Brian Eno’s famous essay, first given as a lecture in 1979, and then printed in Down Beat magazine in July and August of 1983…
Brian Eno delivered the following lecture during New Music New York, the first New Music America Festival sponsored in 1979 by the Kitchen. His remarks were amplified by demonstrations from his own recordings; here we’ve attempted to excerpt the general sense of his more specific points.
The first thing about recording is that it makes repeatable what was otherwise transient and ephemeral. Music, until about 1900, was an event that was perceived in a particular situation, and that disappeared when it was finished. There was no way of actually hearing that piece again, identically, and there was no way of knowing whether your perception was telling you it was different or whether it was different the second time you heard it. The piece disappeared when it was finished, so it was something that only existed in time.
The effect of recording is that it takes music out of the time dimension and puts it in the space dimension. As soon as you do that, you’re in a position of being able to listen again and again to a performance, to become familiar with details you most certainly had missed the first time through, and to become very fond of details that weren’t intended by the composer or the musicians.
The effect of this on the composer is that he can think in terms of supplying material that would actually be too subtle for a first listening. Around about the 1920s — or maybe that’s too early, perhaps around the ’30s — composers started thinking that their work was recordable, and they started making use of the special liberty of being recorded.
I think the first place this had a real effect was in jazz. Jazz is an improvised form, primarily, and the interesting thing about improvisations is that they become more interesting as you listen to them more times. What seemed like an almost arbitrary collision of events comes to seem very meaningful on relistening. Actually, almost any arbitrary collision of events listened to enough times comes to seem very meaningful. (There’s an interesting and useful bit of information for a composer, I can tell you.) I think recording created the jazz idiom, in a sense; jazz was, from 1925 onwards, a recorded medium, and from’35 onwards I guess — I’m not a jazz expert by any means — it was a medium that most people received via records. So they were listening to things that were once only improvisations for many hundreds of times, and they were hearing these details as being compositionally significant.
Now, let’s talk about another aspect of recording, which I call the detachable aspect. As soon as you record something, you make it available for any situation that has a record player. You take it out of the ambience and locale in which it was made, and it can be transposed into any situation. This morning I was listening to a Thai lady singing; I can hear the sound of the St. Sophia Church in Belgrade or Max’s Kansas City in my own apartment, and I can listen with a fair degree of conviction about what these sounds mean. As Marshall McLuhan said, it makes all music all present. So not only is the whole history of our music with us now, in some sense, on record, but the whole global musical culture is also available. That means that a composer is really in the position, if he listens to records a lot, of having a culture unbounded, both temporally and geographically, and therefore it’s not at all surprising that composers should have ceased writing in a European classical tradition, and have branched out into all sorts of other experiments. Of course, that’s not the only reason that they did, either.
So, to tape recording: till about the late ’40s, recording was simply regarded as a device for transmitting a performance to an unknown audience, and the whole accent of recording technique was on making what was called a “more faithful” transmission of that experience. It began very simply, because the only control over the relative levels of sounds that went onto the machine was how far they were from the microphone — like device. The accent was on the performance, and the recording was a more or less perfect transmitter of that, through the cylinder and wax disc recording stages, until tape became the medium by which people were recording things.
The move to tape was very important, because as soon as something’s on tape, it becomes a substance which is malleable and mutable and cuttable and reversible in ways that discs aren’t. It’s hard to do anything very interesting with a disc — all you can do is play it at a different speed, probably; you can’t actually cut a groove out and make a little loop of it. The effect of tape was that it really put music in a spatial dimension, making it possible to squeeze the music, or expand it.
Initially tape recording was a single track, all the information contained and already mixed together on that one track. Then in the mid-’50s experiments were starting with stereo, which was not significantly different. The only difference was that you had two microphones pointing to your ensemble, and you had some impression of a real acousticsound came to you from two different sources as you listened. Then came threetrack recording; it allowed the option of adding another voice or putting a string section on, or something like that. Now this is a significant step, I think; it’s the first time it was acknowledged that the performance isn’t the finished item, and that the work can be added to in the control room, or in the studio itself. For the first time composers – almost always pop composers, as very few classical composers were thinking in this form — were thinking, “Well, this is the music. What can I do with it? I’ve got this extra facility of one track.” Tricky things start getting added. Then it went to four-track after that, and the usual layout for recording a band on four track at that time.
You should remember that everything, including the Beatles Sgt. Pepper’s Lonely Hearts Club Band, was done on four-track until 1968. Normally engineers would do something like this: the drums on one track, the voices spread on two tracks with the guitars and the piano, say, on one of those tracks, and then the strings and additional effects on the fourth track. This was because they were thinking in terms of mono output; eventually, it would be mixed down to one signal again, to be played on radio or whatever. When stereo came in big, it gave them a problem. When they converted to stereo, things were put in either the middle, or dramatically to one side, or you’d hear some very idiosyncratic panning.
Anyway, after four-track it moved to eight track – this was in ’68, I guess — then very quickly escalated: eight-track till ’70, 16-track from’70 to’ 74, 24-track to now when you can easily work on 48-track, for instance, and there are such things as 64-track machines. The interesting thing is that after 16-track, I would say, the differences are differences of degree, not differences of kind. Because after you get to 16-track, you have far more tracks than you need to record a conventional rock band. Even if you spread the drums across six tracks, have the basson two, have the vocals, have the guitars, you’ve still got six tracks left. People started to think, “What shall we do with those six tracks?”
From that impulse two things happened: you got an additive approach to recording, the idea that composition is the process of adding more, which was very common in early ’70s rock (this gave rise to the well known and gladly departed orchestral rock tradition, and it also gave rise to heavy metal music — that sound can’t be got on simpler equipment); it also gave rise to the particular area that I’m involved in: in-studio composition, where you no longer come to the studio with a conception of the finished piece. Instead, you come with actually rather a bare skeleton of the piece, or perhaps with nothing at all. I often start working with no starting point. Once you become familiar with studio facilities, or even if you’re not, actually, you can begin to compose in relation to those facilities. You can begin to think in terms of putting something on, putting something else on, trying this on top of it, and so on, then taking some of the original things off, or taking a mixture of things off, and seeing what you’re left with — actually constructing a piece in the studio.
In a compositional sense this takes the making of music away from any traditional way that composers worked, as far as I’m concerned, and one becomes empirical in a way that the classical composer never was. You’re working directly with sound, and there’s no transmission loss between you and the sound — you handle it. It puts the composer in the identical position of the painter — he’s working directly with a material, working directly onto a substance, and he always retains the options to chop and change, to paint a bit out, add a piece, etc.
Compare that to the transmission intervals in a classical sequence: the composer writes a piece of music in a language that might not be adequate to his ideas — he has to say this note or this one, when he might mean this one just in between, or nearly this one here. He has to specify things in terms of a number of available instruments. He has to, in fact, use a language that, like all languages, will shape what he wants to do. Of course, any good composer understands that and works within that framework of limitations. Finally he has something on the page, and by a process this arrives at a conductor. The conductor looks at that, and if he isn’t in contact with the composer, his job is to make an interpretation of it on the basis of what he thinks the composer meant, or whatever it is he’d like to do. There’s very likely another transmission loss here — there won’t be an identity between what he supposes and what the composer supposes. Then the conductor has the job of getting a group of probably intransigent musicians to follow his instructions, to realize this image of the music he has. Those of you who work with classical musicians know what a dreadful task this is, not to be wished on anyone.
So they come up with something. One can see there’s not necessarily an identity between what the composer — or the conductor — thought, and what they did, so that’s three transmission losses. I’d argue there is another one in the performance of the piece: since you’re not making a record, you’re not working in terms of a controlled acoustic, and you’re not working in a medium that is quite so predictable as a record. If I make a record, I assume it’s going to be the same every time it’s played. So I think there is a difference in kind between the kind of composition I do and the kind a classical composer does. This is evidenced by the fact that I can neither read nor write music, and I can’t play any instruments really well, either. You can’t imagine a situation prior to this where anyone like me could have been a composer. It couldn’t have happened. How could I do it without tape and without technology?
One thing I said about the traditional composer was that he worked with a finite set of possibilities; that is, he knew what an orchestra was composed of, and what those things sounded like, within a range. If you carry on the painting analogy, it’s like he was working with a palette, with a number of colors which were and weren’t mixable. Of course, you can mix clarinets and strings to get different sounds, but you’re still dealing with a range that extends from here to here. It’s nothing like the range of sounds that’s possible once electronics enter the picture. The composer was also dealing with a finite set of relationships between sounds; the instruments are only so loud, and that’s what you’re dealing with, unless you stick one out in a field and one up close to your ear. It was out of the question that he could use something, for example, as the Beach Boys once did — making the sound of someone chewing celery the loudest thing on a track.
Of course, everyone is constrained in one way or another, and you work within your constraints. It doesn’t mean that suddenly the world is open, and we’re going to do much better music, because we’re not constrained in certain ways. We’re going to do different music because we’re not constrained in certain ways we operate under a different set of constraints. I want to explain how multitrack technology works, not electronically, but how it works in spirit. On a 24-track tape recorder you have two-inch tape — it’s that wide — on two big, heavy reels. You have 24 record heads, 24 playback heads. If you want to record a band, you can put one microphone on the bass drum, one microphone on the snare drum, one microphone — on the drummer’s knee-joint if you like — you can separate things very carefully. You can end up with this two-inch piece of tape with 24 distinct signals, and once you’re in this position, you have considerable freedom as to what you can do with each of these sounds.
You can do what the classical composer couldn’t: you can infinitely extend the timbre of any instrument. You are also in the position of being able to subtract or add with discrimination: you can put an echo on the bass drum and not on anything else. The 24-track tape works to separate things off, and keeps them separate until you feed the whole thing back through a mixing head, and you mix it all in some manner of your choice. The mixer is really the central part of the studio.
Most people see a large mixer, and they’re completely bewildered because there are something like 800 or 900 knobs on it. Actually it’s not so complex as it looks – it’s the same thing repeated many times. Since you’re dealing with 24 tracks, everything has to be multiplied by 24; it’s not a very complex system. Each track from the tape recorder plays back on one channel of the mixer. Each individual channel has a whole set of controls that duplicate the other channels; that’s all.
Each channel on the mixer is a long strip. Generally at the bottom is a level control, for how loud you want that channel to play back. Next up, normally, there’s a pan control, for where you want the sound object in the stereo/quad image. Next up is an echo control, and echo is really a separate issue, which has to do with something very unique to recording: briefly, it enables you to locate something in an artifical acoustic space. There’s also equalization — a device by which you can create a timbral change in an instrument, which in rock music is especially important, because many different rock records, in my opinion, are predicated not on a structure, or a melodic line, or a rhythm, but on a sound; this is why studios and producers keep putting their names on records, because they have a lot to do with that aspect of the work. Apart from equalization, there are other facilities which are widely used, such as limiting, compression — which has the effect of altering the envelope of a note or an instrument, so you can do something I’ve been interested in, creating hybrid instruments.
Compression is quite interesting over a whole track; if you’re using severe compression and limiting at the same time, when you push one instrument up, the track is governed so that the overall level will never change. Pushing one instrument up effectively pushes the others down, so all you do is alter the ratio between the instruments where you make a move. I started to use this as a deliberate, compositional, sound-type device; it’s generally been ignored or regarded as a misuse of the equipment before, but I’ll let you judge for yourself. On “Helen Thormdale” from the No New York album (Antilles), I put an echo on the guitar part’s click, and used that to trigger the compression on the whole track, so it sounds like helicopter blades.
Naturally, all of these things are variable throughout the entire course of the music. These are the kinds of things that you, as a listener, don’t generally notice; some of them operate almost subliminally — they are the ambiance of a track, not the obvious aspects of the track. Those are very much the things that traditional production is concerned with. And they allow you to rearrange the priorities of the music in a large number of ways.
We’ve spoken of the transition from the ’50s concept of music to the contemporary concept of mixing. If you listen to records from the ’50s, you’ll find that all the melodic information is mixed very loud — your first impression of the piece is of melody — and the rhythmic information is mixed rather quietly. The bass is indistinct, and the bass is only playing the root note of the chord in most cases, adding some resonance. As time goes on you’ll find this spectrum, which was very wide, with vocals way up there and the bass drum way down there, beginning to compress, until at the beginning of funk it is very narrow, indeed. Things are all about equally loud.
Then, from the time of Sly and the Family Stone’s Fresh album, there’s a flip over, where the rhythm instruments, particularly the bass drum and bass, suddenly become the important instruments in the mix. A timbral change also takes place. The bass becomes a very defined instrument; by the use of amplitude control filters, the bass actually begins to take on a very vocal attack. The bass drum gains a more physical sound, and also has a click to it; generally you’ll find that bass drums are equalized very heavily, something like 1000-1500 cycles, to give a real sharp click. It becomes the loudest instrument in disco — watch the vu meter while a disco track is playing, and you’ll see the needle peak each time the bass drum hits.
Okay. I’ve been talking about some of the possibilities of multi-track recording, which is almost completely what I do. I don’t really have a musical identity outside of studios. Now I’m going to discuss some pieces of mine, because I know how they were made, production-wise, and I can say with confidence how they were built.
Starting with R.A.F, a very obscure B-side of an even more obscure single that came out in ’78 — it’s an interesting piece on a lot of levels. It’s by me and a band called Snatch. This piece started off many years ago; it was just a tatty little tape left over from a mess – around we’d had in the studio which lasted 35 seconds. But that 35 seconds was quite interesting — after that it deteriorated into jamming — but I always kept in mind that I was going to do something with that piece, sometime. I have about 700 pieces like that. Judy Nieland of Snatch suggested doing a reportage piece on the Baader Meinhoff terrorists, and I remembered this piece and pulled it out.
The first thing I had to do was extend it somehow, so I copied the 24-track onto another 24-track machine, four or five times, and I pieced them together, so I had the thing song-length by then. And you’ll hear, in a cleverly disguised fashion, exactly the same parts repeated. Which makes you think that Percy Jones of Brand X is an incredible bass player, because he does every complex, idiosyncratic thing three our four times in a row. That’s a trick I like using.
We had a recording Judy made in Germany of the telephone announcement you could call, where a lady would say, “Good evening, blah blah blah, we’re trying to apprehend the Baader Meinhoff terrorists, this is a recording of one of their voices,” and then the terrorist’s voice would come on, which had been recorded off another telephone when they were making ransom demands. The scenario of this piece was interesting, production-wise, because some of the record is set outside, on the streets, then it suddenly cuts to an airplane which is being hijacked. I wanted to get the effect of going from a very hectic, open space into a very tight, air-conditioned airplane. What I did to achieve that was take all the echo off of everything, and put a very peculiar, tunnel-type echo on things. To me, it works: I get this sense of a contraction of space, and the soft voices working over it. After that it’goes back outside, into the wide world again.
There are two pieces of mine, “Skysaw” from Another Green World, and “A Major Groove” from Music for Films (both Editions EG), which are exactly the same track, mixed differently, slowed down, and fiddled about with a bit. I also gave it to Ultravox for one of the songs on their first album. It’s been a long way, this backing track. Listen to all three, and you hear what kind of range of difference usage is possible. “M386″ on Music for Films is another one that’s had four different lives. This is actually quite similar to what reggae producers have been doing for a while. Once you’re on tape, there are so many variations you can make that you don’t really.need to spend all that money hiring musicians; you can do a great deal with one piece of work. So when you buy a reggae record, there’s a 90 percent chance the drummer is Sly Dunbar. You get the impression that Sly Dunbar is chained to a studio seat somewhere in Jamaica, but in fact what happens is that his drum tracks are so interesting, they get used again and again.
This takes us to reggae, which is a very interesting music in that it’s the first that didn’t base itself around the standard approach of making work by addition. Earlier I said the contemporary studio composer is like a painter who puts things on, puts things together, tries things out, and erases them. The condition of the reggae composer is like that of the sculptor, I think. Five or six musicians play; they’re well isolated from one another. Then the thing they played, which you can regard as a kind of cube of music, is hacked away at — things are taken out, for long periods.
A guitar will appear for two strums, then never appear again; the bass will suddenly drop out, and an interesting space is created. Reggae composers have created a sense of dimension in the music, by very clever, unconventional use of echo, by leaving out instruments, and by the very open rhythmic structure of the music. Then, too, someone like Lee Perry, a producer who’s always been very intelligent as far as using the constraints of the situation goes, might find there’s hiss building up on tracks he’s used over and over. A Western engineer might get frightened by this, and use all sorts of noise reduction and filtration. Perry says, “Okay, that’s part of the sound, so we’ll just add something else to it and use it’ ” This adds an ambiance of weirdness behind what was straightforward reggae.
Which puts me in mind of the first piece on Music for Airports (Editions EG). I had four musicians in the studio, and we were doing some improvising exercises that I’d suggested. I couldn’t hear the musicians very well at the time, and I’m sure they couldn’t hear each other, but listening back, later, I found this very short section of tape where two pianos, unbeknownst to each other, played melodic lines that interlocked in an interesting way. To make a piece of music out of it, I cut that part out, made a stereo loop on the 24-track, then I discovered I liked it best at half speed, so the instruments sounded very soft, and the whole movement was very slow. I didn’t want the bass and guitar — they weren’t necessary for the piece — but there was a bit of Fred Frith’s guitar breaking through the acoustic piano mic, a kind of scrape I couldn’t get rid of. Usually I like Fred’s scrapes a lot, but this wasn’t in keeping, so I had to find a way of dealing with that scrape, and I had the idea of putting in variable orchestration each time the loop repeated. You only hear Fred’s scrape the first time the loop goes around.
There are other examples of things I do with loops and editing based on fairly simple material, to get singular, very rare events I couldn’t have forseen. But perhaps I should mention that you only have control of your studio composition to the pressing plant — then the reproduction is completely arbitrary. So when I mix a record, I mix on at least two speaker systems — and often more than two — so I’m not mixing just for optimum conditions. Most of my records don’t sound good in optimum conditions, where there are very large speakers which are extremely well balanced and have lots of high and low frequencies. I mix, really, for what I imagine most people have medium-priced hi-fi — and for radio a bit as well. It’s the very naive producer who works only on optimum systems.