Forgotten audio formats: Wire recording

From espionage to home recording, the colourful life of the longest-used audio medium.

Read the whole story

 

[dlux]

Peevester[/url]":2kufv9qg]

Wardatrigger[/url]":2kufv9qg]And in a couple more months, Ars will cover music for oscilloscopes.

Which is currently going nuts on youtube - search for Jerobeam Fenderson.

That's pretty impressive. The originating site: http://oscilloscopemusic.com


One antecedent of this is 'Waveform of the Week', which was broadcast on MIT's radio station, WTBS.

(The history of that radio station is likewise interesting, including its call letters: http://wmbr.mit.edu/history.html )

 

[BulkyZ]

Great article. Really, this is the reason I come to Ars.

 

[Antron Argaiv]

Hat Monster[/url]":131citw6]

cbreak[/url]":131citw6]

Jim Salter[/url]":131citw6]

cbreak[/url]":131citw6]The article doesn't actually describe the format, unfortunately, but there certainly is a way in which the data is encoded.

No, there isn't. It's direct amplitude modulation. There's no encoding whatsoever.

An example of encoding would be the old "eight, none, and one" of telephone modem days - referring to eight signal bits, no parity bit, one stop bit. There's absolutely none of that in wire, tape, phonograph, or AM radio - you literally just translate the amplitude modulation of the medium directly into movement of a speaker driver, and presto, you get sound.

It's arguable whether you could constitute FM radio as "encoded" or not - the amplitude modulation of the recorded sound is translated into frequency modulation of a carrier signal, but there's still no encoding to interpret around it, it's just the direct waveform of the original sound on an underlying medium.

Amplitude modulation IS an encoding. Not every encoding is digital.

AM isn't an encoding, it's a modulation. You can have both, but you must have the modulation (modulation is how a different medium is changed to accept the signal: If you do no more than that, you haven't encoded anything). An encoding, as the name suggests, translates the signal into another form where it's a symbolic representation of the original. FM and AM are both not encodings, they're direct transfers into another medium. The offset of the signal controls the power of the AM signal, or the offset of the FM signal, directly.

A very simple encoding would be something like PWM, where the length of the pulse represents the power of the signal. You'd think that's a digital form, but it's still analog: The length of the pulses is your analog component. This can be demodulated by nothing more complex than a speaker cone's inertia, so represents a form of encoding and modulation where the act of demodulation is also the act of decoding.

Another common, and simple, encoding is the digitisation and capture of the signal into a digital form, such as LPCM. LPCM is just a long list of numbers, so it's a valid encoding. Being a storage, rather than transmission, format, it has no modulation.

A more complex encoding is a QAM constellation, where each state of the modulation can be one of 4, 16, 64, 256 or even more states, encoding bits into each individual state (it's similar to a look-up table, and uses one).

LPCM = Linear Pulse Code Modulation...LPC is the encoding, PCM is the modulation. Compare with ADPCM=Adaptive Differential Pulse Code Modulation (ADPC encodes the difference and a scaling factor, between successive samples). LPCM is also Logarithmic Pulse Code Modulation, as used in telephone networks - the difference between code values varies logarithmically, putting more code values close to zero, where zero crossings contain much of the speech data, and fewer codes at higher levels, where fidelity is not so crucial to understanding the speech.

QAM is really a modulation scheme for digital data -- the underlying data can be anything capable of being encoded digitally -- audio, image, or random data.

 

[pusher robot]

Peevester[/url]":2vdjmc11]

Umbaglo[/url]":2vdjmc11]Is this format what was being referred to by all my old tape players that had a switch for various tape formats that included "Metal" as one of them? I don't remember ever seeing any alternate tape format at the time, but I remember thinking "Metal" was certainly a strange one specifically.

No, metal is Chromium Oxide tapes, nothing to do with wire recording. A wire recording looks like a big spool of shiny thread, and is not in a cassette.

I thought it was for my Iron Maiden cassettes.

 

[Mungus the Unhyphenated]

flashback_rtk[/url]":22mazxlh]

a home-recording medium that people could, and did, use to record their own voices, as well as their own music and songs off the radio—thus giving the world its first example of illegal home recording.

Was there any law preventing it at the time? You make it sound as if it was the first observation of a natural phenomenon in a laboratory.

The less complicated tape decks were also getting cheaper and soon became the recorder of choice for universities, schools, independent studios, and, eventually, home recordists who fancied stealing sounds from the radio.

And that's why I ask, insisting on it makes me think you're just trying to educate your readers on your opinion on recording songs from the radio.

It was a very interesting article by the way.

I'm suspecting (hoping?) the /s was implied.

 

[nbs2]

A quibble and a question

Quibble - usage into the 70s yields an eventual lifespan of 72-81 years. Vinyl, at 67, is neither a relative spring chicken or dead, and will likely surpass wire recording's lifetime.

Question - mention is made of a debated stored program computer. Can someone provide more information about this?

 

[WaveRunner]

Antron Argaiv[/url]":1wsbi29g]

Hat Monster[/url]":1wsbi29g]

cbreak[/url]":1wsbi29g]

Jim Salter[/url]":1wsbi29g]

cbreak[/url]":1wsbi29g]The article doesn't actually describe the format, unfortunately, but there certainly is a way in which the data is encoded.

No, there isn't. It's direct amplitude modulation. There's no encoding whatsoever.

An example of encoding would be the old "eight, none, and one" of telephone modem days - referring to eight signal bits, no parity bit, one stop bit. There's absolutely none of that in wire, tape, phonograph, or AM radio - you literally just translate the amplitude modulation of the medium directly into movement of a speaker driver, and presto, you get sound.

It's arguable whether you could constitute FM radio as "encoded" or not - the amplitude modulation of the recorded sound is translated into frequency modulation of a carrier signal, but there's still no encoding to interpret around it, it's just the direct waveform of the original sound on an underlying medium.

Amplitude modulation IS an encoding. Not every encoding is digital.

AM isn't an encoding, it's a modulation. You can have both, but you must have the modulation (modulation is how a different medium is changed to accept the signal: If you do no more than that, you haven't encoded anything). An encoding, as the name suggests, translates the signal into another form where it's a symbolic representation of the original. FM and AM are both not encodings, they're direct transfers into another medium. The offset of the signal controls the power of the AM signal, or the offset of the FM signal, directly.

A very simple encoding would be something like PWM, where the length of the pulse represents the power of the signal. You'd think that's a digital form, but it's still analog: The length of the pulses is your analog component. This can be demodulated by nothing more complex than a speaker cone's inertia, so represents a form of encoding and modulation where the act of demodulation is also the act of decoding.

Another common, and simple, encoding is the digitisation and capture of the signal into a digital form, such as LPCM. LPCM is just a long list of numbers, so it's a valid encoding. Being a storage, rather than transmission, format, it has no modulation.

A more complex encoding is a QAM constellation, where each state of the modulation can be one of 4, 16, 64, 256 or even more states, encoding bits into each individual state (it's similar to a look-up table, and uses one).

LPCM = Linear Pulse Code Modulation...LPC is the encoding, PCM is the modulation. Compare with ADPCM=Adaptive Differential Pulse Code Modulation (ADPC encodes the difference and a scaling factor, between successive samples). LPCM is also Logarithmic Pulse Code Modulation, as used in telephone networks - the difference between code values varies logarithmically, putting more code values close to zero, where zero crossings contain much of the speech data, and fewer codes at higher levels, where fidelity is not so crucial to understanding the speech.

QAM is really a modulation scheme for digital data -- the underlying data can be anything capable of being encoded digitally -- audio, image, or random data.

I'm pretty sure every bit of literature reserves LPC for Linear Predictive Coding...

 

[Thomas Harte]

Antron Argaiv[/url]":3o8u8aad]QAM is really a modulation scheme for digital data -- the underlying data can be anything capable of being encoded digitally -- audio, image, or random data.

QAM doesn't imply anything about being digital — it's amplitude modulation plus a second channel of amplitude modulation, stored in quadrature (i.e. at a phase offset of 90 degrees) from the original. So it's just two channels of AM information composited into one signal.

The colour part of NTSC and PAL television is QAM encoded. It's definitely not digital.

 

[Hopefully Smarter]

According to the Firesign Theatre, the Aztecs invented the wire recorder.

citation: Everything You Know Is Wrong (1974)

 

[flashback_rtk]

Mungus the Unhyphenated[/url]":2h60nu3f]

flashback_rtk[/url]":2h60nu3f]

a home-recording medium that people could, and did, use to record their own voices, as well as their own music and songs off the radio—thus giving the world its first example of illegal home recording.

Was there any law preventing it at the time? You make it sound as if it was the first observation of a natural phenomenon in a laboratory.

The less complicated tape decks were also getting cheaper and soon became the recorder of choice for universities, schools, independent studios, and, eventually, home recordists who fancied stealing sounds from the radio.

And that's why I ask, insisting on it makes me think you're just trying to educate your readers on your opinion on recording songs from the radio.

It was a very interesting article by the way.

I'm suspecting (hoping?) the /s was implied.

There was nothing sarcastic in my post, the question is legit (and still unanswered), was there actually any law preventing the recording of a radio broadcast (music pieces particulary) on wire by that time? Usually, when some new technology appears some time passes until laws regarding its impact on industrial collectives and people get created.

Also, I found the article very interesting and I stated it because I didn't want people, especially the author, to think I was denying its quality by pointing out sections where I suspect he is signaling his opinion.

I don't understand your post, what in mine made you imply I was being sarcastic even to the point of hoping I was?

 

[marsilies]

Peevester[/url]":1tj62ijq]I'm listening to one of the Guthrie songs in the background, and it's giving me chills hearing a 60 year old recording that sounds that good. Does anyone have a link to the uncorrected audio?

This podcast page has links to both uncorrected and corrected audio for a 15 second clip of the concert. It's just of Guthrie talking, not singing, but the difference is astounding:
http://www.ams.org/samplings/mathmoment ... my-podcast

 

[Z1ggy]

one of the reasons i keep coming back to ARS. I love these articles.

 

[ChadD]

bongbong[/url]":30x6lq8l]is it really accurate to call wire recording or tape or vinyl, an audio format?
Im not trying to sound smart alecky or be a nitpicker.
but is this the proper form so i can use it without error since the audio formats i know are flac, wav ,mp3 etc which I KNOW are file fomats.
wont audio device be a better term?

BTW their recovery methods are brilliant and awe inspiring

Of course... the word format does predate the internet. lol

We have always used it when comparing any two items that perform similar jobs in a different non interchangeable way.

Also before the internet people said things like, does that movie store rent VHS or Beta format movies. Old computer magazines used to publish basic language programs that would state which format they where intended for.

So yes audio format is the correct term. I guess if you wanted to get technical you could say Analog audio format. Although if you called Mp3/Flac/ogg ect digital audio format... then what would you classify something like a Sony Mini Disc reader using ATAC... Physical digital audio format?

 

[marsilies]

flashback_rtk[/url]":3sgjgn6t]There was nothing sarcastic in my post, the question is legit (and still unanswered), was there actually any law preventing the recording of a radio broadcast (music pieces particulary) on wire by that time? Usually, when some new technology appears some time passes until laws regarding its impact on industrial collectives and people get created.

It's complicated, since the copyrights for music recording are currently handle by state law for recordings up through 1972. However, since music recordings are derivative works of the original musical compositions, they woud've been covered by copyright of the compositions, and so home recordings of the radio would've at least been violating that copyright. Radio in the US, for example, pays music publishers licensing fees, but doesn't pay for licensing of the actual recording.

http://www.digmedia.org/issues-and-poli ... t-in-music

Except for in the case of terrestrial, analog radio, a separate license must be obtained from the BOTH the copyright owner of the “musical work” AND the “sound recording” as described above, before a particular sound recording of a musical work can be used.

More info:
https://library.osu.edu/blogs/copyright/2013/05/15/198/
http://www.recordingcopyright.org/pdf/AMuscop_fnl.pdf

 

[unequivocal]

I believe modern commercial airplanes use this technology in their black boxes. I think the stainless steel wires are on a continuous loop, so that the black box records some number of hours of data on the wire and the records over the old data.

The advantage is that even if the black box itself is ruptured, the data on the wire wire, even if the wire is severed, will still be recoverable and playable, unless it's been heated over the curie temperature for stainless steel.

 

[b.goody]

Sorry, this has been bugging me for a couple of days. "Wire tapping" had nothing to do with the recording medium. The original use had law enforcement connecting a headset directly to a phone junction or wire. A physical tap into the communication wire. "Wearing a wire" did come from the use of small wire recorders and that may be where the confusion originated.

 

[markratledge]

Old Woody Guthrie sounds like Bruce Springsteen

 

[WaveRunner]

Thomas Harte[/url]":2k243g2r]

Antron Argaiv[/url]":2k243g2r]QAM is really a modulation scheme for digital data -- the underlying data can be anything capable of being encoded digitally -- audio, image, or random data.

QAM doesn't imply anything about being digital — it's amplitude modulation plus a second channel of amplitude modulation, stored in quadrature (i.e. at a phase offset of 90 degrees) from the original. So it's just two channels of AM information composited into one signal.

The colour part of NTSC and PAL television is QAM encoded. It's definitely not digital.

Well no it's not just a second signal at a phase offset of 90 degrees. It's a combination of phase and amplitude offsets according to the limits of the channel and hardware on both ends. What you described only applies to two signals, in common use today you end up sending far more than just two signals (constellations). That sound we use to hear modems make is the establishment of those channel hardware limits using known handshake signals.

 

[Antron Argaiv]

Thomas Harte[/url]":1eurtkqc]

Antron Argaiv[/url]":1eurtkqc]QAM is really a modulation scheme for digital data -- the underlying data can be anything capable of being encoded digitally -- audio, image, or random data.

QAM doesn't imply anything about being digital — it's amplitude modulation plus a second channel of amplitude modulation, stored in quadrature (i.e. at a phase offset of 90 degrees) from the original. So it's just two channels of AM information composited into one signal.

The colour part of NTSC and PAL television is QAM encoded. It's definitely not digital.

Excellent point. AKA Phase-Amplitude Modulation. Two orthogonal channels.

And yes, LPC is almost always Linear Predictive Coding (in fact, that's what I thought he was referring to, until I Googled...guess what comes up first?)

 

[BeowulfSchaeffer]

Loved the Woody Guthrie clips. I'll have to purchase the recording.

 

[mltdwn]

I still have my grandfather's wire recorder sitting in my study. I never use it and have no spools for it but it is a magnificent thing and still in mint condition. They knew how to make products back then. Course I also have his 1950s era circular saw that is made of chromed steel and still works like a champ.

 

[Elsmarc]

charltjr[/url]":deb8kl93]Love this sort of "lost tech" piece, more please.......

Techmoan did a really interesting video on a wire recorder (well, I thought it was interesting )

https://youtu.be/90ihiTwJPCc

I'm with you! This piece blew my mind it was so unexpected.

Closing in on 70, I remember a woman I met about 35 years ago whose father, who was deceased, had a wire recorder. She had it and a bunch of "home" tapes he made when she was a kid stored in her basement. I tried a few spools and they worked (and were quite clear/audible). First time I had heard of them.

 

[CheyenneWills]

I remember as a kid, playing around with one that my grandfather had (he owned a radio repair company back in the mid 1920's). He had a whole bunch of interesting electronic toys. I also remember building a tesla coil with him when I was in 6th grade.

Anyway .. thanks for bringing back fun memories

 

[dlux]

b.goody[/url]":mp4u9t37]Sorry, this has been bugging me for a couple of days.

I see what you did there.

 

[StuartDole]

I had one as a kid - a discarded dictation machine. We played with the wire, recorded things and then reversed the wire to listen to them backwards. "Splicing" the wire was done with a square knot, then trimming the ends. Yeah - you could fit a lot of time on a spool!

We didn't try music - weren't into it that much.

Many thanks for the article!

(PS: In those days a remember the ads for home computers in Scientific American - they were analog panels! Nothing digital. A girlfriend had one - we never figured out how to do anything interesting with it. It was many years later I first got to play with the TTL circuits in a PDP-8.)

 

[twinstronglord]

Hopefully Smarter[/url]":2dq85ffa]According to the Firesign Theatre, the Aztecs invented the wire recorder.

citation: Everything You Know Is Wrong (1974)

I came here to see or make the FT reference. I was hoping I wasn't the only person that would immediately make the connection. Triggered one hell of a nostalgia trip. I recall learning a great deal about audio formats and converting analogue to digital just to preserve my sadly inherited Firesign Theatre albums.

Great article, but what was missing was a look at how they work. But again the wonderfully curious (and pedantic) Ars readers are on the ball with links.

And may as well drop a favorite line from that album

"Hush now. It's a new world, Honey. Nobody gowan have to be a slave all the time no more. We gowan to take to toins. And guess who's toin it is now! Ahuaw huaw huaw!"

 

[palisade]

When I was listening to Woody Guthrie's "Grand Coulee Dam" I pulled up the lyrics on google and found they didn't match. The lyrics I'm hearing go something like this:

This old world has seven wonders a travelling salesman tells
Some gardens and some towers (flowers?), I guess you know them well
But, ahhhh the greatest wonder in Old Sam's fair land
Oh in King Columbia River it's that big Grand Coulee Dam

She heads up the Canadian Rockies where the rippling waters glide
Comes a-rumbling down her canyon to meet that salty tide
Of that wide Pacific Ocean where the sun shines in the west
And that big Grand Coulee country that's the land I love the best

She rides down a granite canyon and she bends across the lean
Like a silver running stallion down her seabed she does sheen
Catched a ride upon the biggest thing yeah built by human hands
It's that King Columbia River and the big Grand Coulee Dam

In the misty crystal glitter of that wild and windward spray
We carved the mighty history of the sacrifices made
Well she ripped our boats a-splinter, so she'd give us things to dream
Of the day that Coulee Dam would cross that wild and wasted stream

Uncle Sam took up this challenge in the year of Thirty three
For the farmer and the worker and for all of you and me
He said, "Roll along Columbia. You can ramble to the sea."
But river while you're rollin' you can do some work for me

Now up in Washington and Oregon you hear the factories hum
Making chrome and making manganese and light aluminum
And there flies a super rocket ship across this land of pots and pans
Spawned upon that King Columbia by the big Grand Coulee Dam

In the misty crystal glitter of that wild and windward spray
We carved the mighty history of the sacrifices made
Well, she tore our boats to splinters but she gave us dreams to dream
Of the day the Coulee Dam would cross that wild and wasted stream

 

[real mikeb_60]

Great article!

Glad the "wiretap" thing was corrected. That one bugged me when I saw it. Also, "on the wire" may have meant on the recording in some cases, but also frequently meant that something was transmitted as in "put that story on the wire" referring to AP/UPI/etc. wire services. Ambiguous without context.

Difference between modulation and encoding as I see it: if the original signal is recoverable and understandable without processing, even if it's difficult to understand, it's modulation. If it requires some kind of processing to recover ANY signal (i.e. if without processing the signal is gibberish), it's encoded. There's probably a grey area where a modulated signal is adjusted in such a way, or so noisy, that it might be impossible to understand without doing something to it first - indistinguishable from encoding, though perhaps not intentionally so.

The old wire recordings of Woody might be thought of as a poor analog encoding job. The original signal was there "in the clear" but was so buried in noise and other problems that it wasn't usable without processing. Lots of forensic audio work is in that problem space. As fixed, it's a great record!

 

[real mikeb_60]

Jim Salter[/url]":qre5zwxd]

cbreak[/url]":qre5zwxd]Amplitude modulation IS an encoding. Not every encoding is digital.

No, it really isn't - amplitude modulation is exactly what it says on the tin - modulation. The only way to label what goes on in wire, phonograph, or standard two-channel cassette "encoding" is if you want to get super pedantic and call it "direct" encoding - the equivalent of multiplying a number by one.

Encoding is something above and beyond the original data, which needs to be interpreted or decoded rather than being just blindly played back as part of the main data stream. And no, that doesn't need to be digital - an analog example is the old Dolby Pro Logic encoding that hid an extra couple of channels inside a two-channel recording medium.

Edit: I forgot about recording bias to tape, as pointed out above. Bias would qualify as extremely simple encoding.

Before Dolby Pro Logic, there was Columbia SQ quadraphonic. It didn't do a really great job, but also didn't require a special cartridge with response to 50 kHz to play back - ordinary equipment would work. It was essentially a matrix system.

Impecunious college students in my era got a similar effect by adding a 3rd speaker in the rear, wired across the hots of the amplifier (with a L-pad to adjust level in fancier setups), to get phase effects that were interesting. Moody Blues had some songs that moved around the dorm room in strange and wonderful ways.

As for bias: is it encoding if nothing has to be done to reverse it at the playback stage? Bias is only used during (magnetic tape) recording; it's irrelevant during playback unless it's at a low enough frequency to require filtering to exclude it from the playback.

 

[Hinton]

What makes Valdemar Poulsen a Danish-American? I can't find him having any relation to USA on either the English or Danish version of Wikipedia.

 

[ChadD]

unequivocal[/url]":23l9pxzh]I believe modern commercial airplanes use this technology in their black boxes. I think the stainless steel wires are on a continuous loop, so that the black box records some number of hours of data on the wire and the records over the old data.

The advantage is that even if the black box itself is ruptured, the data on the wire wire, even if the wire is severed, will still be recoverable and playable, unless it's been heated over the curie temperature for stainless steel.

From the article;
"Wire’s ability to absorb far greater heat than tape and, if used with certain recorders from the Armour collection, to keep going for hours on end, meant it was still useful for certain black-box applications; it was still being used in planes as late as the early 1970s."

It was tech that was cutting edge in the 50s... some of the airlines fleets included planes of that vintage up until the early 70s. Wire wasn't nearly as durable as may first guess. Yes it was SS, but your talking about some very very fine wire... if it did break it would be a nightmare trying to splice back together.

Current boxes use solid state memory and record a few hours of voice and 24 hours or so of data. Previous tape units only recorded 30 mins or so of voice in a loop as you suggest. I don't know exactly how long the old wire models recorded. The article suggest some models recorded for hours. I assume the CVR portion of the box had to be swapped out between flights. The first black boxes used steel foil, which was much the same concept but had less issues with broken wire.

 

[baksdesign]

jerminator[/url]":3sg601s3]Very interesting read on a technology that I've seen in old movies, and didn't think it was quite this ubiquitous. Good stuff. I liked the Guthrie guitar sticker "This machine kills fascists." We'll be seeing more of that sentiment soon again looking at where we're all headed.

Yep Obama and Hillary are out, and next one will be Trump.

Looking forward to the total destruction of two party system in future.

 

[stormcrash]

One other use of magnetic wire recording during World War II was by the Ghost Army (23rd HQ Special Troops) for their sonic deceptions. They needed a playback mechanism for the sounds of troops and equipment movement to sound like a large force was secretly amassing in a location (kind of ironic, sound like a secret movement). They used wire because unlike records was not prone to skips or other similar playback issues that would give away the deception.

 

[dlux]

stormcrash[/url]":34zuvpok]...the Ghost Army...

Those guys were stealthy, but also strong:

(That's a topic that's fascinating on its own.)

 

[MCYL]

Brings the term "wiretapping" a new context.

 

[DNSGeek]

Jim Salter[/url]":3ehje8ks]

fragile[/url]":3ehje8ks]Coming from an era where everything was analogue, be it reel-to-reel tape, cassette tape, records etc, then this is definitely a format.

Digging back in memory, I think what audio engineers used to call "format" in the analog days wasn't ever really a tightly-defined term. It generally encompassed physical medium, encoding, and modulation in one all-encompassing if somewhat vague term - because you never really had much call to separate the layers back then; they all went together.

For that matter, the "encoding" was usually "none whatsoever" when it came right down to it. Wire or tape are both direct modulation of a magnetic medium with no encoding. Phonograph is direct modulation of a physical medium with no encoding.

Actually, most phonograph recordings are indeed encoded, using equalization curves for better low-frequency reproduction. See https://en.wikipedia.org/wiki/RIAA_equa ... RIAA_curve for information on the most common curve.

 

[flashback_rtk]

marsilies[/url]":2ylhkr6r]

flashback_rtk[/url]":2ylhkr6r]There was nothing sarcastic in my post, the question is legit (and still unanswered), was there actually any law preventing the recording of a radio broadcast (music pieces particulary) on wire by that time? Usually, when some new technology appears some time passes until laws regarding its impact on industrial collectives and people get created.

It's complicated, since the copyrights for music recording are currently handle by state law for recordings up through 1972. However, since music recordings are derivative works of the original musical compositions, they woud've been covered by copyright of the compositions, and so home recordings of the radio would've at least been violating that copyright. Radio in the US, for example, pays music publishers licensing fees, but doesn't pay for licensing of the actual recording.

http://www.digmedia.org/issues-and-poli ... t-in-music

Except for in the case of terrestrial, analog radio, a separate license must be obtained from the BOTH the copyright owner of the “musical work” AND the “sound recording” as described above, before a particular sound recording of a musical work can be used.

More info:
https://library.osu.edu/blogs/copyright/2013/05/15/198/
http://www.recordingcopyright.org/pdf/AMuscop_fnl.pdf

First, thanks for the links and information. I started reading those and finally went a bit further.
We're actually talking about the late 40s and early 50s so the copyright law was the 1909 Act by then. This act created the first compulsory mechanical licence to deal with mechanical reproductions. It was made with piano rolls in mind but audio recording fell into the category shortly after. The licence let third parties make mechanical reproductions without asking permission to the work intellectual property owner as long as there was already at least one mechanical reproduction published by him, and he was noticed that the third party was gonna make a reproduction. The third party had then to pay a fixed fee of two cents (per copy I believe) without having to ask every time a new mechanical reproduction was gonna be made. I understand that mechanical recordings are the actual recordings and the work covered by the copyright is the composition but I think only the composition could be copyrighted (and this changed in 1972). What I fail to see is if that applied to home recordings as well or just to commercial (and published, not performed) works. I should look deeper into it. My source is primarily this link I found on wikipedia:
http://www.copyright.gov/docs/regstat031104.html

 

[Oak]

Danathar[/url]":31n0bej5]I wonder, was there ever an attempt to do digital recording onto a wire? Something incredibly long lasting might be useful somehow.

I'd imagine the bitrate would be pathetically low by today's standards, but with enough wire you could record something digitally and archive it for a VERY long time.

The article said

A year after Woody’s live gig, the US National Bureau of Standards (now the National Institute of Standards and Technology) built the world’s third-ever stored-program computer (or the fourth, depending on who you believe), and this SEAC device actually used wire recorders to store digital data.

It's not necessarily as long lasting as you imagine, however. Like tape recording, it's magnetic, and thus subject to similar problems of demagnetization.


I do worry about how ephemeral digital data is, though. I wish the Internet Archive's records could periodically be etched into something very durable and copies sent to multiple locations, but that would be an enormous task given the volume of data put on the web per year, nowadays. (Some of the first few years of its captures, from the 1990s, could perhaps be done at least. Well, it would be technologically possible. The legal and cost issues could still be significant.) I notice that occasionally I hit 404s on archive.org on pages that the Wayback Machine's interface offers links to (e.g., try hitting the ▶ link, to try to go to a 15 November 2012 capture, in the timeline bar above this page), and consequently I wonder to what extent even the Internet Archive is decaying over time as system migration errors accumulate and even presumably redundant file copies are occasionally lost simultaneously. (To say nothing of what would happen if the IA stopped getting enough funding or ever suffers severe legal challenges.)

Some interesting articles on the disappearing Web:
The Internet's Dark Ages/Raiders of the Lost Web (The Atlantic)
The disappearing web: Information decay is eating away our history (Gigaom)
Learning from failure: The case of the disappearing Web site (First Monday)

 

[Oldmanalex]

I had an uncle who was working on high tech wire recorders towards the end of WWII. Then some captured German tape recorders were brought into the lab, and he was one of those who analyzed their performance. After that analysis was completed, he was no longer working on wire recorders.