The head-to-head success of the device in the open marketplace revived interest in a format that the world had left for dead. The standards committees had hated the mp3, but the customers sure loved it. This success brought attention, and soon Fraunhofer was cutting other deals. Macromedia licensed the mp3 for use with its multimedia Flash codec; Microsoft licensed it for an early version of Windows Media Player; a start-up satellite radio provider named WorldSpace licensed it for broadcasting to the Southern Hemisphere. The overall revenue from these deals was modest—enough to justify the technology’s continued existence, but not enough to justify the thousands of man-hours and millions of dollars Fraunhofer had spent in development.

And so, toward the end of 1996, Fraunhofer was preparing to retire the mp3. Its development was complete, and there was no longer anyone actively working on it. The plan was to shift the technology’s limited customer base to the second-generation Advanced Audio Coding, which was now nearing completion. AAC had delivered on its promise. It was 30 percent faster than the mp3 and employed a variety of new techniques that allowed it to compress files with perfect transparency even beyond the 12-to-1 goal. After 14 years, Seitzer’s vision was real, and when Fraunhofer submitted the AAC technology for standardization in late 1996, the event formally marked the mp3’s obsolescence.

What happened next was like an episode of Star Trek. A mysterious case of amnesia struck every member of the crew. Brandenburg, Grill, and Popp; Gerhäuser, Eberlein and Herre; even Seitzer was afflicted. The Fraunhofer team generally had excellent memories, and could often recall events more than twenty years past with great clarity and precision. They were good record keepers too, and the stories they told of the early days could almost always be corroborated with photographs and documentation. But when it came to the mysterious period of late 1996 to early 1997, every one of them drew a blank. No one—not one—could remember the first time they’d heard the word “piracy.”

The Fraunhofer team were no strangers to the Internet, but the Internet they knew was a collaborative tool for research and commerce, not some grimy subculture of anonymous teenage hackers. In their naiveté, they had not seen what was coming. Somewhere in the underworld, L3Enc, the DOS-based shareware encoder Grill had programmed several years back, was being used to create thousands upon thousands of pirated files. Meanwhile, somewhere else in the underworld, the commercial WinPlay3 player that supposedly self-destructed after twenty uses had been cracked, enabling full functionality. Together, the two were now being distributed in chat rooms and websites as a bundled package.

That wasn’t all. Some of the Warez Scene groups were also providing direct links to Fraunhofer’s FTP server, along with stolen serial numbers for L3Enc and WinPlay3. By the middle of 1996, Fraunhofer’s database administrators would have seen a spike in the FTP traffic for mp3 software. By late 1996 the surge in downloads of L3Enc and WinPlay3 would have been impossible to ignore. After years of neglect, there was finally interest in mp3 software—but, amazingly, none of the Fraunhofer researchers could recall the details of this remarkable turnaround.

The official Fraunhofer narrative only resumed on May 27, 1997, when Brandenburg, in America for a conference, was handed a copy of USA Today. There, buried on page eight of the newspaper’s “Life” section, in an article by the music journalist Bruce Haring, was the first ever mention of the mp3 in the mainstream press. “Sound Advances Open Doors to Bootleggers,” read the headline. “Albums on Web Sites Proliferate.” Included in the article was a short interview with an 18-year-old Stanford University freshman named David Weekly.

In late March of this year, Weekly put 110 music files—including cuts from the Beastie Boys, R.E.M., Cypress Hill and Natalie Merchant—on his personal Web server, run through the university system. Soon, more than 2,000 people a day were visiting, representing more than 80% of Stanford’s outgoing network traffic.

Brandenburg recognized the importance of this development. He knew he needed to remember this moment, and he knew he needed to bring it to the attention of his colleagues back at Fraunhofer as well. So he cut the article out of the newspaper with a pair of scissors.

Brandenburg disapproved of piracy. Everyone at Fraunhofer did. These men were inventors who made a living by selling their intellectual property, and they deeply believed in both the letter and the spirit of copyright law. They were not participants in the file-sharing subculture, and they never pirated music files themselves. Upon Brandenburg’s return to Germany, they prepared a course of corrective action. They reported some of the more brazen hackers to the authorities, and they scheduled a meeting with the Recording Industry Association of America, the music industry’s lobbying and trade group, at their headquarters in Washington, D.C., to warn them of what was occurring.

Brandenburg arrived at the RIAA meeting that summer with an enhanced piece of tech: the copy-protectable mp3. Although his recent experience had showed how this protection could be disabled by technical experts, Brandenburg believed that the majority of casual downloaders would never make it past this hurdle. At the meeting, he demonstrated the use of the file, then urged the RIAA to adopt this technology at once. The best way to get ahead of mp3 piracy, he believed, was to provide a legal substitute.

He was informed, diplomatically, that the music industry did not believe in electronic music distribution. To him this was an absurd argument. The music industry was already engaged in electronic distribution. To the recording executives those racks of CDs at the mall might look like inventory, but to an engineer they were just an array of inefficiently stored data. Brandenburg explained his position again, but his patient, methodical style of scientific argumentation failed to ring the appropriate alarm bells. So he got on a plane and went home.

Why didn’t they listen? The RIAA would later offer various explanations:

The first explanation was that Brandenburg’s argument was self-serving. To sell mp3s legally, the industry would have had to license them from Fraunhofer, and that would have been expensive. Given the number of pirated files being hosted online, Brandenburg’s proposal might even have looked like blackmail, although this was certainly not his intention.

The second explanation was that the RIAA was not actually in charge of the music industry. The opposite was true: it was just a lobbying arm that took its orders from the Big Six. RIAA employees were Beltway insiders who talked to the legislators about copyright policy, or private detectives who worked with law enforcement to hunt down bootleggers, or accountants who certified gold and platinum records. They weren’t capital allocators, and they didn’t have the authority to make large-scale investments in digital distribution technology. Brandenburg had scheduled a meeting with the wrong people.

Still, if they’d really cared, the RIAA could at least have referred Brandenburg to a major label. But they didn’t do that either. And that was for a third reason, the best explanation of all: their technical people told them not to. The studio engineers hated the mp3. These were the knob-twiddling soundboard jockeys who actually mixed the albums. Responsibility for the sound quality of recorded albums fell to them, and, in their consensus opinion, the mp3 sounded like shit.

This guildlike resistance to the technology proved to be the biggest hurdle to early adoption. In one regard, the studio engineers had a point. The cracked version of L3Enc floating around on the Internet did not produce high-quality audio, and even a casual listener could easily distinguish between a compact disc and the early pirated files. But it went beyond that—the studio engineers were irredeemable audiophiles who regarded even high-quality mp3s with disdain. For them, capturing the subtle acoustic qualities of recorded music was a professional obligation that bordered on obsession. Now Brandenburg was proposing to irretrievably delete 90 percent of their life’s work.