Just when you thought you've heard everything in music, comes a blast from the '60s....
That's the 1860s....
Almost two decades before Thomas Edison unveiled his tin foil cylinder phonograph, a little known French scientist named Édouard-Léon Scott de Martinville was also researching audio recording.
He built a device called a "phonautograph" that recorded sound waves.
However, unlike Edison, his device had no means of playback. His recordings were made on plate glass and later paper and stored, unplayed for nearly 150 years. There were no known means of playing the recordings without permanent damage to the extremely delicate grooves.
Finally, just a few years ago, with the aid of lasers and computer audio restoration, a 10 second snippet of a human voice singing "Au Clair de la Lune", recorded in February of 1860 was recovered from these recordings. An earlier 1859 recording of a tuning fork and possibly a bit of a human voice recorded in 1857 were also found. But that recording was too short to identify positively.
You can hear the "Au Clair de la Lune" recording below. Granted, the fidelity is extemely low, just barely recognizable. But it's history:
There is now a web site dedicated to the digitizing and preservation of phonautpgraph recordings:
The oldest playable recording up until then was the Lambert Talking Clock from 1878. Unlike Edison, who was recording and playing on weak tin foil at that time, Frank Lambert used a sturdy solid lead cylinder. Here is an MP3 of that (the audio on that too was barely recognizeable.):
And until the recent playback of the 1860 phonautograph recording, this 1888 wax cylinder of Handel's Israel In Egypt was considered to be the earliest known surviving music recording. Again it's low fi, but eerily beautiful:
It was recorded on a wax covered cardboard cylinder, hence the heavy surface noise - especially at the end.
It has been speculated that ancient etched pottery COULD hold sound vibrations from as far back as 1000 B.C. But that hasn't been proven yet. Pottery clay itself is among the worst substances to make a recording on and unless the person making the pottery was singing VERY loudly to tool etching a groove - a possibility, but a very distant one. I doubt it...