Earliest record to be played tonight for first time in 134 years

October 26, 2012 • 11:00 am

According to CBC News,

People in Schenectady, N.Y., will get a chance tonight to hear something that hasn’t been heard publicly since 1878: a tinfoil recording made that year on what was then revolutionary equipment, Thomas Edison’s phonograph.

Gathering at the Museum of Innovation and Science, people will hear a rough-sounding 78 seconds of music and voice, as well as an explanation by Carl Haber of Lawrence Berkeley National Laboratory how he and his colleagues recovered the sound using digital techniques.

The tinfoil that Edison used was extremely thin and tolerated only a few plays before it began to wear down and even puncture

. . . The recording opens with a 23-second cornet solo of an unidentified song, followed by a man’s voice reciting Mary Had a Little Lamb and Old Mother Hubbard. The man laughs at two spots during the recording, including at the end, when he recites the wrong words in the second nursery rhyme.

“Look at me; I don’t know the song,” he says. At a time when music lovers can carry thousands of digital songs on a player the size of a pack of gum, Edison’s tinfoil playback seems prehistoric. But it opens a key window into the development of recorded sound.

“In the history of recorded sound that’s still playable, this is about as far back as we can go,” said John Schneiter, a trustee at the Museum of Innovation and Science, where it will be played Thursday night in the city where Edison helped found the General Electric Co.

This photo provided by the Museum of Innovation and Science in Schenectady, N.Y., shows Thomas Edison’s 1878 tinfoil phonograph. (Associated Press)


h/t: Lynn

37 thoughts on “Earliest record to be played tonight for first time in 134 years

    1. Sheeezzz! I don’t follow Meat Loaf, just know of him from his musical heydays and a few snippets here and there. But, I would not have guessed that and for some reason it is disappointing.

      Not like Gary Oldman “I AM VERY DISAPPOINTED” in The Fifth Element, but kind of like when you take a sip of a cheap Earl Gray tea that has the wonderful smell but no taste whatsoever.

  1. At the usual youtube prefix plus “/watch?v=7Vqvq-f-UtU”, you can hear the earliest recording there is, from 9th April 1860, of the song “Au clair de la lune” on Édouard-Léon Scott de Martinville’s phonoautograph.

    It was designed to record sound, but not play it back, bizarrely enough. Berkeley scientists digitally re-mastered it a few years back.

    1. “It was designed to record sound, but not play it back, bizarrely enough. ”

      I suspect in those days (before the invention of electronic amplifiers?) it was just technically easier to record than to play back. Simply because there would have been enough sound power for the recording process (‘just speak loudly into this trumpet please’) but not enough power available from a needle in a groove (or whatever method was used) to produce a usefully audible sound.

  2. I expect you could probably work up a laser playback system that could read the tinfoil without damaging it further.

    1. That was my first thought, too…but then I quickly followed it with thinking that you could take high-resolution photographs from different angles to construct a three-dimensional model of the recording, and that, once you have said model, you can do all sorts of interesting analyses.

      I would hope that whatever method they’re going to use to play back this recording, it doesn’t involve physical contact.

      I’d also love to hear that cornet solo!


      1. The first I heard of what is now CD technology was a mention of a new type of turntable that would read the groove with a laser wothout making contact with the vinyl. Supposedly it would not only reduce the potential for scratching, but also be able to ignore existing scratches and skips. The music would still have been analog. Instead, digital took over.

        1. I vaguely remember hearing of the same. I also vaguely remember that a few such turntables actually made it to market, but that they were so ludicrously expensive that nobody ever bought them.

          Probably not a bad thing for an archivist to have…except that I’d hope that most archivists would have access to the original master plates and / or studio tapes….


          1. See http://en.wikipedia.org/wiki/Laser_turntable

            I’ve once attended a less that fully convincing demo of an early ELP system. My understanding is that they have achieved some practical level of workability in the meantime.
            The panel of which I was a member at the time had to decide on means of digitising early audio archives. In the end, conventional Simon Yorke transcription turntables won the day. I believe the Library of Congress uses them too.

            Speaking of the vinyl era: the problem of master tapes is that some, not necessarily the earliest, degrade at a much faster rate than vinyl, which is basically inert under adequate storage conditions. Of course, for many of these tapes, only one copy exists: the original. (We’re not talking Abbey Road here.) I’m in the process of rescuing an ethnomusicological tape archive, and the oxide comes unstuck from the tape support. You have one go at transcription, and that’s it.

            1. Thanks for that Wikipedia article; it matches what I recall.

              I’ve heard of the problems with tapes, too, and I’ve wondered: We’ve had some very significant advancements in electronic signal processing in the past 3/4 century. Shouldn’t it be possible to design something that doesn’t require physical contact with the tape in order to read the signals encoded on it?

              We’ve also had some pretty significant advancements in materials science. Can’t the tapes be somehow fixed / encased / whatever?

              I’m sure the answers would be “yes” for somebody at JPL with their budget ,with the main question being whether or not anybody with the interest to do something about it can afford the R&D to make those interests a reality….


              1. If you search for ‘tape baking’, you’ll find a number of mundane recipes, some of which actually do work on some kinds of tape.

                Since you mention “significant advancements in materials science”, it is ironic that such advancements may just be responsible for part of the problem. New polyurethane binders used in the 1970s and ’80s were the proximate cause of oxide shedding. Ampex was the most prominent culprit — and victim. I’m working with BASF and Agfa tapes from the 1950s and ’60s that are mechanically as good as new.

                The point is, it would be incommensurably cheaper to allocate the very modest resources required to digitising what’s left of our analogue tape legacy rather than trying to tackle the problem of adhesive polymer decay head-on.

        2. There actually *are* analog laser-read disks, namely the Laserdisc format. Originally both video and sound were stored with an analog technique, later, digital sound was added.

          1. I remember a movie laser disc player connected to a large projection TV proudly displayed at a wealthy friend of a friend’s party outside of LA in the later ’80s. The disc was about the size of an audio LP disc. Can’t remember the movie that was playing, may have been Star Wars, but remember being impressed by how good the picture and sound were.

            Laserdisc made it out of the box, then quickly disappeared. Never realized it was an analog recording technique, but that makes sense.

  3. I have an imprecise recollection of reading an article in (Science News? Discover?) about some archeologists who observed that the thin groove on a ceramic ancient Greek water or wine jar had been produced by having a needle do the engraving while the potter’s wheel turned. Since the depth was very regular the needle clearly was not being held by a human…but there were still small variances in the groove…. Supposedly they used a laser to read the depth and tried playing the information back as audio and got a result that sounded enough like people talking and a dog barking that the people who scoffed at the idea that the groove could be a recording were left wondering if it could be….but the sound reproduction was still bad enough that the people who wanted the groove to be a recording had to admit that they were not convinced that it was an inadvertent recording. Unfortunately I have not even the vaguest citation for this….

    1. The plot of the X-FIles episode Hollywood A.D. is partly based on a piece of ancient pottery that supposedly recorded the voice of Jesus in just such a fashion. Instead of being a real plot, it was part of a movie based on the adventures of Scully & Mulder.

    2. As long as we’re trading vaguely-remembered news items… I vaguely remember that the pottery in question was Egyptian, but that the article appeared in a April 1 issue of the journal of the Audio Engineering Society, i.e. April fool!.

    3. Largely a canard, but one to which every new generation of archaeologists is susceptible.

      I first read about it in a Soviet ‘popular science’ magazine for kids in 1966 (NB: before the Woodbridge letter of 1969). Soviet-era sci-fi was rather good at such things.

      Dreamed of investigating it myself. Which I did, as a young archaeologist. But the real physics of pottery making are defeating.

      A good overview of experiments and hoaxes:

  4. Mechanical recording in general fascinates me. The fact that sounds can make an impression on a plastic medium and can then be reproduced by rubbling a piece of metal across it seems more wonderous to me than any other invention.

    You may have seen that the LOC has put online its collection of Victor cylinders some time ago (http://www.loc.gov/jukebox/).

    There is also an outstanding collection and archival project at UC Santa Barbara (http://cylinders.library.ucsb.edu/).

    I remember probably about 10 years ago hearing on NPR that the LOC had been struggling to find a format/medium for permanent preservation of sound recordings. With the changes in tape media and then optical media, there was a concern about choosing a standard that would accessible over the long term. The answer? Good, old-fashioned shellac discs. The goal here, course, is not fidelity, but accessibility, and all you need for playback is a needle and horn speaker.

    (Incidentally, this is an issue at the National Archives, too, where they get accessions of older electronic storage media that have fallen into disuse.)

    1. I think I heard the same NPR story that you did, about the Library of Congress transferring all the most important recordings to 78s (or maybe wax cylinders) for the ultimate in long-term storage and playability. Alas, it was broadcast on April 1. It was a few years, to the day, after their story on pickle trees in Montana, I believe.

      1. Oh noes! But it makes such sense!

        Alas, you are correct. A 2003 April Fools prank. That may be the longest it’s ever taken me to get a joke.

  5. I recently read that when Edison’s recordings were first played back to interested parties touring his lab, the astonishment was such that women often fainted at the “magic” of the device.

  6. as an audiophile, i think today’s quality of recording and playback formats are awful. Vinyl is garbage and so are all the digital formats . A real audiophile listens to wax cylinder.

    1. You’re a piker! I have a full complement of musicians in stasis tubes in my basement, ready to play any live music I want at a moments’ notice.

      1. ya i’m sure they sound great with their compression, gated reverbs, and their pro tools to capture their memorable performances…….please. ifthere ain’t no wax, it lacks.

    1. SWEET! Thank you!

      I can tell that the cornettist is doing some triple tonguing, but I don’t recognize the solo. It might have been something improvised on the spot…but maybe, with the (hopefully) wide attention this is about to get, somebody might actually recognize it.

      Also…did that laughter sound forced and / or theatrical to anybody else?


  7. “In the history of recorded sound that’s still playable, 1878 is about as far back as we can go,” said John Schneiter, a trustee at the Museum of Innovation and Science” ..

    Hmm. Would a 1200 year old bird call recording qualify? Many find the evidence for intentional design to be compelling.

    A chirped echo at Chichen Itza’s Kukulkan pyramid is coded into the design of its four long staircases (tread length and riser height). You can hear and see it at


    A handlap there stimulates an echo that sounds like the primary call of the resplendant quetzal, a bird venerated by the Maya and their predecessors for over 2000 years.

    Older non-Maya pyramids have chirping staircases too. But the echo at Kukulkan became strongest after two of its four decaying staircases were renovated.

    David Lubman


    1. Some of those echoes are certainly remarkable. Whether they’re intentional or just an artefact of the step layout (or possibly the first discovery was accidental and it could have been deliberately replicated later) is something one could argue about (and I’m sure it has been).

      But I’d say the sound is synthesized rather than recorded.

  8. Edison was not the first person to come up with this idea; he was merely the first with the resources to produce a working version.

    Édouard-Léon Scott de Martinville invented the “phonautograph” for producing a visual recording of sound and from this, Charles Cros developed and registered the “paleophone” which would it seems (had he been able to afford to produce a working version), have actually been better than Edison’s effort.

    Edison is rather overrated. To see how adrift he was, read about his farcical efforts to discredit Tesla’s alternating current power production in favor of his own d/c version.

Leave a Reply to John DanleyCancel reply