MP3@25: 1 – Tom’s Diner

MP3@25 | 1: Tom’s Diner | 2: Napster | 3: iMP3 | 4: At the Movies | 5: Legacy |6: Timeline | 7: Gift Shop


The 1982 a capella recording of Suzanne Vega’s original song Tom’s Diner – like Vega’s own voice – has become synonymous with the history of the MP3’s creation. As with all legends, the story is complicated.

Suzanne Vega with a certificate presented to her at Fraunhofer IIS (Karlheinz Brandenburg stands behind her), 3 August 2007. Photo: Daniel Karmann.

Legend has it that Karlheinz Brandenburg, one of the engineers at the Fraunhofer Institute who oversaw testing and refinement of the MP3 compression algorithm, heard this song and believed it would make the perfect test subject via which MP3 compression could be fine-tuned to perfectly reproduce the sound of the human voice. Brandenburg has himself offered multiple (conflicting) versions of the story, and the absolute truth may never be known.

The First Recording of a Human Voice vs ‘The First MP3’: At the top, the image of the sound of Édward Léon Scott de Martinville’s voice singing the song ‘Au Clair de la Lune’ as etched by a needle into charcoal-darkened paper by his own invention, the Phonautograph – the first listenable recording of a human voice (1860). Below, a portion of the code inside ‘the first MP3 file’, Suzanne Vega’s ‘Tom’s Diner’ (a capella), viewed as a text file (2020).

The following portion of MP3@25 presents four versions of Tom’s Diner which together deconstruct and rejoin Vega’s work(s) with the ongoing legacy of the MP3.


Suzanne Vega
Tom’s Diner (a capella) 1987

Recorded 1986-1987. From the album Solitude Standing, released on A&M/PolyGram Records, 1 April 1987. Written by Suzanne Vega, 18 Nov 1981. Produced by Steve Addabbo and Lenny Kaye.

Part truth, part internet legend, the role played by Suzanne Vega (aka ‘The Mother of the MP3‘) and this song in the birth of the MP3 is primarily apocryphal; yet, one of the engineers credited with fine-tuning the MP3 encoding algorithm, Karlheinz Brandenburg, did indeed use Tom’s Diner (a capella) as one of several MP3 test tracks, and it was indeed selected because Brandenburg was impressed by the quality of its recording of Vega’s (almost) unaccompanied voice.

Suzanne Vega
Tom’s Diner (a capella on Edison wax cylinder) 2012

Recorded at Building 11, Thomas Edison National Historical Park, Glenmont, West Orange, Essex County, New Jersey, United States in February 2012.

In 2012, Vega was invited to the Thomas Edison National Historical Park in New Jersey, which preserves the home and laboratory of Thomas Edison, originally designed in 1887. While at the park, Vega was recorded onto a specially produced 2-minute Edison wax cylinder singing Tom’s Diner unaccompanied in attempt to reclaim Vega’s song and performance for the analog audio world.

Ryan Maguire
moDernisT 2014

Conceived, produced, and recorded by Ryan Maguire

Ryan Maguire‘s conceptual composition amplifies the frequencies in Vega’s 1987 recording that are stripped away when the original track is encoded into the MP3 format. These ghostly echoes amplify what has been lost, giving voice to parts of the original recording which now frequently go unheard in our contemporary world of compressed digital audio as the current standard for recorded sound. A detailed explanation of the composition process is published on Maguire’s website.

Tom’s Diner Edison Wax Cylinder MP3 file 2020

Created by Museum of Portable Sound Director & Chief Curator John Kannenberg on the 25th anniversary of the MP3, 14 July 2020. Released exclusively via this exhibition website.

Exclusive to this exhibition, we have created an MP3 file of the audio of Vega’s Edison wax cylinder performance of 2013, which you can download by clicking on the Napster screenshot above. It is encoded at a consistent bitrate of 128kbps. It is yours to download, keep, and share.


Exhibition Home | Onto Part 2 > | Gift Shop