How "Tom's Diner" Tuned the MP3

Image credit: 

Suzanne Vega's catchy tune has made her "Mother of the MP3" -- though it took a while. Vega wrote "Tom's Diner" as an a cappella song way back in 1982. By 1984 it has been released on an obscure folk compilation, and didn't appear on Vega's studio albums until 1987's Solitude Standing. In 1990, the song was remixed by The DNA Disciples, adding a danceable beat and instrumentation -- this version hit the Billboard Hot 100, peaking at #5 in the US.

So what does this have to do with MP3? Well, after its release in 1987 audiophiles began using Vega's a cappella track to test speaker systems for clarity. It was considered a good, warm recording of a human voice -- something that could reveal flaws in an audio setup. Working at the Fraunhofer Society in Germany in the 90's, audio engineer Karl-Heinz Brandenburg was hard at work developing the MP3 audio compression scheme. Brandenburg used Vega's a cappella version of "Tom's Diner" to tune the compression system, playing the track before and after compression was applied to tell whether MP3 sounded good enough. He figured Vega's song would be a tough track to compress (as it was already favored by audiophiles), and would be a good test for whether MP3 was really listenable. Although many audiophiles ended up hating MP3, Brandenburg seems to have done pretty well for himself -- MP3 became an incredibly popular technology. On the choice of "Tom's Diner," Brandenburg recalled: "I was ready to fine-tune my compression algorithm...somewhere down the corridor, a radio was playing 'Tom's Diner.' I was electrified. I knew it would be nearly impossible to compress this warm a cappella voice."

In last week's New York Times, Vega reminisced about the song and her career as a "two-hit wonder" (the other hit was "Luka"). From her article:

So Mr. Brandenberg gets a copy of the song, and puts it through the newly created MP3. But instead of the "warm human voice" there are monstrous distortions, as though the Exorcist has somehow gotten into the system, shadowing every phrase. They spend months refining it, running "Tom's Diner through the system over and over again with modifications, until it comes through clearly. "He wound up listening to the song thousands of times," the article, written by Hilmar Schmundt, continued, "and the result was a code that was heard around the world. When an MP3 player compresses music by anyone from Courtney Love to Kenny G, it is replicating the way that Brandenburg heard Suzanne Vega."

So goes the legend. The reason I know what that MP3 originally sounded like is that last year I was invited to the Fraunhofer Institute in Erlangen, Germany, where I met the team of engineers who worked on the project — including Mr. Brandenberg, who I had met once before at the launch of the Mobile Music Forum in Cannes in 2001.

All the men are obviously intelligent, but Karl-Heinz is a character. He stands out, because he looks like a mad scientist. His hair and tie always look as if they have been blown askew in a stiff wind, and he taps the tips of his fingers together constantly, smiling beatifically.

Read the rest for one artist's thoughts on her nearly thirty-year career in music -- and the unexpected resonance of a song scribbled on paper back in 1982.

More from mental_floss...

September 30, 2008 - 10:27am
submit to reddit