Search Gear
 

Tech Page: Credit Where Credit Is Due

June 1, 2008
share
FIG. 1: Édouard-Léon Scott de Martinville''s phonautograph consisted of a barrel-shaped horn and stylus that etched images of sound waves onto a rotating piece of paper coated with soot.

FIG. 1: Édouard-Léon Scott de Martinville''s phonautograph consisted of a barrel-shaped horn and stylus that etched images of sound waves onto a rotating piece of paper coated with soot.

The “Tech Page” column usually focuses on emerging technologies that might be applied to music and audio in the future. This month, however, I'd like to profile something from the past that paved the way to the recording tools now used by electronic musicians every day. I'm not talking about the work of Thomas Edison, who is credited with inventing audio recording, but rather of Édouard-Léon Scott de Martinville, a Parisian typesetter who managed to capture sound nearly two decades before Edison's famous recording of “Mary Had a Little Lamb” on a sheet of tinfoil.

Interestingly, Scott did not intend to record sound for playback. Instead, he wanted to create a visual representation, much like a waveform display today. To that end, in the 1850s Scott invented a device he called the “phonautograph,” which consisted of a barrel-shaped horn attached to a stylus (see Fig. 1a). Sound waves entering the horn caused the stylus to vibrate, which etched squiggles onto a sheet of paper covered with a layer of soot from an oil lamp (see Fig. 1b). The paper was mounted on a rotating drum that also moved horizontally along its axis as it turned, so the stylus traced a spiral, much like a wax cylinder.

Some of the images, called “phonautograms,” survive to this day, stored in a Parisian patent office and at the French Academy of Sciences. But no one paid much attention to the images until recently, when audio historian David Giovannoni and his team decided to find as many phonautograms as they could and attempt to re-create the sounds represented on them.

The resulting phonautogram was intended to allow the recorded speech to be transcribed.

The resulting phonautogram was intended to allow the recorded speech to be transcribed.

After digitizing the images with a high-resolution scanner, Giovannoni took the files to Lawrence Berkeley National Laboratory in Berkeley, California. There, Carl Haber and Earl Cornell analyzed the traces using an optical-imaging technique originally developed for particle-physics experiments and later applied to restoring and archiving other forms of early sound recordings. (For more on this, see “Tech Page: In the Groove” in the November 2004 issue of EM.) The oldest phonautograms, dating from 1853 and 1854, yielded only squawks, but Giovannoni found one from 1860 in more pristine condition. On it, a vocalist could be heard singing a fragment of the song “Au Claire de la Lune”.

This feat was accomplished by First Sounds (firstsounds.org), an informal organization of researchers dedicated to making humanity's earliest sound recordings available to all people for all time. Others working with Giovannoni, Haber, and Cornell include Richard Martin and Meagan Hennessey of Archeophone Records (archeophone.com), and Patrick Feaster, an expert in the history of the phonograph who teaches at Indiana University.

As mentioned earlier, Scott's aim was to represent sound visually, not record it for playback. As a typesetter and librarian who published a book on the history of shorthand, he apparently thought that sound recording might improve stenography, and that people could learn to discern what recorded voices were saying by looking at the traces left by the stylus. It's difficult to imagine how anyone could discern what was being said by looking at something like Fig. 1b, but the 19th century was the dawn of the Industrial Revolution, when mechanization seemed to be the answer to every problem.

The “Au Claire de la Lune” recording was made 17 years before Edison received a patent for the phonograph, though most historians do not believe that Scott's work obviates Edison's achievement. After all, the American inventor was trying to record sound for playback, a goal not shared by Scott. Still, hearing the scratchy, noisy reconstruction of the 1860 phonautogram transports us back to an era rife with potential that is only now being fully realized.

Show Comments

These are my comments.

Reader Poll

Do play more hardware or software synths?


See results without voting »