Pshhhkkkkkkrrrrkakingkakingkakingtshchchchchchchchcch*ding*ding*ding*
June 3, 2012 1:38 AM Subscribe
You may remember the sounds your old dial-up modem used to make, but do you know why it was making those sounds and what was happening during each part? The Atlantic explains the Mechanics and Meaning of That Ol' Dial-Up Modem Sound.
I'm pretty sure that's the sound of a 56k digitally modulated (PCM) modem, probably of the v90-v92 / v42-v42bis variety. This is basically the last of the true Plain Old Telephone System analog/voice line modems before ISDN and ADSL took over and engineers stopped trying to cram more and more bits down the legally available center audio bandwidth of a POTS analog phone line.
If so the initialization procedure for one of these modems is actually a bit more complicated than described in the notes, but that's the general gist of it. There's some error/noise correction testing stuff going on along with data compression initialization.
Earlier modems were much "dumber" and operated with much more simple modulation schemes like frequency-shift keying.
Older modems without error correction and data compression had a lot less of that extremely complicated hissing/static noise during both handshaking and operation. Here's a modem being operated in audible mode (but without the full handshake/dialup part) at what sounds like 300 baud or below. You can actually hear both the carrier tone that maintains the "sync" between the two modems as well as the individual characters being sent/received that are modulating that carrier tone.
Here's the sound of another 300 baud modem connecting. Though I'm not sure of the protocol/type - for some reason it sounds foreign to me like it's not a true Rockwell/Hayes-compatible Smartmodem.
I used to be able to tell not just speed but other parameters by ear, like if the modem was using data compression or if it was full duplex or simplex. You could even get a feel for what BBS your friends were dialing. Not just from recognizing the DTMF touchpad tones and decoding the dialed number by ear - but also by the handshakes. Some BBSes had particular handshakes depending on the modems and software they were using, or would only connect at 2400 baud even if it was a faster modem.
See also: Modem.
posted by loquacious at 2:42 AM on June 3, 2012 [36 favorites]
If so the initialization procedure for one of these modems is actually a bit more complicated than described in the notes, but that's the general gist of it. There's some error/noise correction testing stuff going on along with data compression initialization.
Earlier modems were much "dumber" and operated with much more simple modulation schemes like frequency-shift keying.
Older modems without error correction and data compression had a lot less of that extremely complicated hissing/static noise during both handshaking and operation. Here's a modem being operated in audible mode (but without the full handshake/dialup part) at what sounds like 300 baud or below. You can actually hear both the carrier tone that maintains the "sync" between the two modems as well as the individual characters being sent/received that are modulating that carrier tone.
Here's the sound of another 300 baud modem connecting. Though I'm not sure of the protocol/type - for some reason it sounds foreign to me like it's not a true Rockwell/Hayes-compatible Smartmodem.
I used to be able to tell not just speed but other parameters by ear, like if the modem was using data compression or if it was full duplex or simplex. You could even get a feel for what BBS your friends were dialing. Not just from recognizing the DTMF touchpad tones and decoding the dialed number by ear - but also by the handshakes. Some BBSes had particular handshakes depending on the modems and software they were using, or would only connect at 2400 baud even if it was a faster modem.
See also: Modem.
posted by loquacious at 2:42 AM on June 3, 2012 [36 favorites]
What an awesome article. Thank you for posting it bjrn, not least because it contains a link to the wonderfully named Museum of Endangered Sounds.
posted by MuffinMan at 2:50 AM on June 3, 2012 [4 favorites]
posted by MuffinMan at 2:50 AM on June 3, 2012 [4 favorites]
1, tick
9, t.t.t.t.t.t.t,tick
eeeeeeeooohhhwwwwwaaa............this is a recording, ....
%*&!
posted by Mblue at 2:54 AM on June 3, 2012
9, t.t.t.t.t.t.t,tick
eeeeeeeooohhhwwwwwaaa............this is a recording, ....
%*&!
posted by Mblue at 2:54 AM on June 3, 2012
Oh, I meant to mention something about recordings and modems:
Yeah, you can replay recorded modem noises to another modem and sometimes they'll be able to decipher it and you can replay the session, at least with older/slower modem signals.
I'm not so sure about v90/v92 PCM modems, though. It may not work even with CD quality audio recordings since the sample rate will introduce aliasing and errors, and it's extremely likely it won't work at all with compressed audio like an MP3 or MPEG file.
But "modem recordings" are the basis of the data compact cassette.
posted by loquacious at 3:02 AM on June 3, 2012 [1 favorite]
Yeah, you can replay recorded modem noises to another modem and sometimes they'll be able to decipher it and you can replay the session, at least with older/slower modem signals.
I'm not so sure about v90/v92 PCM modems, though. It may not work even with CD quality audio recordings since the sample rate will introduce aliasing and errors, and it's extremely likely it won't work at all with compressed audio like an MP3 or MPEG file.
But "modem recordings" are the basis of the data compact cassette.
posted by loquacious at 3:02 AM on June 3, 2012 [1 favorite]
I'm not an expert at this, but I think the gloss provided of the handshaking sequence is kind of off, actually. In particular the connection is almost definitely using LAPM, and so working out parity and stopbits and that sort of thing isn't happening. The specifications for all these protocols are freely available from the ITU. In the v.92 specification (which is pretty dry reading) it outlines the sequence (starting at around page 45). It describes four phases:
v.90/v.92 is kind of black magic, anyway, in that it relies on the assumption that at least one end of the connection is connected to the phone system via a digital connection (like a T1 or PRI).
posted by orthicon halo at 3:03 AM on June 3, 2012 [7 favorites]
- Phase 1, network interaction;
- Phase 2, channel probing and ranging;
- Phase 3, equalizer and echo canceller training and digital impairment learning
- Phase 4, final training.
v.90/v.92 is kind of black magic, anyway, in that it relies on the assumption that at least one end of the connection is connected to the phone system via a digital connection (like a T1 or PRI).
posted by orthicon halo at 3:03 AM on June 3, 2012 [7 favorites]
Handshaking was always a bad metaphor for what was happening there. Handshakes don't have a sound. A better metaphor is bird songs. Creatures were using an analog channel to set up a digital reality; we'll fuck or we won't fuck. Even birds can do that.
posted by twoleftfeet at 3:24 AM on June 3, 2012 [2 favorites]
posted by twoleftfeet at 3:24 AM on June 3, 2012 [2 favorites]
I was lucky enough to have something of a ringside seat for this, as a friend was (possibly still is) a senior figure on the ITU committee that decided the v. standards for modems, and I followed developments closely between around 1990 and 2000 when a lot of the clever stuff was being agreed. (Being a rapporteur for an ITU committee is a good gig. You get diplomatic status, and access to the special diplomatic duty-free shops at airports. What, you didn't know about those?)
As Orthicon Halo says, the original article is simplistic to the point of being, well, wrong and Orthicon's description is accurate. My particular favourite bit of the link establishment is the boingy-boingy bit where the modem is sending frequency and amplitude modulated sweeps down the line so that working with the other end, it can work out the characteristics of the connection. Clever stuff that sounds cool, which always gets my vote.
That decade was the golden age for modem design. We came in at 1200/1200 and simple-minded modulation where you could actually hear the bits, we left at 56k and what sounded like white noise - actually, massively error-corrected and compressed frequency and amplitude modulated bitstream, with the same techniques as used to squeeze the last drops out of the radio link with the Voyagers. We started with concepts that would have been recognisable to wartime radio teletype engineers, and finished with digital signal processing indistinguishable from magic (well, unless you understand Claude Shannon. Which I can only pretend to do).
Learned a lot from that. I also learned a lot from the modem industry politics of the time. Once were giants - Hayes, US Robotics, Rockwell - and their fortunes waxed fat with the discovery of the Internet by the common people. This coincided, by no means coincidentally, with the rapid surge in modem capabilities brought about through Moore's Law getting to the point that you could have cheap DSP, combined with a surge in general computing capabilities brought about through Moore's Law getting to the point that you could have reasonably sophisticated interfaces.
In centuries to come, we will look back at that decade as something extraordinary.
But corporations have no sense of occasion and no grace whatsoever. The modem companies got greedy and refused to play the standardisation game that had got them so much. Instead of there being one 56k standard, there were two, with the two consortia staring each other out in a winner-takes-all pissing competition. Alas for the poor dears, commoditisation was also happening, but instead of concentrating on crossing that bridge into broadband land, the modem companies delighted in annoying their customers (ISPs had to have two different banks of modems with different dial-up numbers, or just plumped for one which meant punters had to choose their modems according to which ISP they were going to use) and those members of the ITU V groups who just wanted to get a proper standard. It also annoyed those of us who were watching the scene, as there was a very great deal of BS from the modem companies as they sought to explain their marketing-led cupidity as being in some way a product of unique engineering prowess.
Of course, there was a final standard in the end, but by then they'd screwed the pooch. You may have noticed that the modem companies did not survive, and not for there being a lack of modems in the world; ADSL chips are basically hundreds of modems in parallel (I lie, but that's the general idea) and every digital phone relies entirely on modem technology - electromagnetic reality being analogue and all (I lie again, but that's the general idea).
So listen carefully to that song of the modem. It holds within its microcosm the story of how technology and humans goad each other into the future, from a mythical time when the digital first truly awoke.
Boingy-boingy hisssss....
posted by Devonian at 4:56 AM on June 3, 2012 [105 favorites]
As Orthicon Halo says, the original article is simplistic to the point of being, well, wrong and Orthicon's description is accurate. My particular favourite bit of the link establishment is the boingy-boingy bit where the modem is sending frequency and amplitude modulated sweeps down the line so that working with the other end, it can work out the characteristics of the connection. Clever stuff that sounds cool, which always gets my vote.
That decade was the golden age for modem design. We came in at 1200/1200 and simple-minded modulation where you could actually hear the bits, we left at 56k and what sounded like white noise - actually, massively error-corrected and compressed frequency and amplitude modulated bitstream, with the same techniques as used to squeeze the last drops out of the radio link with the Voyagers. We started with concepts that would have been recognisable to wartime radio teletype engineers, and finished with digital signal processing indistinguishable from magic (well, unless you understand Claude Shannon. Which I can only pretend to do).
Learned a lot from that. I also learned a lot from the modem industry politics of the time. Once were giants - Hayes, US Robotics, Rockwell - and their fortunes waxed fat with the discovery of the Internet by the common people. This coincided, by no means coincidentally, with the rapid surge in modem capabilities brought about through Moore's Law getting to the point that you could have cheap DSP, combined with a surge in general computing capabilities brought about through Moore's Law getting to the point that you could have reasonably sophisticated interfaces.
In centuries to come, we will look back at that decade as something extraordinary.
But corporations have no sense of occasion and no grace whatsoever. The modem companies got greedy and refused to play the standardisation game that had got them so much. Instead of there being one 56k standard, there were two, with the two consortia staring each other out in a winner-takes-all pissing competition. Alas for the poor dears, commoditisation was also happening, but instead of concentrating on crossing that bridge into broadband land, the modem companies delighted in annoying their customers (ISPs had to have two different banks of modems with different dial-up numbers, or just plumped for one which meant punters had to choose their modems according to which ISP they were going to use) and those members of the ITU V groups who just wanted to get a proper standard. It also annoyed those of us who were watching the scene, as there was a very great deal of BS from the modem companies as they sought to explain their marketing-led cupidity as being in some way a product of unique engineering prowess.
Of course, there was a final standard in the end, but by then they'd screwed the pooch. You may have noticed that the modem companies did not survive, and not for there being a lack of modems in the world; ADSL chips are basically hundreds of modems in parallel (I lie, but that's the general idea) and every digital phone relies entirely on modem technology - electromagnetic reality being analogue and all (I lie again, but that's the general idea).
So listen carefully to that song of the modem. It holds within its microcosm the story of how technology and humans goad each other into the future, from a mythical time when the digital first truly awoke.
Boingy-boingy hisssss....
posted by Devonian at 4:56 AM on June 3, 2012 [105 favorites]
Though far more common back in to 90's and early aughts, every once in awhile you will still hear snippets of this sound used as a background noise in local advertising as an indication of "technology" or "find us on the web!". Usually car dealership ads.
posted by Thorzdad at 4:57 AM on June 3, 2012 [2 favorites]
posted by Thorzdad at 4:57 AM on June 3, 2012 [2 favorites]
I had just about every generation of modem, starting with an acoustically coupled 300 baud box that did hardly any handshaking at all; I went on to own 1200 baud, 2400 baud, 9600 baud, 14.4K, and finally 56K modems, and with each upgrade the handshake started the same but got longer as additional capabilities were probed and tested. Much of the complexity of the handshake sequence is due to downward compatibility; that 56K modem was meant to be able to call a BBS with an ancient 300 baud set and make connection. The 56K handshake would have been a lot simpler if it only ever needed to connect to other 56K modems (though it would still have to negotiate line quality).
posted by localroger at 4:59 AM on June 3, 2012 [1 favorite]
posted by localroger at 4:59 AM on June 3, 2012 [1 favorite]
it contains a link to the wonderfully named Museum of Endangered Sounds
I want to hug that website, and make it some hot cocoa.
posted by louche mustachio at 5:00 AM on June 3, 2012 [5 favorites]
I want to hug that website, and make it some hot cocoa.
posted by louche mustachio at 5:00 AM on June 3, 2012 [5 favorites]
Yeah, this explanation is rather questionable. Most of what the handshake is really doing is about the two modems trying to characterize the properties of the line (ranging). Both ends send reference signals for the other end to record and analyze. This lets the modems work out how much delay is on the line and what frequencies the line is attenuating, perhaps because of the presence of things like load coils that were used in the olden days. Once the channel is characterized, the modems negotiate a modulation scheme and symbol rate -- trying to go too fast on a line with heavy attenuation in certain bands would be counterproductive. Being able to adjust based on the specifics of the line conditions was one of the things that allowed speeds to progress to higher speeds.
It's kind of like when you evaluate a lens for a SLR camera -- you look through it and see where the image is sharp, where there's chromatic aberration, etc. If the corners or edges are a bit blurry then you change the framing to avoid using them, or if the whole picture is off then you just plain don't use it for things that require sharpness. The modem is doing a similar thing trying to judge the quality of the line.
posted by Rhomboid at 5:00 AM on June 3, 2012 [1 favorite]
It's kind of like when you evaluate a lens for a SLR camera -- you look through it and see where the image is sharp, where there's chromatic aberration, etc. If the corners or edges are a bit blurry then you change the framing to avoid using them, or if the whole picture is off then you just plain don't use it for things that require sharpness. The modem is doing a similar thing trying to judge the quality of the line.
posted by Rhomboid at 5:00 AM on June 3, 2012 [1 favorite]
There are lots of people who remember the modem sounds and enjoy imitating the handshake. (previously on metafilter)
I've been doing lots of PSK31 communication recently, which is only 31.5 baud variable width varicoding, but also only uses about 31.5 Hz of bandwidth. One very neat thing with SDR waterfall displays is that bandwidth is a literal width of the communication on the screen. Some communication modes are very low "bandwidth" in terms of data transmitted, but very high bandwidth in terms of the space that they consume on the RF channel (I'm looking at you, FM voice).
PSK and CW (Morse code) are fascinating in how little band they use -- 31 Hz and 150 Hz -- yet allow for real discussion to occur. Since PSK is a SSB encoding, multiple transmitters can even be on the same carrier frequency and not overlap if their baseband encoding frequencies differ. The overlap of various faint PSK31 chats fits nicely into a 3 KHz audio passband and makes for pleasant listening, even if it isn't human parsable like CW.
posted by autopilot at 5:55 AM on June 3, 2012
I've been doing lots of PSK31 communication recently, which is only 31.5 baud variable width varicoding, but also only uses about 31.5 Hz of bandwidth. One very neat thing with SDR waterfall displays is that bandwidth is a literal width of the communication on the screen. Some communication modes are very low "bandwidth" in terms of data transmitted, but very high bandwidth in terms of the space that they consume on the RF channel (I'm looking at you, FM voice).
PSK and CW (Morse code) are fascinating in how little band they use -- 31 Hz and 150 Hz -- yet allow for real discussion to occur. Since PSK is a SSB encoding, multiple transmitters can even be on the same carrier frequency and not overlap if their baseband encoding frequencies differ. The overlap of various faint PSK31 chats fits nicely into a 3 KHz audio passband and makes for pleasant listening, even if it isn't human parsable like CW.
posted by autopilot at 5:55 AM on June 3, 2012
Sounds like the author would have been better off writing an Askme!
posted by orme at 6:12 AM on June 3, 2012 [1 favorite]
posted by orme at 6:12 AM on June 3, 2012 [1 favorite]
Yeah, you can replay recorded modem noises to another modem and sometimes they'll be able to decipher it
Information Society had put 300baud recordings of data at the end of their CDs more than once, the first was a text file of a story and the other was a bunch of text and a URL. IIRC, the first one was given to the studio with instructions on how it should be put on the disc, but a "helpful" engineer faded out the static at the end, rather than the abrupt cutoff, for aesthetic purposes by effectively damaging the data stream.
posted by AzraelBrown at 6:13 AM on June 3, 2012 [3 favorites]
Information Society had put 300baud recordings of data at the end of their CDs more than once, the first was a text file of a story and the other was a bunch of text and a URL. IIRC, the first one was given to the studio with instructions on how it should be put on the disc, but a "helpful" engineer faded out the static at the end, rather than the abrupt cutoff, for aesthetic purposes by effectively damaging the data stream.
posted by AzraelBrown at 6:13 AM on June 3, 2012 [3 favorites]
it contains a link to the wonderfully named Museum of Endangered Sounds
That's a really crappy sounding rotary phone they've got. It was a nice sound, solid and reassuring. But maybe that sound is gone forever now. Maybe there aren't any rotary phones left in good enough shape to get a recording of what they really sounded like.
posted by DarkForest at 6:15 AM on June 3, 2012
That's a really crappy sounding rotary phone they've got. It was a nice sound, solid and reassuring. But maybe that sound is gone forever now. Maybe there aren't any rotary phones left in good enough shape to get a recording of what they really sounded like.
posted by DarkForest at 6:15 AM on June 3, 2012
I think the phone sound matches up fine with the picture, which seems to be a WE-302, from the 30's. There are certainly more 'modern' ones around, from the 60's, which sound smoother. I thought I had one, but I can't find it now. I'm pretty sure my sister has one in her kitchen as the main house phone.
The endangered sound I loved hearing a little while back is the sound of a tone arm softly landing on an LP, heard through headphones- that soft scratching. Takes me back...
posted by MtDewd at 6:41 AM on June 3, 2012 [2 favorites]
The endangered sound I loved hearing a little while back is the sound of a tone arm softly landing on an LP, heard through headphones- that soft scratching. Takes me back...
posted by MtDewd at 6:41 AM on June 3, 2012 [2 favorites]
Previously on Metafilter: a 300 baud modem in a wooden box from 1964. There's an excellent DEFCON presentation by the owner of the modem, and at the end, he uses it to decode the Information Society track mentioned above by AzraelBrown.
posted by zsazsa at 7:03 AM on June 3, 2012 [1 favorite]
posted by zsazsa at 7:03 AM on June 3, 2012 [1 favorite]
Maybe there aren't any rotary phones left in good enough shape to get a recording of what they really sounded like.
My mother has a rotary Bell Model 500 that still works perfectly. She prefers to use a cordless pushbutton phone because, usually when you dial a phone number, a voice mail system answers. When the electricity goes out, the POTS phone can still dial out.
posted by Multicellular Exothermic at 7:03 AM on June 3, 2012
My mother has a rotary Bell Model 500 that still works perfectly. She prefers to use a cordless pushbutton phone because, usually when you dial a phone number, a voice mail system answers. When the electricity goes out, the POTS phone can still dial out.
posted by Multicellular Exothermic at 7:03 AM on June 3, 2012
I still hear the modem sound all the time; in construction, a good percent of business is still done by fax. It's still faster to drop a proposal (or request for same) into the fax machine and type in the number than it is to scan in a hand-written proposal or print to PDF, find the recipient in your contacts, and email it to them. And you don't want to send something plaintext because there are tossers out there that will change your text and swear that's what you sent them in your original bid.
The sounds I miss? the "EhhhhhhhhhEh_chnk_Eh_chnk_Eh_chnk_Eh_chnk_" of my TI99/4a recording BASIC programs onto cassette tape. The "Ffff...tkt...tkt...tkt" of auto-reverse on a cassette. And from my own growing up, the slimy, gritty, rotate-ey noise that lifters in a 350 Chevy make when you push a wrench to cover your tragus (ear flap) and put the other end of the wrench on just the right place on the valve cover to check the valvetrain.
posted by notsnot at 7:05 AM on June 3, 2012 [2 favorites]
The sounds I miss? the "EhhhhhhhhhEh_chnk_Eh_chnk_Eh_chnk_Eh_chnk_" of my TI99/4a recording BASIC programs onto cassette tape. The "Ffff...tkt...tkt...tkt" of auto-reverse on a cassette. And from my own growing up, the slimy, gritty, rotate-ey noise that lifters in a 350 Chevy make when you push a wrench to cover your tragus (ear flap) and put the other end of the wrench on just the right place on the valve cover to check the valvetrain.
posted by notsnot at 7:05 AM on June 3, 2012 [2 favorites]
Is there no love for the Telebit 19.2K ?
posted by Obscure Reference at 7:19 AM on June 3, 2012
posted by Obscure Reference at 7:19 AM on June 3, 2012
When the electricity goes out, the POTS phone can still dial out.
And if the rotary dial breaks, you can just tap the number out on the hangup button!
posted by Meatbomb at 7:40 AM on June 3, 2012 [2 favorites]
And if the rotary dial breaks, you can just tap the number out on the hangup button!
posted by Meatbomb at 7:40 AM on June 3, 2012 [2 favorites]
And if the rotary dial breaks, you can just tap the number out on the hangup button!
Yeah, you could use that trick to make calls on phones that were supposed to be 'receive-only'. AFAIK, there was no such thing for real, they'd just disable the keypad or dial. You could just flash the hangup to simulate a dial, and call anyone you wanted.
They used to charge extra for touch-tone service. Like, for real. It cost more money to dial more efficiently, taking fewer phone company resources.
And wow did the phone companies hate modems. Hated them intensely. If something like modems were to come out now, with the absolute capture of government by corporations, there's no way a dial-up Internet could happen.
posted by Malor at 8:02 AM on June 3, 2012 [6 favorites]
Yeah, you could use that trick to make calls on phones that were supposed to be 'receive-only'. AFAIK, there was no such thing for real, they'd just disable the keypad or dial. You could just flash the hangup to simulate a dial, and call anyone you wanted.
They used to charge extra for touch-tone service. Like, for real. It cost more money to dial more efficiently, taking fewer phone company resources.
And wow did the phone companies hate modems. Hated them intensely. If something like modems were to come out now, with the absolute capture of government by corporations, there's no way a dial-up Internet could happen.
posted by Malor at 8:02 AM on June 3, 2012 [6 favorites]
And wow did the phone companies hate modems
I remember when you could not own a phone. You leased your phone from the phone company, and you paid extra to have more than one. The entire network, including every device connected to it at the far end and every piece of wire, including those in your house, was considered property of the phone company.
This is the reason the phone companies hated modems; they were uncontrolled "foreign attachments" hanging on THEIR PRECIOUS NETWORK. But they hated all uncontrolled things hanging on their network equally; you could not buy phones from anyone else. This was the case until 1984 when the monopoly was broken and part of that was making it possible to own your own phone and for third parties to make and sell them.
Even after that the phone companies tried to charge more if you had more than one phone while whining about extra ringer load and whatnot from too many uncontrolled phones, but as mechanical ringers became replaced by much less demanding piezo sounders that complaint got pretty hollow.
posted by localroger at 8:14 AM on June 3, 2012 [3 favorites]
I remember when you could not own a phone. You leased your phone from the phone company, and you paid extra to have more than one. The entire network, including every device connected to it at the far end and every piece of wire, including those in your house, was considered property of the phone company.
This is the reason the phone companies hated modems; they were uncontrolled "foreign attachments" hanging on THEIR PRECIOUS NETWORK. But they hated all uncontrolled things hanging on their network equally; you could not buy phones from anyone else. This was the case until 1984 when the monopoly was broken and part of that was making it possible to own your own phone and for third parties to make and sell them.
Even after that the phone companies tried to charge more if you had more than one phone while whining about extra ringer load and whatnot from too many uncontrolled phones, but as mechanical ringers became replaced by much less demanding piezo sounders that complaint got pretty hollow.
posted by localroger at 8:14 AM on June 3, 2012 [3 favorites]
Gah, not 1984, it was Carterfone in 1968 that made third party handsets and such legal. And get off my lawn.
posted by localroger at 8:19 AM on June 3, 2012 [1 favorite]
posted by localroger at 8:19 AM on June 3, 2012 [1 favorite]
The reason phone companies hated modems wasn't because they were uncontrolled, although the "it's dangerous to plug anything into our incredibly important network" line was used for a long time to justify things (the fact that telco approvals have largely gone away with no noticeable detriment is one sign that it was an excuse, not a reason).
The real reason was that phone companies often also had a monopoly on data services to businesses, which they charged an absolute fortune for. Anything that threatened this, whether internal or external to the telco, was to be resisted at all costs. That's the major reason ISDN was so half-heartedly promoted in the UK; the data services part of BT was older, bigger and well-established, and made damn sure that ISDN was no threat to it by suffocating it in its cradle.
Game of Thrones has nothing on the Game of Telephones for backstabbing, treachery and sticking it to the poor bloody peasants.
posted by Devonian at 9:03 AM on June 3, 2012 [7 favorites]
The real reason was that phone companies often also had a monopoly on data services to businesses, which they charged an absolute fortune for. Anything that threatened this, whether internal or external to the telco, was to be resisted at all costs. That's the major reason ISDN was so half-heartedly promoted in the UK; the data services part of BT was older, bigger and well-established, and made damn sure that ISDN was no threat to it by suffocating it in its cradle.
Game of Thrones has nothing on the Game of Telephones for backstabbing, treachery and sticking it to the poor bloody peasants.
posted by Devonian at 9:03 AM on June 3, 2012 [7 favorites]
We had really crappy dial-up, and I could always tell from the modem whether or not it had connected. My boyfriend at the time didn't believe me. I thought it was obvious, but then he also didn't believe that I could hear when the car engine was low on oil. Until it threw a rod.
posted by oneirodynia at 9:16 AM on June 3, 2012 [5 favorites]
posted by oneirodynia at 9:16 AM on June 3, 2012 [5 favorites]
My folks still have one of those Model 500 rotaries as the main phone. I think they initially kept it for so long because they weren't going to give Pac Bell no $1.95 a month extra for that touch tone service. Mind you, this is a woman whose first telephone number was '3' — as in the call was for them if the phone rang three times.
posted by benito.strauss at 10:25 AM on June 3, 2012 [1 favorite]
posted by benito.strauss at 10:25 AM on June 3, 2012 [1 favorite]
Though far more common back in to 90's and early aughts, every once in awhile you will still hear snippets of this sound used as a background noise in local advertising as an indication of "technology" or "find us on the web!". Usually car dealership ads.
The last time I heard the familiar noise was about 2 weeks ago. I needed some cash, so I went to a standalone ATM.
*card swipe*
*PIN*
"yes, $20"
"$3 fee?! wtf. *yes*"
*boing CHHHHHHHH*
posted by ninjew at 10:30 AM on June 3, 2012
The last time I heard the familiar noise was about 2 weeks ago. I needed some cash, so I went to a standalone ATM.
*card swipe*
*PIN*
"yes, $20"
"$3 fee?! wtf. *yes*"
*boing CHHHHHHHH*
posted by ninjew at 10:30 AM on June 3, 2012
Awesome. The Atlantic post is full of bunk, but the comments in this thread are gold. I am struck with sudden nostalgia (okay, sort of) for the days when I'd dial in and, based on the length and type of the noises my shiny 56k modem made, would know before Windows reported the connection as live how fast it was going to be. Fallback to 33.6k or 28k was always so upsetting.
And then I got a cable modem before all of my IRC buddies did, and would (as mischievous 13-year-olds are wont to do) drop in to a channel and blurt out "+++ATH0", then giggle my ass off 2 minutes later when they all got dropped by the server.
posted by spitefulcrow at 10:31 AM on June 3, 2012
And then I got a cable modem before all of my IRC buddies did, and would (as mischievous 13-year-olds are wont to do) drop in to a channel and blurt out "+++ATH0", then giggle my ass off 2 minutes later when they all got dropped by the server.
posted by spitefulcrow at 10:31 AM on June 3, 2012
Ah yes, that first Radio Shack 300 baud modem that had no dialer and a single call/answer switch and big connect button. Probably the first electronic gizmo I saved my allowance and bought myself. Living with the constant fear of siblings picking up the phone mid connection or mother finding out that somebody tried to call for hours and got the busy signal. I had to learn enough Amiga Basic to patch in a download function into their terminal program so I could download a real terminal emulator, and then still had to fixup downloads because they were padded out to a blocksize and the Amiga loader would freak out over the extra bytes at the end. Almost as bad as the previous cassette recorder monster on the Apple, with volume dials taped in place or endless fiddling and retries to get the playback to match the recording volume. Couldn't afford one of the dedicated recorders that had fixed gain. Then after uncountable plays of Star Trek the tape wears out. Youngsters are so lucky today.
posted by zengargoyle at 10:42 AM on June 3, 2012
posted by zengargoyle at 10:42 AM on June 3, 2012
The first modems I ever saw were the size of a VCR (in 1973, before VCRs). So old I can't find any pictures online. I don't recall if they were 300 baud or 1200, but they didn't make these noises because they were on leased lines, always connected, and didn't need to dial out; didn't need to have a speaker.
The first ones I got trained on were the IBM 3863/4/5. Most of the TP stuff I worked on then was bi-sync and 2400-9600 baud, also leased line. The idea that you can get 56Kbps asynchronous on a public switched telephone network still floors me.
posted by MtDewd at 11:05 AM on June 3, 2012
The first ones I got trained on were the IBM 3863/4/5. Most of the TP stuff I worked on then was bi-sync and 2400-9600 baud, also leased line. The idea that you can get 56Kbps asynchronous on a public switched telephone network still floors me.
posted by MtDewd at 11:05 AM on June 3, 2012
I remember building my first 300 baud modem. I used a couple of RadioShack style 2 1/4 inch paper cone speakers to make an acoustic coupler. You use one speaker as the transmitter and the second speaker in reverse as a microphone receiver (yeah, that works). The speakers are just about the right size to cover the handset for a Bell 500 telephone. I fashioned the acoustic cups from a couple of strips cut from a foam plastic table placemat. I see that you can still buy National Semiconductor MM74HC943 300 baud modem chips on the gray market in Asia.
posted by JackFlash at 11:06 AM on June 3, 2012 [2 favorites]
posted by JackFlash at 11:06 AM on June 3, 2012 [2 favorites]
Hey, author here. I think there are a couple of misunderstandings about what I was trying to do here, Orthicon and Devonian. We had to use an actual incarnation of a modem sound, but we were trying to get at something like the general case. I know there are a lot of details that get obscured when you're trying to abstract out of a specific modem sound to a general case, but I still think that's more valuable to readers than giving them the *exact* v.92 specifications for a device that very few of them are ever going to use again in their lives.
So that's what we did. We made the graphic and I ran it by a bunch of people who were all like, "Yes, that is, in rough terms, what's happening." And then of course you post it and suddenly you're just horribly wrong about everything according to some subset of people who really *live* this stuff. I think it's sort of a category mismatch between highly technical readers and everyone else.
Here's my analogy: If you wanted to explain a car to someone who didn't understand anything about cars and you point to some new Cadillac on the street, you might explain things about the way the thing worked that weren't exactly perfect for the 2012 Cadillac but that are true for cars generally. I should have made it clearer in my post. I think it's implicit in that we didn't go into the specific type of modem and sound that was embedded, but I should have made it explicit.
AND, of course, if you do want to get in touch with me to create a highly detailed, ultra specific explanation of the sound, I'm all ears and would be happy to run that as well. I'm at alexis.madrigal[at]gmail.com.
posted by alexismadrigal at 11:09 AM on June 3, 2012 [13 favorites]
So that's what we did. We made the graphic and I ran it by a bunch of people who were all like, "Yes, that is, in rough terms, what's happening." And then of course you post it and suddenly you're just horribly wrong about everything according to some subset of people who really *live* this stuff. I think it's sort of a category mismatch between highly technical readers and everyone else.
Here's my analogy: If you wanted to explain a car to someone who didn't understand anything about cars and you point to some new Cadillac on the street, you might explain things about the way the thing worked that weren't exactly perfect for the 2012 Cadillac but that are true for cars generally. I should have made it clearer in my post. I think it's implicit in that we didn't go into the specific type of modem and sound that was embedded, but I should have made it explicit.
AND, of course, if you do want to get in touch with me to create a highly detailed, ultra specific explanation of the sound, I'm all ears and would be happy to run that as well. I'm at alexis.madrigal[at]gmail.com.
posted by alexismadrigal at 11:09 AM on June 3, 2012 [13 favorites]
It's important to point out that two factors doomed hardware modem design. The first has already come up -- the growth of high speed links like ADSL and Cable -- but I'm unconvinced that the trickery required to operate over those relatively high bandwidth links was the same sort needed to function over extended telephone networks. A cruise ship and a Vespa both have engines, but you know, they're solving different problems.
The other nail in the coffin, however, was software modems. As neat as a hardware device is, everything a modem does can be approximated with a sound card hooked up to a phone line with a suitable software driver sending and receiving the appropriate noises. So-called "Winmodems" weren't perfect, but they were good enough. And man, were they cheap.
posted by effugas at 11:20 AM on June 3, 2012
The other nail in the coffin, however, was software modems. As neat as a hardware device is, everything a modem does can be approximated with a sound card hooked up to a phone line with a suitable software driver sending and receiving the appropriate noises. So-called "Winmodems" weren't perfect, but they were good enough. And man, were they cheap.
posted by effugas at 11:20 AM on June 3, 2012
Fallback to 33.6k or 28k was always so upsetting.
1998 me is jealous. Anytime my modem connected above 26.4k was a guarantee that it would disconnect within 5 minutes.
posted by Mister Fabulous at 11:46 AM on June 3, 2012
1998 me is jealous. Anytime my modem connected above 26.4k was a guarantee that it would disconnect within 5 minutes.
posted by Mister Fabulous at 11:46 AM on June 3, 2012
Hi Alexis, welcome to MetaFilter!
We pick a good nit here about an awful lot of things but it's generally well-intentioned and fun. Of course you raise a good point about simplifying a complex process for a general audience: the audience here is -- and I'm probably generalizing inappropriately here -- a lot more willing to applaud it when people who know something specialized really get into the details than the audience you write for.
I hope you'll have a look around the place and maybe even decide to stick around.
posted by gauche at 12:28 PM on June 3, 2012
We pick a good nit here about an awful lot of things but it's generally well-intentioned and fun. Of course you raise a good point about simplifying a complex process for a general audience: the audience here is -- and I'm probably generalizing inappropriately here -- a lot more willing to applaud it when people who know something specialized really get into the details than the audience you write for.
I hope you'll have a look around the place and maybe even decide to stick around.
posted by gauche at 12:28 PM on June 3, 2012
Devonian: "The modem companies got greedy and refused to play the standardisation game that had got them so much. Instead of there being one 56k standard, there were two, with the two consortia staring each other out in a winner-takes-all pissing competition."
I seem to remember most x2 modems were field upgradable to V.90. Was that not the case for the K56Flex modems?
You know, it still amazes me that they were able to make modems good enough that whether or not the host end was on a robbed bit signalling ISDN line mattered or not. They needed to know if that one bit out of 8 in every 6th frame was actually carrying voice data or not. The fact that I could ever even once have a modem connect and successfully use over 80% of the phone company's digital rate is freakin' amazing.
Think about that for a second. The phone company is transmitting all this within its network in all digital form. Only once it reaches your particular copper pair does it get turned into analog form. Yet somehow that modem can screech and beep and boop with accurate enough timing to avoid having most of the bits it's trying to send get mangled on the way across the network and reach 80% of the total available capacity of the channel.
It's largely a black art to me, but boy howdy was it interesting. The really interesting thing I just remembered is how similar to DSL the later standards actually are. On a good modem, you can connect to the far end and then get an amazing amount of diagnostic info. Thinking back, I remember it looking suspiciously like some of the DSL modem diagnostic info I've seen. I don't think it used a technique like DMT, but I sure remember it having similar bit binning. I wish I had a phone line so I could play with my Courier (the 33.6 version field upgraded to x2 and then v.90, all for no charge..good thing since I remember the modem being $200 or more) and find out.
posted by wierdo at 12:38 PM on June 3, 2012 [1 favorite]
I seem to remember most x2 modems were field upgradable to V.90. Was that not the case for the K56Flex modems?
You know, it still amazes me that they were able to make modems good enough that whether or not the host end was on a robbed bit signalling ISDN line mattered or not. They needed to know if that one bit out of 8 in every 6th frame was actually carrying voice data or not. The fact that I could ever even once have a modem connect and successfully use over 80% of the phone company's digital rate is freakin' amazing.
Think about that for a second. The phone company is transmitting all this within its network in all digital form. Only once it reaches your particular copper pair does it get turned into analog form. Yet somehow that modem can screech and beep and boop with accurate enough timing to avoid having most of the bits it's trying to send get mangled on the way across the network and reach 80% of the total available capacity of the channel.
It's largely a black art to me, but boy howdy was it interesting. The really interesting thing I just remembered is how similar to DSL the later standards actually are. On a good modem, you can connect to the far end and then get an amazing amount of diagnostic info. Thinking back, I remember it looking suspiciously like some of the DSL modem diagnostic info I've seen. I don't think it used a technique like DMT, but I sure remember it having similar bit binning. I wish I had a phone line so I could play with my Courier (the 33.6 version field upgraded to x2 and then v.90, all for no charge..good thing since I remember the modem being $200 or more) and find out.
posted by wierdo at 12:38 PM on June 3, 2012 [1 favorite]
Not to mention analog phone lines being upper-bandpassed at around 3.2KHz, so how do you get the quantity of information to exceed the carrier frequency?
Lots of technology.
posted by MtDewd at 1:20 PM on June 3, 2012
Lots of technology.
posted by MtDewd at 1:20 PM on June 3, 2012
I always thought of it as a computer's love song.
posted by Sebmojo at 1:21 PM on June 3, 2012 [3 favorites]
posted by Sebmojo at 1:21 PM on June 3, 2012 [3 favorites]
I've got a miniature milling machine out in the garage with three medium size stepper motors on it. Any one of them in motion has a discernible pitch and rhythm depending on what speed its running at -- with all three of them going it sounds almost exactly like a 300 baud modem. Kind of homey actually.
posted by Tell Me No Lies at 2:28 PM on June 3, 2012
posted by Tell Me No Lies at 2:28 PM on June 3, 2012
I really liked my old 2400 that ran off of line current. It was half the size of a pack of cards - really, the size was limited by the DB-25 serial port - and I had it duct-taped to the back of my Mac Portable, the mini-DIN-8 adapter cable looped and zip-tied. I could log on to the local dial-up ISP from pretty much everywhere there was a (non-PBX) phone jack. Brand new, it was cheaper than a 2nd hand Mac Portable modem card.
posted by Slap*Happy at 9:04 AM on June 4, 2012
posted by Slap*Happy at 9:04 AM on June 4, 2012
Hey gauche: I mean, as a reader, of course I love the MF community's ability to pick the best nit around. As an author, you're like, "But.. but.. but..."
Even when I disagree or have an explanation, though, I still feel like it's worth it to really try to hear the criticism. Keeps you from ever feeling too confident about knowing what is going on in the world.
posted by alexismadrigal at 11:07 AM on June 4, 2012 [1 favorite]
Even when I disagree or have an explanation, though, I still feel like it's worth it to really try to hear the criticism. Keeps you from ever feeling too confident about knowing what is going on in the world.
posted by alexismadrigal at 11:07 AM on June 4, 2012 [1 favorite]
> Instead of there being one 56k standard, there were two
This also happened earlier at 9600 bps with the Telebit Trailblazer, Hayes Express 96, US Robotics HST. Eventually there was a standard, V.32.
In 1992, my office had a bunch of 2400 bps modems and one USR Courier HST. USR offered Couriers in HST, V.32 (when it became available) or both to BSS sysops at about half price to push their standard.
posted by morganw at 3:53 PM on June 4, 2012
This also happened earlier at 9600 bps with the Telebit Trailblazer, Hayes Express 96, US Robotics HST. Eventually there was a standard, V.32.
In 1992, my office had a bunch of 2400 bps modems and one USR Courier HST. USR offered Couriers in HST, V.32 (when it became available) or both to BSS sysops at about half price to push their standard.
posted by morganw at 3:53 PM on June 4, 2012
> In 1992, my office had a bunch of 2400 bps modems and one USR Courier HST.
Shit, that can't be right. 28.8 must have been the new hotness at the time, not 9600.
posted by morganw at 3:56 PM on June 4, 2012
Shit, that can't be right. 28.8 must have been the new hotness at the time, not 9600.
posted by morganw at 3:56 PM on June 4, 2012
Uh, no, that's exactly right. 28.8 in 1992 would have been alien wonder technology. Here's a press release from November 1993 where USR announces their first high end products to support V.FC, the proprietary pre-standard 28.8 modulation. The final V.34 spec that "officially" launched 28.8 wasn't ready until September 1994.
posted by Rhomboid at 5:38 PM on June 4, 2012
posted by Rhomboid at 5:38 PM on June 4, 2012
(although by 92 you probably would have had v.32bis 14.4k symmetric or a related proprietary model like the Zyxel that could do an asymmetric 16.8.)
posted by Rhomboid at 5:41 PM on June 4, 2012
posted by Rhomboid at 5:41 PM on June 4, 2012
bjrn: "You may remember the sounds your old dial-up modem used to make"
It must be nice to live somewhere where you can refer to dialup like you're talking about passenger pigeons, but on my folks' road in coastal Maine, everyone is still on dialup. Fairpoint is outright refusing to provide basic DSL, and we're not exactly in the boondocks.
Considering the Web is optimised for high-speed now, even loading Google is a chore.
posted by dunkadunc at 11:48 PM on June 4, 2012
It must be nice to live somewhere where you can refer to dialup like you're talking about passenger pigeons, but on my folks' road in coastal Maine, everyone is still on dialup. Fairpoint is outright refusing to provide basic DSL, and we're not exactly in the boondocks.
Considering the Web is optimised for high-speed now, even loading Google is a chore.
posted by dunkadunc at 11:48 PM on June 4, 2012
Another old-timer here who dates back to the 300 baud era.
localroger's correction was interesting to me because I'd have thought the same as he — I certainly don't recall third-party phones sold much, if at all, before the deregulation and break-up and I'd thought that the phone company's ability to disallow you to connect your own phones to their lines only ended at that time. Huh. I suppose that what happened is that the anti-trust breakup and deregulation greatly accelerated or began a trend that was possibly but mostly unrealized up to that point.
Anyway, it's pretty amazing how much things have changed in my adult lifetime online. From dialing up to DJNR and CompuServe and Genie and BBS's in the 80s, to Prodigy and PCL/MA/AOL and ISPs in the 90s, to my first broadband (cable modem) connection in 1998.
And here I am, just moved to Kansas City four days ago and hotly anticipating the eventual rollout of Google's gigabit Internet access initiate. I'd not paid any attention to it since they announced they'd chosen KC, KS and was disappointed about it because my family lives in MO. And no one ever bothered to mention to me that, apparently, a month later the outcry led Google to expand it to Kansas City, Missouri, too. Woohoo!
Meanwhile, though, I'm annoyed at my 3Mbps connection when I've been accustomed to a 20Mbps one for the last two years. Hell, I had 5Mbps upstream. For the next couple of months, I'm staying with my mom and her husband and I — admittedly somewhat ungratefully — groused to her about how my ABQ connection was seven times faster and the one Google is going to make available here is 333 times faster. Incidentally, it will be symmetric, which is actually very smart of them since I'm certain that will be important for encouraging grass-roots innovation in ultra-high speed internet apps. When anyone in their garage can serve high-speed content and be mostly limited by their processing, not network, bandwidth...that is a different ecosystem than the one we're used to now. It's more like, as it happens, the one we lived in during the BBS era.
posted by Ivan Fyodorovich at 1:34 AM on June 5, 2012
localroger's correction was interesting to me because I'd have thought the same as he — I certainly don't recall third-party phones sold much, if at all, before the deregulation and break-up and I'd thought that the phone company's ability to disallow you to connect your own phones to their lines only ended at that time. Huh. I suppose that what happened is that the anti-trust breakup and deregulation greatly accelerated or began a trend that was possibly but mostly unrealized up to that point.
Anyway, it's pretty amazing how much things have changed in my adult lifetime online. From dialing up to DJNR and CompuServe and Genie and BBS's in the 80s, to Prodigy and PCL/MA/AOL and ISPs in the 90s, to my first broadband (cable modem) connection in 1998.
And here I am, just moved to Kansas City four days ago and hotly anticipating the eventual rollout of Google's gigabit Internet access initiate. I'd not paid any attention to it since they announced they'd chosen KC, KS and was disappointed about it because my family lives in MO. And no one ever bothered to mention to me that, apparently, a month later the outcry led Google to expand it to Kansas City, Missouri, too. Woohoo!
Meanwhile, though, I'm annoyed at my 3Mbps connection when I've been accustomed to a 20Mbps one for the last two years. Hell, I had 5Mbps upstream. For the next couple of months, I'm staying with my mom and her husband and I — admittedly somewhat ungratefully — groused to her about how my ABQ connection was seven times faster and the one Google is going to make available here is 333 times faster. Incidentally, it will be symmetric, which is actually very smart of them since I'm certain that will be important for encouraging grass-roots innovation in ultra-high speed internet apps. When anyone in their garage can serve high-speed content and be mostly limited by their processing, not network, bandwidth...that is a different ecosystem than the one we're used to now. It's more like, as it happens, the one we lived in during the BBS era.
posted by Ivan Fyodorovich at 1:34 AM on June 5, 2012
> It must be nice to live somewhere where you can refer to dialup like you're talking about passenger pigeons
Ouch. Yeah. I think it's been over a decade since I was on dial-up, and didn't consider people who aren't so lucky. I'm sure it must suck to be on a slow connection when so many sites assume you can pull down data at megabytes per second.
About the article. I read the it with the mindset that things were probably simplified a whole lot to keep it understandable and readable. I read a piece once (I think by Terry Pratchett) where the thesis was that we aren't being taught the truth in school, it's all lies. But lies that lets us understand some part of what we're learning about. I think he had some specific example, but I can't remember what it was. But something along the lines of you learning something in school about the universe. And then you get to high school and learn that what you were taught was wrong, and that actually works differently. And then you learn even more about it, and discover that what you were taught in high school was wrong too, and that actually works differently, and so on and so on.
Anyway, I just wanted to jump in and thank everyone here for posting such amazing comments, and for Alexis Madrigal for coming over and joining the discussion.
posted by bjrn at 11:05 AM on June 5, 2012 [1 favorite]
Ouch. Yeah. I think it's been over a decade since I was on dial-up, and didn't consider people who aren't so lucky. I'm sure it must suck to be on a slow connection when so many sites assume you can pull down data at megabytes per second.
About the article. I read the it with the mindset that things were probably simplified a whole lot to keep it understandable and readable. I read a piece once (I think by Terry Pratchett) where the thesis was that we aren't being taught the truth in school, it's all lies. But lies that lets us understand some part of what we're learning about. I think he had some specific example, but I can't remember what it was. But something along the lines of you learning something in school about the universe. And then you get to high school and learn that what you were taught was wrong, and that
Anyway, I just wanted to jump in and thank everyone here for posting such amazing comments, and for Alexis Madrigal for coming over and joining the discussion.
posted by bjrn at 11:05 AM on June 5, 2012 [1 favorite]
For the longest time, my mother referred to the modem as the "beep boop bop boop beep boop bop."
posted by Madamina at 12:09 PM on June 6, 2012
posted by Madamina at 12:09 PM on June 6, 2012
I recall sometime around the 1200 baud years (my niece the engineer says "only really _old_ guys say 'baud'") I read a study by some phone company forward planning department that foretold vast profits to be made off of long distance charges as dialup BBSes became increasingly popular, especially in congested areas where several area code lines had to be crossed. And they were fiddling with the regulations to be able to charge the extra long distance rate for modem traffic that went out of the local dialing area and _back_ to a BBS in the same area, I think. Yes, that was slightly before the Internet ate their lunch.
posted by hank at 2:01 AM on June 12, 2012
posted by hank at 2:01 AM on June 12, 2012
One time I called the modem on my computer from a regular phone and tried to whistle the connection sounds. I connected at 300 baud. Then I started making garbley noises into the phone and I watched nonsense type itself out on the terminal.
posted by Galaxor Nebulon at 7:49 AM on June 17, 2012 [1 favorite]
posted by Galaxor Nebulon at 7:49 AM on June 17, 2012 [1 favorite]
That's not so impressive, as your name clearly indicates that you are a robot, Mr. Galaxor Nebulon.
posted by benito.strauss at 8:41 AM on June 17, 2012
posted by benito.strauss at 8:41 AM on June 17, 2012
That was one of the few miasmic Soundcloud pop-up comment streams I really loved reading.
posted by obscurator at 11:50 AM on June 22, 2012
posted by obscurator at 11:50 AM on June 22, 2012
« Older Kyle McDonald Explains FaceTracker | Miami 1981 Newer »
This thread has been archived and is closed to new comments
A mystery solved.
I love that sound more than I reasonably should.
I think I will make it my ringtone.
posted by Mezentian at 2:01 AM on June 3, 2012 [2 favorites]