The First Recording of a Human Voice (1860) — Édouard-Léon Scott’s Phonautograph Sings “Au Clair de la Lune”

What is this eerie recording?

This short, warbly clip might sound like someone gurgling underwater while trying to sing. It’s actually the first known audio record of a human voice, traced on paper on April 9, 1860.

Listen: “Au Clair de la Lune” (reconstructed)

How the phonautograph captured sound (but couldn’t play it)

Scott’s phonautograph converted air vibrations into a visual trace on paper using a membrane and stylus. The device was intended for analyzing speech and music—not for playback. Thomas Edison would later invent the phonograph in 1877, which finally enabled recorded sound to be heard.

Who made the first voice recording?

French inventor Édouard-Léon Scott de Martinville created the phonautograph and produced several paper recordings in the late 1850s–1860. The snippet here is believed to be a line from “Au Clair de la Lune.” Scholars have debated whether the voice is female or possibly Scott’s himself—the fragment is too degraded to be conclusive.

How did we finally hear it in 2008?

In the 2000s, audio historians and physicists used optical scanning and digital restoration to convert the phonautograms’ lines back into audio—without touching the fragile paper. In March 2008, a playable version was unveiled, marking the first time the 1860 voice trace could be heard as sound.

FAQs — First Human Voice Recording

What is the earliest known recording of the human voice?

A phonautograph trace made on April 9, 1860, in Paris—later reconstructed digitally to sound like a fragment of “Au Clair de la Lune.”

What is a phonautograph?

An 1850s device that recorded sound visually by tracing vibrations onto soot-blackened paper. It did not play recordings back.

Who invented it?

Édouard-Léon Scott de Martinville, a French printer and inventor focused on the science of acoustics and speech.

How did researchers play the 1860 trace in 2008?

By scanning the phonautogram at high resolution and using software to convert the drawn waveform back into audio—no needles, no physical contact.