A Ghostly Voice from the Past

O.K. Folks,

This blog is about something so strange even I have to say it's ghostly. I read in the newspaper (the new york Times) about this recording that was done in France in 1860 that pre-dates Edison's tin-foil experiment of recording "Mary had a little Lamb".

Seems this Frenchman names Scott, was a tinkerer and Librarian and he was coming up with a way to write down what the human voice was saying, but this was so it would be deciphered by some new means in the future, Scott was not thinking of an audio recording but a type of writing form in which people did not have to write their words down, just speak into a tube and it would be written down by a machine and then scientists in the future would find a way to translate the waves into the written word.

Little did he know, so he used a roll of heavy duty paper and black soot put it on a small hand cranked drum placed a fine needle point on it that was attached to a speaking tube and then it was cranked while someone either spoke, sang or played an instrument into the tube.

The result was wavy lines that recorded the voice, much like the up and down grooves on a record. This Scientist---whose name I forget----is an audiophile researcher, he had heard about Scott's experiment and have come across several of these lampblack soot recordings, several done in 1857 and 1858 were of people talking and someone playing a guitar, but they were not very clear, all there was pops and hiss.

Now how did this Scientist convert these recordings? Well he created a virtual "needle", using light (much like how the sound is transfered to movie film), separated the sound (using computers) into 16 tracks to separate the hiss and pops, allowed for the unevenness of the hand cranking and did find something. So this was promising.

But he needed to find a really clear lampblack recording, he found out that Scott had left one with a diagram of his invention at this institute for mechanical science in France, and the recording was a gem.

He was able to reproduce a voice of a young girl, possibly Scott's daughter, singing an old french folk song called "Au Clair du lune" and the line is about seeing Peirrot by moonlight.

If you go to the New York Times Website for March 28, 2008 they will have attached two recordings---one from 1931 (very clear) of a opera singer singing this song (listen to this one first) that will give you an idea of how it is suppose to sound. The second is Scott's recording it runs for 10 seconds and it is ghostly, it is the tune, but there is something so eerie about it.

I had to play it several times just to feel comfortable about it, the voice of a young girl singing an old french folk tune from almost a hundred and fifty years ago before Edison developed the phonograph. And preserved on a flimsy piece of paper with easily smuggable soot, how quick it could be to wipe away something like that with one careless swoop.

It made me think of those ghost hunters EVP's (electronic voice phenomena), but this is real from a girl long dead and now just reaching out to us across the expanse of time.

Well the scientist presented this finding to about 150 scientists and audio researchers and was greeted with resounding applause, I don't know if it was for the fact that the recording was recoverable and the scientists techniques worked, if it was for Scott or for his long dead daughter's performance---maybe all three.

Oh and don't worry about Edison's place in history, his is still solid, because he was looking for recording sound, not writing sound, but when I looked at the picture of the recording it looks much like how we "read" sound now, so they were both on the same track this just pushes it back further.

Another thing I was thinking about is that some writer is going to write a suspense story (a la Da Vinci Code) using this re-discovered science. I'm looking forward to that.