Taking Stock Of The MP3 At Mid-Life
Last week, Joel Rose wrote about the compact disc on its 30th anniversary, but it could have been an obituary. In the last decade, CD sales in the United States have dropped by more than two thirds, fulfilling a cycle that dates back to wax cylinders and 78 rpm discs: the 20 to 30 year lifespan of a format, followed by the rise of a new technology. So we decided to look at the format that usurped the CD's place in music listener's ears and hard drives, if not always hearts.
The MP3's history is one of innovation and betrayal. Invented in the 1990s by engineers at the Fraunhofer Institute for Digital Media Technology in Germany, it took years of work to create, but its widespread adoption as the CD's successor was triggered by an act of theft. Jonathan Sterne, a professor at McGill University in Montreal, dives into that history in his book MP3: The Meaning of a Format. The MP3 — a format that so many of us now play through technology embedded in our mobile phones — stands on the shoulders of research that began as a way of limiting the sonic reproduction of the human voice over telephone lines nearly a century ago.
As Sterne follows the developments that led from that innovation, he takes into account two factors that intersect at a different point on each new format: (A) how we experience the playback of a format (what he calls its "sensual dimension") and (B) the standards of technology and infrastructure that enable it to reach an audience. In the late 1990s, the CD was a high-quality format and the music industry was built around its distribution. CD players were portable, programmable and cheap. Early MP3s, by comparison, sounded terrible and there was virtually nowhere to get them legally. So how did the the upstart come to dominate the marketplace?
NPR: Why do you think the MP3 overtook the CD?
Jonathan Sterne: "Well, there's so many reasons. It's smaller. It's the whole 10,000 songs in the palm of your hand kind of sell that Apple does. So it's much, much easier to maintain and organize a collection. It's much more portable than a compact disc. And it's much easier to share and exchange. MP3s are very convivial technology. They're not copy protected, mostly. And because they depend on a particular configuration of the Internet — that there's this big infrastructure where we can easily send stuff back and forth. People say they're children of the digitization of music and the compact disc, because if you didn't have that, you couldn't have MP3s, and that's true. But even more so, I think that they're sort of children of high speed internet, especially in the U.S. and Canada, although you'd find very different MP3 economies on the streets of major cities including New York, but also in places like Africa where people are selling memory sticks or CDs that are basically unauthorized copies of whole artists' collections.
"So MP3s are much easier to share, much more portable, all these technologies have been built around them and so they're very successful that way. And so by 1997, MP3 actually surpasses pornography as the most searched for item on the internet. That's in part because people could also get the means to encode and decode the format for free. And very quickly thereafter Apple and Microsoft and people start signing deals with Fraunhofer, the people who own the rights to what was then finally called MP3. And then all of a sudden it was this incredibly valuable property, but at the same time something you and I don't pay for as users. We pay for the hardware to play it back and maybe we pay for the recordings."
NPR: About a year ago, we had a conversation with Karlheinz Brandenburg, one of the guys at Fraunhofer who helped to invent the MPEG Layer 3 encoding process. One of things he told us is that the MP3 has a birthday. July 14, 1995 is the day that they decided on ".mp3" as the file extension. So we can say that the MP3 is 17 years old, but from your perspective, where it its lifespan is the format right now?
JS: "If you go with [the theory that] every format has a 30 year lifespan, it's definitely mid-life. It is the most common form of recorded audio in the world, so that's not going to go away overnight. It's compatible with not only the devices that people use but the ways in which they use music in their lives. And the format itself has been hugely improved since 1995, so the encoding protocols are different. People generally encode at a higher bit rate. People have come up with some pretty good tricks for encoding audio that make it sound better than it once did. I mean, there are all these other competitor formats out there like AAC which Brandenburg and his people developed afterwards and OGG which is open source. So there are competitors, but none of them have what economists and technological historians call the level of path dependency that MP3 has. In other words, if your entire music collection or a large portion of it is coded as MP3, when you buy a new device to play it back, you're going to want that device to play MP3s. It's been widely adopted. People know what it is, they know how to use it and using it is part of the expectations that they bring to, say, buying a new audio system. So plenty of people will buy audio playback equipment that can't play records, but very few people will buy audio playback equipment that cannot play MP3s."
NPR: In the book, you use the terms "channel efficiency" and "aesthetic experience" to describe two criteria by which you can judge the success of each format. I think of those criteria, in basic terms, as convenience of access and sound quality.
JS: "But quality defined in a certain way. So from an engineering point of view, people talk about it very similar to how close it is to reality in terms of definition. One of the most profound things that psychophysicists have shown over and over again is that the human senses aren't actually 100 percent high definition in the way that a microphone or a speaker might be."
NPR: And that's actually hugely important to the process of encoding MP3s.
JS: "One of the things that's so fascinating about this technology is precisely that it presupposes the limitations of the human senses. Although as it turns out — and this is sort of what I discovered in writing the book — many of the major communications developments over the past century have been built specifically, like explicitly on this assumption and operation wise in order to produce channel efficiency. Every format balances this and makes decisions and a lot of that is situational as well in terms of how it's going to be used and how it's going to be engaged with. But over time, those decisions come to matter less and less, and the look and sound and feel of the format actually becomes culturally iconic in and of itself. So there's a whole generation of, not only music fans, but musicians who believe — well they don't believe, they feel — if you listen to drums recorded to tape, they sound right; if you listen to drums recorded to digital they sound wrong. So there are all these adaptations for digital recording systems to make drums sound more like they were recorded to tape for that reason. So we come to things that would've been, in the time, considered imperfections. We think about the use of vinyl noise in hip hop music even today as a sort of aesthetic thing. So these imperfections or compromises actually become part of the sound and something people expect to hear."
NPR: The subtitle of your book is "The Meaning of a Format." What is a format? It's not just the means by which I listen to music; it's not an iPhone, right?
JS: "Right. Format is ... I call it the intersection between the sensual dimensions of a medium and standard protocols and operating routines. Think about the differences between what's called standard definition and high definition TV. It's not just an aesthetic experience, although that's a huge part of it, right? The 'Oh my God, I can see the blades of grass on the golf course' or 'I can see the puck move' or whatever. But it's also what part of the spectrum broadcasters are broadcasting on. It's what hardware you use and how that hardware interfaces with other hardware.
"The reason I'm interested in the MP3 specifically as a format is because I think a lot of the important historical changes in media are now happening in the sort of spaces between hardware and user experience — things sort of below the threshold of your average consumer's perception, like it's not something people pay attention to. They pay attention to the new iPhone. They don't necessarily pay attention to the format in which they receive the music except inasmuch as they want their devices to be compatible with each other. So if you want to actually tell the story of media as devices that produce sort of sensory experiences for people, or with people you could say, and also as sort of economic and technical and cultural institutions, then I think looking at formats is one really good way to do that."
NPR: So explain the MP3 then, in those terms.
JS: "The MP3 is a set of technologies that have a set of assumptions about perception that are really interesting. Let's use the song as our unit, which is like a track on a compact disc. An MP3 is important because it's so much smaller, in terms of the amount of hard drive space or bandwidth it takes up, than the same track on a compact disc. That means it's easier to store, it's easier to transmit. It achieved this smallness through some decisions that are made in the encoding of an MP3. It takes the music thousands of times a second. It says 'What parts of the sound is a person not going to hear?' It removes those parts of the frequencies and then encodes it. So it's actually taking out part of the file that it thinks will be inaudible to you. Now a lot of people listen to MP3s and say they don't sound as good or whatever, but in fact the problems that they hear with them are often not the missing audio, but rather sort of artifacts of the format itself. You could say the equivalent of clicks and pops and scratches on a record. And you know, they don't like that, just as people who were inventing the CD didn't like the clicks and pops and scratches on a record, which of course, now we all think is wonderful and fascinating. So it has this sensuous dimension to it.
"And what's amazing is not that MP3s are perfect — they're not — it's that they work as well as they do, and part of that is because of how people listen to music. If you think about my commute to work today: I get on the metro in my neighborhood and then I get off the metro at school. If I'm listening to music, I'm listening to it on these very cheap ear buds off a device in very noisy environment. Even if I'm listening to the highest possible definition audio, I'm not going to hear most of it. And most people now listen to most music in these sort of imperfect situations in a state of distraction. And in a way, the MP3 actually accounts for it in its code. So that's the sensual part of it.
"But then the standards part of it is really interesting too because MP3s are the most common recorded audio in the world — for the moment. Any device that plays back digital sound pretty much is expected to be able to play MP3s, and so it's a way of perpetuating particular institutional arrangements in the consumer electronics industry. My favorite story about this is actually in 2001, Sony, the conglomerate, did two things that seemed to contradict each other. Sony Records joined a suit against Napster for, you know, Napster promoting the sharing of music files. And Sony Consumer Electronics released a CD player that played back MP3s, so people who did have MP3s that — at that point — they had probably acquired either by ripping their own collection or by downloading through services like Napster, they could play them back."
NPR: Looking at the MP3 compared to other dominant formats, if you're thinking about both convenience and quality, how does this format compare?
JS: "I think it depends on what kind of music you listen to. So at like a knee jerk level, I find MP3s vastly superior to cassettes in terms of sound quality. But I know that there are whole labels dedicated to releasing music on cassette now and are building around that sound, so they would probably disagree with me. MP3s probably can reproduce lower bass than a lot of records can. Are we talking about like a $50,000 audiophile system or something that somebody's going to have in their dorm room? Again, it's a variable thing.
"What I would say is every medium makes compromises. So people always talk about the arbitrary — especially with digital; They say 'Well it's so arbitrary, it doesn't reproduce sound below 20Hz or above 20kHz' and those are arbitrary limits. But you could say the same thing about the RIAA curve that's applied to records, which is also a way of compensating for some of the physical limitations of the medium. So yeah, the thing about them is they sound pretty good. If you're listening on expensive speakers and paying attention, the CD version or the lossless version of something could well sound better than the MP3 version, but you also might not be able to tell, especially if you don't know what to listen for. There's a great quote — and I don't remember if it was Brandenburg or it was Heinz Gerhäuser or one of the other Fraunhofer people who wrote this — but they're setting the standard and somebody says, 'Well you know, it's like when you buy a new house and you look out the living room window and at first you're really impressed with the view, and then time passes and suddenly you notice that there's a big crack in the window.' People don't necessarily, at first, if they're not paying attention — and some people never learn or care to pay attention — notice that crack in the window. And it's the same thing with the MP3. They may never notice issues in the stereo imaging or how it deals with certain kinds of transients or things like that. Those things may completely be irrelevant to their enjoyment of music."
NPR: But the MP3 can get pretty close to CD quality, too, depending on how you encode. And in terms of channel efficiency ...
JS: "In terms of channel efficiency, it's way, way, way better. For one thing, it's much smaller. It's so small you can't hold one in your hand in the way — you can hold thousands of them in your hand inside a hard drive or inside flash memory — but you can't hold just one, whereas a CD you can't hold thousands in your hand. In fact, holding hundreds is bad for your back. So in terms of channel efficiency, there's no contest."
NPR: It's like having a house that's enormous — so enormous and has so many windows — that the crack is less meaningful.
JS: "Yes. But again, it matters to users because some people who listen to music want channel inefficiency. They don't want efficiency. Think about someone who deliberately chooses to listen to records today. That is an aesthetic decision. I'm going to shop for music in different places. I've got to flip it every 15 to 20 minutes. So people actually do choose channel efficiency. As with definition, it's not an unmitigated good. People come to love the inefficiencies just as they come to love the distortions in a format and that can be an important part of the listening experience too. So while our MP3s are, without question, more efficient in their use of storage space and bandwidth and much more portable than compact discs, the efficiency — just as the definition — may not be the value that the music listener is pursuing."
NPR: Why isn't the MP3 already a thing of the past?
JS: "I think the main thing is path dependence. People say there are new higher quality formats. You can basically stream CD quality audio now in the form of lossless files, but I think there's a lot of resistance to it. For one thing, anybody with a large MP3 collection probably still didn't pay for a large proportion of that music. Second, MP3s are just so completely common. It's the most obvious thing: It's the format you're most likely to get digital music in and it is the format that any digital device is most likely to be able to play. So as long as those two conditions obtain, it is unlikely that MP3s will go away.
"I think to challenge them, it's not just a matter of improving the quality of sound reproduction or whatever when you look at the changes from one format to another. It's always about changing context of use, changing ways of dealing with users by companies, and then struggles among industries and sometimes regulators about how to shape media and how to distribute music I guess. So you know people often say to me 'Well we can play much better quality audio even over the internet, why are we still using MP3?' And the answer is because it was never about just ever increasing technical standards or higher and higher definition. It's always about all these others things at once. And if that's not enough, a couple of people have done some not very well documented but interesting studies of university undergraduates where they play the same song in many different formats, and not surprisingly, the undergrads tend to think that the MP3 sounds more like music is supposed to sound, just as your baby boomer is going to say music should sound like it sounds on records."