Michel Chion’s analog tape techniques

Guest Contribution by Michel Chion

For this month’s theme, french  composer and researcher Michel Chion very kindly allowed me to translate a tutorial he wrote about analog tape techniques. This tutorial is actually an appendix of a book on the much larger subject of concrete music, La musique concrète, art des sons fixés (the examples given by Chion are also from music pieces, although nothing forbids the same techniques to be applied in sound design).

Talking about analog tape techniques may seem a little offbeat since our expectations over a tutorial generally gravitate around a fast way to learn how to use recent or new technology. A tutorial on analog tape techniques may not feel practical at all: these machines are bulky and cumbersome compared to a mouse and a keyboard, most of people don’t have access to them (even if they’re only one-click away on ebay). More importantly, they’re considered old, obsolete, technology. But, is this really true? Simply because they’re not fabricated anymore, it doesn’t mean analog tape recorders don’t have unique features worth exploring. Besides that, as shown by this tutorial, their use may influence the way we perceive sound at our work :

1.  concerning its time/duration, we’re not facing a visual representation of sounds as in a DAW environment where 1 hour or 1 millisecond differ only by a zoom movement at a monitor screen, therefore the realization of sound as something evolving in time becomes much more acute with analog systems;
2.  concerning its material, plastic nature: most of the techniques rely on manually manipulating the tape or are aimed to create variations over sonic texture;
3. finally, the fact that these machines are mechanical devices multiplies the number of variables in contrast with the predictability of digital systems: the same thing never happens twice, accidents are bound to occur and a mistake might actually become an expressive tool.

That being said, I hope you enjoy Chion’s tutorial below:

“With an analog tape machine, we can:

a) Record and playback a sound

No further explanation required

b) Play a sound backwards

To play a sound backwards, which is only possible with a 2-track tape machine, gives the most varying results, sometimes really trivial and mechanic, sometimes surprising ones. The only way to know is by doing it. The greatest advantage of the analog tape recorder in comparison with the same process executed in a DAW environment is that you don’t alter the sound by giving it a “digital texture”.

c) Change the speed of the recorder to transpose a sound or choose a slower speed in order to have a smoother sound regarding transients

The choice between two speeds, sometimes even three or four on certain recorders, was, at first, a resource given to the user to spend less tape (in the same way that old VCRs gave you the possibility to record at SP or EP, for example). It was possible then to record large sections at a slower speed, losing quality but saving tape, or to record at a higher speed for the best quality available but spending two times the amount of tape.

This feature opened the possibility to transpose a sound, but it could also be used to make a sound smoother or sharper depending on the choice of speed or by working with different speeds in recording and reproduction.

The simple choice between 7 1/2 ips (inches per second) or 15 ips has some interesting consequences: we can choose to work at a slower speed just because the sound will have a smoother texture. In my compositions, it is not unusual for me to blend sounds recorded at 7 1/2 ips, even 3 3/4 ips, with others recorded at 15 ips. This gives a richer, less uniform, combination of textures.

I did not get the opportunity to work with 30 ips recorders, avid tape consumers that allowed very detailed editing possibilities. These were mostly used during the fifties by composers like Stockhausen, Pierre Henry, Luc Ferrari…When I entered the GRM (Groupe de Recherches Musicales), the professional standard was already 15 ips.

As it is best known, altering the speed transposes the sound, making it higher or lower in pitch at the same time it gets shorter or longer by being played faster or slower.

d) Fast rewind with or without the tape touching the playback head

The fast rewind of the tape, if it touches the playback head, creates high pitched swirling sounds, most of the time annoying , but eventually interesting for a few seconds. Ivo Malec used this technique quite well at his 1970 piece, Luminétudes. Avoid doing this for a long time or too often because it can wear out the playback head of the tape recorder.

If you do the same rewind movement keeping the tape away from the playback head, close but without actually touching it, you can achieve a very poetic effect which is kind of a washed-out sound: the playback head still reads the signal but with a lot less high frequencies creating an unreal feel. That is how I did the sounds for the beginning of the Offertoire part of my Requiem.

e) Use or creation of a wow (flutter) effect

An analog tape machine, even a high quality one, has always a certain degree of wow, which is caused by subtle irregularities of the playback speed generating small variations in pitch. This wow effect is stronger at low playback speeds ( 7 1/2 ips and 3 3/4 ips, where it’s quite clear) and more noticeable in steady tones.

This “defect” can actually work in our favour if we are trying to create a sound that is unstable and fluttering.

I have recorded at a low speed a few times just to have this kind of irregularity, mechanically originated, which gives a trembling effect to the sound, making everything more alive. At the same time, we can interfere with the free movement of the tape, obstructing it partially, or making it oscillate using our hands, a piece of cardboard or even a pen. I did this at the beginning of my Credo.

I must clarify that this operation has nothing to do with slowing down the speed of a sound as in a varispeed effect. The sound is recorded and played at the same low speed. There is no variation between these two stages and the pitch and duration of the sound therefore are not affected by this technique. Nevertheless, this effect produces a very subtle oscillation, hardly noticeable, that could be compared to a vocal or instrumental vibrato.

If I have the sound of a sustained note created using a synthesizer, an electric or mechanic organ, etc. and I record this steady note on tape, I’m able to make it vibrate by using my hands to interfere with the regularity of its running – grabbing the tape between my thumb and my index finger and hitting it rhythmically, slowing it down slightly, etc. This is what I call défilement contrarié (partially obstructed running). The tape is still running, but not at the same rhythm as the motor. This is for me one of the advantages of an analog system. More flexible due to their mechanical design, they accept much better these kinds of intervention, these “mistakes” that end up becoming creative tools. The same happens when we compare, for example, an analog distortion and a digital one; in the first case, sound is not destroyed, only slightly changed, acquiring a different color.

Once again I must say that I don’t think that everyone should convert to the exclusive use of “analog” systems, but they still can have their use combined with digital technology.

f) Blocking of the tape in the record position, allowing to run the tape manually while recording

Not all tape machines allowed to do this, but the famous Uher Royal de Luxe (sic), a consumer tape recorder with four speeds that wasn’t very expensive in the sixties, gave that possibility. If, by using our hand, we make the tape run slower, the recorded sound will have its pitch raised when played later; the opposite will happen if we make the tape run faster. The most interesting use of this technique is to make constant variations of speed that will generate some interesting glissando like effects. It’s like a scratching disc effect, but achieved in the opposite way since we have to go slower during record to achieve a higher pitch and faster for a lower pitch.

g) Multiplayback

Some tape recorders like the above mentioned Uher Royal de Luxe (mine is out of order, but I guess that there must be still dozens of these recorders here in France still able to provide service) allowed to use two tracks at the same time and make a mono mix out of them. By adopting this technique with another Uher recorder, my good friend Ghedalia Tazartes accomplished real masterpieces in the seventies by the use of juxtapositions of his voice.

More simply, because of the existence of these two separate tracks and the built-in speaker of the Uher, I could record myself first on track 1 and then on track 2, but hearing simultaneously what I had recorded before on track 1. So, just like this, between 1969 and 1972, I did a lot of duos with myself. For example, I would play a small section on a prepared piano and afterwards I would add the sound of a vocal or a flute. I didn’t even have to use headphones as long as I set the level of the speaker not very loud and avoided pointing the microphone towards it.

h) Mechanical editing

The editing process is mechanical on these machines just like it was with film editing. It’s done with a special scotch tape and scissors. You don’t really need a tape splicer for cutting. It takes no more than fifteen minutes to explain how to do it, but months to really learn how to do it fast and efficiently. The editing process is rightfully considered a process of articulation and construction, but it is also a capital moment of transformation and discovery. A powerful way to create sounds is to choose specific, meaningful moments of a recording in a process that I call “time framing” (cadrage temporel) of a sound, one by which the original recording can become a whole different sound. These possibilities are incredibly rich as long as the sound from departure is a lively one and presents variations at its course. The same can’t be said for a sound that is steady in nature, like a telephone tone – in this case, it’s clear that nothing will happen.

These possibilities of reframing a sound are really unpredictable – a sound that seems trivial and boring may give us something great and mesmerizing if we reframe it properly. To find these moments, I often use this technique of random playback. I choose a certain part of the tape without listening to it and hit play. After a lot of trials, I finally hear something like an attack that seems surprising or a beautiful form emerging from the recording. As important as this is the moment where I choose to interrupt the sound and hit stop.

Of course we can do the same thing on a computer, this sort of blindfolded reframes. Nevertheless, in order to come back to what is specific of the analog tape recorder, and since I just mentioned the attack of a sound as something important, let’s not forget that at tape editing a transition is harder or softer according to the inclination of the cut (the more perpendicular, the harder it is).It’s true that we can do something similar on a computer, but with analog tape I often do bad aligned transitions on purpose in a way that the playback is slightly affected at the junction of the two pieces of tape, which creates a kind of “hiccup” effect that I find interesting. Even the clicks that we hear at the junction of softer sounds joined by a perpendicular cut may have a nice effect at the right context: it’s the same with photography when we consider the expressive use of defects like an out of focus image, a lens flare, etc.

i) “Crayonnés” (Sonic overprints)

“Crayonné” or “crayonnage” is the name I gave to a technique I invented in 1988 and that I’ve been using extensively ever since, from my Dix études de musique concrète up to La vie en prose. A few other composers and colleagues of mine like Christine Groult and Lionel Marchetti have adopted it in their works as well and also teach this technique to their students.

This process – which I describe in my composition Crayonnés ferroviaires (1992) – is possible to achieve with all kinds of analog tape recorders, but it’s better executed with a Revox B 77 since this recorder has the three heads (erase/playback/record, aligned in that exact order) positioned in a way that is easier to reach them or using the more expensive, but still at a consumer level, Revox A 700. It would be even easier to achieve if you have at your disposal a professional studio recorder like an A 80 or PR 99 Studer machine.

What is the “crayonné”? It’s the use of a two track magnetic tape as a surface over which we can do various sonic overprints by erasing, or not, the previous recorded sound at a certain spot. While recording something, I intervene manually over the contact surface of the tape with both the erase and record heads ( alternatively you can use a cotton swab for this or anything else you think is better suited). From moment to moment, we choose to record more or less on the tape erasing more or less what has been previously recorded. By keeping the tape a little bit apart from the erase head, the sound, instead of just disappearing, becomes muffled and lower in volume. When we put the tape again in contact with the erase head, slowly or abruptly, we hear the sound of the previous layer disappear.The same applies to the record head. If we move the tape away from it, the new sound is still recorded, but it becomes muffled and filtered with a lower level.

In both cases, that can be performed at the same time by using both hands or with the help of a second person, the interesting thing is the creation of a very subtle gradation between a “completely erased” state and a “not erased at all” state of the previous layer. A sound can appear progressively rich and precise with its full spectrum or can be gradually faded away. The same applies for the new layer that we are recording.

At the opposite side, we can use the same technique to do a fast, random and dazzling sound montage by putting the tape abruptly against the record head or by moving it away fast from the erase head, which makes a new sound “appear” suddenly or a previous recorded one reemerge from the first layer.

A random montage, I said. Why not? This process is very much like a live performance in which we can not anticipate many things. But we can listen immediately to what we just did. That’s when the randomness of the process finishes and we take control by choosing our best moments from the various versions we’ve recorded. That’s how I did the movement Danse de l’Ombre, from my 24 Préludes a la vie, and the most eventful parts of Gloria.

j) Use of the inertia of the motor

This inertia is characteristic of analog systems. Playback isn’t instantaneous as in Pro Tools for example; it’s always preceded by a small “hiccup”, which is never exactly the same. This is due to the mechanical nature of this kind of equipment. When we use an analog tape recorder, we put the tape a few seconds before the point where we actually want to begin, more precisely where the leader is (a leader tape is a nonmagnetic tape used at the beginning or end of a reel or as a marking for the beginning of a selection within a tape).

In Purgatoire, a work of François Bayle inspired by the the poem of Dante Alighieri, at the moment where one of the characters says: I am Aglaor who became a rock, we hear quite clearly Bayle pushing the play button of the recorder over the I am, which ends up sounding like an engine starting. This effect, to my knowledge never repeated by Bayle, is really strong and in tune with the situation depicted: it gives the intensity and the violence needed for that specific line. I applied the same technique on my Requiem. In his 1970 work, Luminétudes, Ivo Malec did an amazing use of this technique.

k) Reversed tape

It consists of reversing the tape in a way that the nonmagnetic side of it is the one touching the playback and record heads, which is at first considered an error . By doing this we can still hear the recorded sound but it gets muffled as if heard behind a closed door. The effect is sometimes disappointing, sometimes interesting.

l) Record a sound played trough the built-in speakers

Some professional recorders such as the Studer A 80 have a built-in speaker of terrible audio quality. The sound played trough this speaker becomes a caricature of the original and we can record such a caricature by placing a microphone next to the speaker. In Gloria, some pathetic and “strangled” voices were recorded this way.

From all the techniques described above, there are three or four that I think were created by me (the “crayonné” and the “défilement contrarié” specially), but anyone can discover others. Of course these techniques are based upon a limited number of parameters, but what we must realize is that each one of them is still an individual case depending on numerous variables. Everything plays a part at the final result: the characteristics of the recorder we use, the characteristics of the tape, the sounds we choose to work with, etc. The macroscopic aspect of it, the concrete nature of these operations done with a mechanical system (at the opposite side of the microscopic, non-variable world of digital media) is what assures this diversity. We must not think of these as “defects”: a straight line drawn by hand must not be considered a faulty one compared to another straight line done with a ruler; it only has a different quality to it.

Michel Chion is a french composer and researcher who has devoted a large part of his career to the study of the audiovisual relationship in cinema. Some of of his most important books on this subject were translated to english by Claudia Gorbman: “The Voice in Cinema”, “Audio-Vision: Sound on Screen” and “Film, a sound art”. For more info, please consult his personal website: michelchion.com