his name.
Over the course of more than three years I endeavored to gain his trust. Sitting in the living room of his sister’s ranch house, we often talked for hours. The things he told me were astonishing—at times they seemed almost beyond belief. But the details all checked out, and once, at the end of an interview, I was moved to ask:
“Dell, why haven’t you told anybody any of this before?”
“Man, no one ever asked.”
CHAPTER 1
The death of the mp3 was announced in a conference room in Erlangen, Germany, in the spring of 1995. For the final time, a group of supposedly impartial experts snubbed the technology, favoring its eternal rival, the mp2. This was the end, and the mp3’s inventors knew it. They were running out of state funding, their corporate sponsors were abandoning them, and, after a four-year sales push, the technology had yet to secure a single long-term customer.
Attention in the conference room turned to Karlheinz Brandenburg, the driving intellectual force behind the technology and the leader of the mp3 team. Brandenburg’s work as a graduate student had pointed the way to the technology, and for the last eight years he had worked to commercialize his ideas. He was ambitious and intelligent, with a contagious vision for the future of music. Fifteen engineers worked under him, and he oversaw a million-dollar research budget. But with the latest announcement, it looked as if he had led his team into a graveyard.
Brandenburg did not possess a commanding physical presence. He was very tall, but he hunched, and his body language was erratic. He constantly rocked on his heels, lurching his gangly body forward and back, and when he talked, he nodded his head in gentle circles. His hair was dark and kept too long, and his nervous, perpetual smile exposed teeth that were uneven and small. His wire-frame glasses sat over dark, narrow eyes, and stray hairs protruded like whiskers from his scraggly beard.
He spoke quietly, in long, grammatically perfect sentences, punctuated with short, sharp intakes of breath. He was polite, and overwhelmingly kind, and he always tried his best to put people at ease, but this only made things more awkward. When he talked, he tended to dwell on practical matters, and, perhaps sensing boredom on the part of the listener, he would occasionally pepper this rambling technical discourse with weakly delivered, unfunny jokes. In his personality were united two powerful antiseptic forces: the skepticism of the engineer, and the stuffy, nation-specific conservatism they called typisch Deutsch.
He was brilliant, though. His mathematical talent was surpassing, and he held his contemporaries in thrall. These were men who had excelled in difficult academic disciplines and who had spent their lives near the top of competitive fields. They were not, as a rule, given to intellectual modesty, but when they talked of Brandenburg, their arrogance subsided and they reverted to quiet, confessional tones. “He’s very good at math,” said one. “He really is quite smart,” said another. “He solved a problem I could not,” said a third, and this, for an engineer, was the most terrible admission of all.
When challenged on a point, Brandenburg would pause, then squint, then subject the contrasting claim to a piercing scientific dismissal. In disagreement, his voice grew almost imperceptible, and in his responses he was guarded in the extreme, careful to never make an assertion without the data to back it up. In the conference room then, as he lodged his final objection to the committee, the mp3 went out with a whisper.
Defeat was always bitter, but this one was more so since, after 13 years of work, Brandenburg had solved one of the great open questions in the field of digital audio. The body of research the committee was dismissing went back decades, and engineers had been theorizing about something like the mp3 since the late 1970s. Now from this murky scientific backwater something beautiful had emerged, the refined product of a line of inquiry that went back three generations. Only the suits in the room didn’t care.
Brandenburg’s thesis adviser, a bald, stentorian computer engineer by the name of Dieter Seitzer, had started him down this path. Seitzer himself was indebted to his own thesis adviser, an obsessive investigator named Eberhard Zwicker, the father of an obscure discipline called “psychoacoustics”—the scientific study of the way humans perceive sound. Seitzer had been Zwicker’s protégé, his experimental audio subject, and, most important, his mortal opponent. For nearly a decade, the two had met every weekday after lunch for a game of table tennis, during which, over the course of an hour, Zwicker would school his pupil on the liminal contours of human perception while blasting ping-pong balls at his head. Zwicker’s chief finding, accrued over decades of research with real-world test subjects, was that the human ear did not act like a microphone. Instead it was an adaptive organ, one that natural selection had determined should 1) hear and interpret language and 2) provide an early warning system against enormous carnivorous cats.
The ear was only as good as it needed to be to achieve these goals, and no better. Thus, it had inherited a legacy of anatomical imperfections, and Zwicker’s research had revealed the unsuspected breadth of these errors. For example, anyone could distinguish two simultaneous tones separated by a half note or more, but Zwicker had found that, by moving the tones closer together in pitch, he could trick people into hearing just one. This effect was especially true when the lower-pitched tone was louder than the higher one. Similarly, any listener could distinguish between two clicks spaced a half second apart, but Zwicker had found that, by shortening this interval to just a few milliseconds, he could trick the ear into combining them. Here, too, increasing the relative loudness of one of the clicks made the effect more pronounced. The aggregate effect of these “psychoacoustic masking” illusions meant that reality, as humans heard it, was something of a fiction.
With time, Seitzer began to outplay the master. Zwicker was an anatomist, and his insights were products of the analog era. Seitzer, by contrast, was a computer scientist, and he anticipated the coming era of digitization. In particular, he suspected that, by exploiting Zwicker’s research into the ear’s inherent flaws, it might be possible to record high-fidelity music with very small amounts of data. This unique education gave him an unusual perspective. When the compact disc debuted in 1982, the engineering community celebrated it as one of the most important achievements in the history of the field. Seitzer, practically alone, saw it as a ridiculous exercise in overkill. Where the sales literature promised “Perfect Sound Forever,” Seitzer saw a maximalist repository of irrelevant information, most of which was ignored by the human ear. He knew that most of the data from a compact disc could be discarded—the human auditory system was already doing it.
That same year, Seitzer applied for a patent for a digital jukebox. Under this more elegant model of distribution, consumers could dial into a centralized computer server, then use the keypad to request music over the new digital telephone lines that Germany was just beginning to install. Rather than pressing millions of discs into jewel cases and distributing them through stores, everything would be saved in a single electronic database and accessed as needed. A subscription-based service of this kind could skip the manifold inefficiencies of physical distribution by hooking the stereo directly to the phone.