Projects » performance practise

Live Performance with Pure Data + MIDI Controller

Over the last four years I have been evolving a live performance practice using various iterations of software and a laptop computer. The current iteration of this practice uses a MIDI controller keyboard to control a series of patches written in the open source music development environment Pure Data.

The patches are built so as to allow me to perform improvised electroacoustic music, either solo or as part of an ensemble of other improvisers playing more traditional acoustic instruments. During development the following three core requirements have made themselves apparent for attaining this goal.

  1. agility – being able to move from one sound to another and then another, and back again, rapidly, reliably, and as instantaneously as possible;
  2. spectral integration – being able to fit in with a group of other musicians playing acoustic instruments, without dominating, while still being differentiable, through high-level filtering and textural control; and
  3. physical communicatian – being able to communicate with other musicians, and the audience, through eye contact and body language.

1. Agility

Agility is a key part of free improvised music. Agility means being able to jump from one sound to another, and then back again. It means having direct hair-trigger control over the entirety of the sound output. If, as Xenakis says, “[w]ith the aid of electronic computers, the composer becomes a sort of pilot: pressing buttons, introducing coordinates and supervising the controls of a cosmic vessel sailing in the space of sound,” (Iannis Xenakis, Formalised Music, 1971) then the electronic performer needs the ability to turn this cosmic vessel on a 50c piece, to execute instant full-stops and reversals, and even to jump from one kind of space to another, instantaneously.

From an instrument design point of view, this would seem to mean that direct control over any significant parameter or set of parameters in the synthesis engine is required, through a physical controller on the MIDI controller. However, this need must be balanced with a limited number of physical controllers, and also the recognition that with only two hands, the performer must be able to make significant musical changes to the sound output.

In tackling this problem I have had to make a lot of concessions, and this is the part that is still under development. What it boils down to is a recognition that, within a particular subset of the instrument, sets of controls have interdependant relationships on the sound. This is what leads to a common problem when first learning to play analog synthesis equipment: you turn a lot of knobs and hear no change at all to the sound output, and then suddenly reach a setting where any touch causes radical changes to the sound. A computer scientist I studied under in conversation described this effect as ‘finding local maxima and minima on an n-dimensional surface,’ where n is the number of knobs. It’s hard, in other words. It’s like trying to tune a radio with three tuning dials, instead of just one, where finding a station requires each of the three dials to be in a particular position. (This is edging on to cybernetics and control systems theory, which are fascinating interdisciplinary areas of study that not many people are exploring right now, but they should be.)

But we can use this. In most case you want to exercise a subtle control of the sound, and it is only sometimes that the sudden control is necessary. So at the software level you engineer your instrument to enable these kind of interdependant relationships, and then practise, practise, practise so that your muscle memory knows the positions to put the knobs to get the level of radical change to the sound that you need.

2. Spectral Integration

Spectral integration is about not dominating the entire spectrum with electronic sound. Electronics are capable of generating every possible audible frequency, whereas acoustic instruments tend to be limited to particular frequency bands. This difference can make it difficult to integrate electronic instruments with acoustic instruments in a band situation, with the electronic musician sometimes finding themselves in a position of supplying a sonic backdrop, the acoustic musicians playing figure to the electronic musician’s ground.

The other side of this issue is the problem of becoming lost in the sound of all the instruments. Without a well-defined spectral centre it can be difficult for the particular sounds made by the electronic musician to be located within an overall soundfield. This problem extends right from the audience, who sometimes complain that they can’t tell what the electronic musician is contributing, through the rest of the band members, who cannot pin down something to musically respond to, right through to the electronic musician themself, who may not be aware of exactly which sounds they are creating.

The solution I have settled on to this problem is to provide a pair of filters with very high Q – one highpass and one lowpass – and to have these tied to two controllers on the MIDI keyboard, physically adjacent to one another. This enables me to readily restrict my sound to a narrow spectral range, one that can complement the spectral ranges of the other musicians with whom I am playing. This is further augmented with a high-level textural control mechanism, enabling me to modulate from sonically smooth to sonically rough material through another series of controllers, again in order to better integrate with what is going on with the acoustic instruments.

3. Physical Communication

Physical communication means using body language to communicate musical intent and musical narrative development to other musicians and to an audience. This is a particular problem for computer-based musicians, since it is so tempting to make use of the computer screen to provide a visual feedback mechanism for the instrument, perhaps in place the physical/motor feedback offered by a traditional acoustic instrument. However, using this kind of feedback both ties up the performer’s eyes and can lead to a kind of rigid body posture, seriously hindering any communication they might have through body language.

This body language is important. In 1753, C.P.E. Bach wrote, in Essay on the True Art of Playing Keyboard Instruments:

A musician cannot move others unless he too is moved. He must of necessity feel all the ‘affects’ he hopes to arouse in his audience, for the revealing of his own humour will stimulate a like humour in the listener … Those who maintain that all of this can be accomplished without gesture will retract their words when, owing to their own insensibility, they find themselves obliged to sit like a statue before their instrument. Ugly grimaces are, of course, inappropriate and harmful; but fitting expressions help the listener to understand our meaning.

last updated Thursday 19th June 2008