Listening to the OLD stuff
Very cool: a bit of sound from 1860 has been played back. The 19th century phonautograph couldn't play back sound, but made a visual record of waveforms; for first time audio has successfully been extracted.
The folks who did it: FirstSounds.org
Selected press coverage (not including the particularly sucky examples)
NY Times article
The Age, AU
NPR
Short crunchgear article
The press coverage has been interesting. I first heard about this from a wacky short tv news piece saying something like "Scientists have discovered a sound recording from 1860 ... Almost 20 years before Thomas Edison invented the first sound recording!" with no further explanation. Ms. Hollie witnessed me making an exasperated gesture at the tv set and saying, "What, have they extracted sound from a phonautograph? Or what??"
I can remember speculation going back at least 20 to 25 years ago that someday someone would be able to figure out some technology to extract audio from a phonautograph paper. That the phonautograph predated Edison's phonograph was no secret to those with some interest in early audio. This is not to discount the significance of the scientific achievement of playing it back -- it is more an observation of how the media tend to report things. No doubt if something significant and startling was discovered in old presidential papers from a 120 years ago, we'd see examples presenting it along the lines of: "Historians have discovered that the United States used to have a president called ''Grover Cleveland'', who has been totally forgotten!"
It will be interesting to see what else might come of these developments. I hope we'll get to hear some of the Edison tinfoil recordings again.
I note FirstSounds.org/Sounds already has a few other things up, the only one earlier recognizable as something is a tuning fork from 1859. Regarding an 1857 phonautogram, "his recording methods were not yet sophisticated enough at this time to yield audibly recognizable results." I wonder if this is an absolute threshold or one of current reconstructive technology.
There have been suggestions at least since the late 1960s that pots on a potter's wheel just might accidentally record sound. "Archaeoacoustics". A few archaeologists have contemplated that, just maybe, somehow, we may be able to listen to bits of conversation from thousands of years ago, perhaps listening to spoken Etruscan or Linear A. And other archaeologists and historians have found this dream, while tantalizing, pretty funny.
(Hm, doing a quick google while preparing this post has turned up a few things I've missed, including an April Fool's Day prank claiming recordings from Pompeii a couple years ago, and an "X-Files" tv episode with a pot with a recording of the voice of Jesus! upen.edu language log; Pottery recording)
I've long wondered if eventually better audio fidelity might be extracted from early recordings by some sort of computerized reverse engineering to compensate for the audio strengths and weaknesses of early recording devices.
Speaking of extracting hidden data from early audio, a dozen years ago a friend told me he was playing around with a NASA sonar program on his computer and tried it on some snippets of acoustic recordings-- where there was a pure tone like a chime or bell, and making a 2-d image. He said in a few cases he'd get a circle pattern, a few others a square. He thought he was getting a sonar picture of the inside of the recording horn.
The folks who did it: FirstSounds.org
Selected press coverage (not including the particularly sucky examples)
NY Times article
The Age, AU
NPR
Short crunchgear article
The press coverage has been interesting. I first heard about this from a wacky short tv news piece saying something like "Scientists have discovered a sound recording from 1860 ... Almost 20 years before Thomas Edison invented the first sound recording!" with no further explanation. Ms. Hollie witnessed me making an exasperated gesture at the tv set and saying, "What, have they extracted sound from a phonautograph? Or what??"
I can remember speculation going back at least 20 to 25 years ago that someday someone would be able to figure out some technology to extract audio from a phonautograph paper. That the phonautograph predated Edison's phonograph was no secret to those with some interest in early audio. This is not to discount the significance of the scientific achievement of playing it back -- it is more an observation of how the media tend to report things. No doubt if something significant and startling was discovered in old presidential papers from a 120 years ago, we'd see examples presenting it along the lines of: "Historians have discovered that the United States used to have a president called ''Grover Cleveland'', who has been totally forgotten!"
It will be interesting to see what else might come of these developments. I hope we'll get to hear some of the Edison tinfoil recordings again.
I note FirstSounds.org/Sounds already has a few other things up, the only one earlier recognizable as something is a tuning fork from 1859. Regarding an 1857 phonautogram, "his recording methods were not yet sophisticated enough at this time to yield audibly recognizable results." I wonder if this is an absolute threshold or one of current reconstructive technology.
There have been suggestions at least since the late 1960s that pots on a potter's wheel just might accidentally record sound. "Archaeoacoustics". A few archaeologists have contemplated that, just maybe, somehow, we may be able to listen to bits of conversation from thousands of years ago, perhaps listening to spoken Etruscan or Linear A. And other archaeologists and historians have found this dream, while tantalizing, pretty funny.
(Hm, doing a quick google while preparing this post has turned up a few things I've missed, including an April Fool's Day prank claiming recordings from Pompeii a couple years ago, and an "X-Files" tv episode with a pot with a recording of the voice of Jesus! upen.edu language log; Pottery recording)
I've long wondered if eventually better audio fidelity might be extracted from early recordings by some sort of computerized reverse engineering to compensate for the audio strengths and weaknesses of early recording devices.
Speaking of extracting hidden data from early audio, a dozen years ago a friend told me he was playing around with a NASA sonar program on his computer and tried it on some snippets of acoustic recordings-- where there was a pure tone like a chime or bell, and making a 2-d image. He said in a few cases he'd get a circle pattern, a few others a square. He thought he was getting a sonar picture of the inside of the recording horn.