Social top


Musical structure, shifting perspectives and composition

An essay on how choice, time, and geography have influenced personal methodologies


About two and half years ago, I travelled from my home country England to spend nine months composing in Norway, assisted by a small research grant from the Norwegian government. The work was to be carried out at NoTAM (Norwegian network for music, technology and acoustics), which is an institution where anyone living in Norway can spend time working with various computer composition facilities, at no cost to the user, and without being affiliated with an academic institution. NoTAM survives on small financial support, good will, and regular composer visitors. Most people who have heard of NoTAM are familiar with the name through using NoTAM's numerous free sound analysis and transformation programmes. One of the most important aspects of the initial period at NoTAM was the change of scenery (and I am not referring to the Norwegian climate!). The encounter with different computer tools and experiencing other composers' working methods – whether I think them appropriate or not for my own work – has been invaluable. I am quite sure that the learning process for a composer is never ending. The months at NoTAM were a productive period, and the stay in Norway became longer than initially intended. Indeed, I am still here at the moment.

In writing for an issue of eContact featuring woman composers, it seems appropriate to mention my current situation as a female electroacoustic composer in Norway. There are not enough electroacoustic composers of either sex to begin thinking about gender issues. My everyday colleagues are all male, except one female who works mainly with instrumental music, and one vocal-electronics performer who commutes between New York and Oslo. Concerning computer music, we all want the same: more performances of good compositions, more financial support to produce concerts, and better integration of computer music into the education system. My understanding of the historical side of computer music in Norway is easy to summarise, if not completely accurate: some composers experimented with both tape and electronic techniques from 1970's to early 90's, and then decided that pure instrumental music was either more enjoyable, or more successful with the funding bodies. There is now a small body of mainly younger computer music composers. They find role models on international, as opposed to national levels. When I think about Norway's small population (about 4.5 million spread over a 2000km long country) and its historically slow uptake of contemporary art, the body of composers involved in computer music in Norway does not seem so small after all! Our concert and promotion organisation, NICEM (which is the Norwegian section of the ICEM) consists of five active board members, and just over 100 members in total. All work is voluntary, and although there are good and bad years with funding, we do continue to produce serious electroacoustic concerts each year, which are frequented by a significant audience. Nevertheless, I think globalisation has brought general relief amongst our practising computer music composers.


The intention of the following text is to attempt to summarise my current compositional intentions and how they have changed over the past five years. The clearest change, or more accurately 'clarification', has concerned the methods I use to create and control musical structure. Attempts to capture aspects of the physical and natural world, either as main or minor factors in the composition, are common to many works. The trend has been to capture these aspects in both intrinsic and extrinsic materials, most recently with methods that are non-conflicting and integral to both sound and structure in a non-compromising fashion. This trend has been manifest in tape, live and installation works. In the following I have focused on tape works mainly because the ideas are clearer, and easier to illustrate with sound examples.


To identify genuine origins of the ideas described above would be a difficult task. At what level would one begin? If taking a high level view, one could argue that the origin would lie in every waking experience or perception. In taking a lower level view, the discussion would require a detailed analysis of the work of the many composers who have had similar aims. This problem is hardly a justification for avoiding a historical or musicological discussion. However within the current context, in suggesting origins to my present musical output I will take reference from two of my own works composed in 1995.

Puzzle Wood (tape composition, 1995)

In Puzzle Wood,sound sources and transformations were partly chosen for their implication of wind, landscape and energy. This work was an attempt to create a landscape through layering foreground, mid- and background sound materials. Sound example 1 illustrates such material with a lengthy extract taken from Puzzle Wood, 0'00 to 2'50. (The example also presents a drone-like sound continuity. At the time it seemed inappropriate to use sharp articulations and silences because our perception of the world, while eyes and other senses are open, consists of an uninterrupted stream of information). During the composition process the intention of creating a landscape became absorbed, and partly obscured, by concepts concerning the structural grouping of elements for perceptual purposes. The implications of wind, landscape and energy are subsequently manifest mainly as vague extra-musical ideas, and are minor in the structural organisation of materials.

Earth Haze (tape composition, 1995)

In Earth Haze, the intention was also to evoke a natural environment through the selection of sound materials and computer transformations. One of the differences between Earth Haze and Puzzle Wood is that in Earth Haze these intentions were the main point of the composition. Many sound materials were selected to evoke the natural environment through aural recognition of the sounding material. Source sounds were also selected to create connotative contrast. 'Natural' sounds, where only the material is suggested, such as the sound of stones knocking together, are thus relatively abstract, while sounds originating in an 'urban' source (where the complete sounding body and mechanism is suggested, such as dustbin lorries and traffic), provided connotative and extra-musical contrast. Sound example 2 is an extract illustrating this contrast in direct connotation, and via degrees of surrogacy (Earth Haze, 7'30-10'00). Even when a sound has undergone no computer transformation, its interpretation will depend on the listener. This sound example partly contains extreme computer transformations which, to some extent, serve to obscure the clarity of the source. One could argue that in such an example the less ambiguous sound information is too fleeting to have significant effect on the listener. However, over the course of the piece, many such subtle suggestions can build into significant effect.

The structural use of space is designed to reinforce the evocative intentions of the work. The use of close-microphone studio recorded source sounds are intended to connect acoustically or connotatively with the implications of space and distance inherent to environmental source material, and to other materials that have undergone certain transformations. Reverberation effects are not included because of the need to imply and 'expanse' rather than a 'room'. Often, these connections serve to mark expansion or convergence in musical structure. Sound example 3 takes an extract from 3'35 – 4'57 where relatively dry sounds (at 19 seconds and 42 seconds into the example) give way quickly to material containing significant spatial connotations achieved through combining environmental source sounds with extreme computer transformation. This repeated contrast of small and large spatial implication mark a point of arrival in the divergence of sound transformation and musical structure.

In retrospect, whether such connections between sound materials and spatial dispositions are evident enough to define the hierarchical and temporal organisation of structure for the ear, is under question.

Same idea, different approach

Through searching for a method to more clearly and accurately articulate natural phenomena in sound and structure, numerical data proved useful. In two recent works, Viva La Selva and The Utility of Space, compositional structures were based upon natural phenomena by extracting data from real-world processes, or from the pattern a process leads to. Nature’s formations are created by, or are part of, complex processes about which the sciences are only approaching an understanding. In many cases, understanding is aided through numerical modelling techniques, or by accurate observational methods. Numerical models create data in an attempt to simulate some aspects of a process or pattern. This is in contrast to observation, where data are obtained directly from the subject.

If numerical data from either modelling or observational methods are to be used in a composition, there are several questions that need asking: Which aspects of the subject do I want to capture in the music? Which aspects of macro and micro musical structure should be controlled, and what aspects of the data are musically most appropriate? Do I need to re-scale data sets such that the listener can perceive a musical ‘mapping’? Such questions are difficult to answer, and are different in every context. To investigate these simple questions it is necessary to have at hand reasonably complete sets of data describing the pattern or process. Often, data that describes a strong visual trend is not the most appropriate for musical mapping. For example, a numerical description of the shape of a landscape (commonly achieved by manipulating a fractal algorithm) may be less musically appropriate than a description of the process of erosion. In this instance, the description of the pattern has less importance than the description of the process.

Through wanting to avoid appropriating an algorithmic process to create musical materials and structures, I decided to obtain control data from real-world observation. This has so far resulted in the two projects discussed below: one where I collected and analysed my own material (through lack of sufficient existing data), and one where data and a subsequent numerical model were supplied by another source.

Project 1: Viva La Selva

In Viva La Selva the idea was to embody, in the manner described above, the natural environment of a relatively undisturbed rain forest. Different data sets were used to control different aspects of sound materials and their structural arrangement. These data sets were created from four channel sound recordings over a continuous 24-hour period, made on a field trip to a rain forest (in fact 2 hours of recordings were separated by 1 hour of no recording due to technical requirements of changing recorder batteries). Four microphones were positioned in a pyramid, approximately 20 meters apart (the exact geometry was measured using string and a compass). By measuring the differences in sound arrival times between all pairs of microphones, the position of louder sounds, such as those made by monkeys and birds, can be determined with an accuracy that depends on the distance from the array. (For more information about the methodology the reader can refer to Barrett, 2000).

The recordings were analysed to reveal the location of animals and the direction in which they were moving, the temporal placement of sounds, and the long-term energy content in seven one-octave bands frequency bands. In addition, temperature variation, rainfall, and daylight times were noted. The amount of vocalisation activity was noted through a simple aural analysis of animal diversity and sound intensity. Often, interesting temporal information was heard in animal calls not loud enough to allow location identification. In these instances, only the time information was extracted.

Materials from the recordings were used as sound-sources for the composition. One could argue that the best way to embody the chosen source, in intrinsic and extrinsic materials through the long- and short-term structure of the composition, would be to simply edit the 24-hour recording into a manageable duration for normal listening with our normal attention span. (1 hour would probably be the upper limit for home listening because this is the duration of a normal CD). Although this is a possible approach, I would argue that, in this instance, it only captures a glimpse of the aural environment, saying very little about the experience of actually being in the location. To capture all the information our other senses receive, as well as assumptions, surprises and fears the mind invents, requires significant sound transformation and rearrangement. To remain accurate to the concept, the majority of these transformations and rearrangements were governed by information extracted from the environment in question – information obtained by analysing the four channel sound recordings.

Examples of compositional application

(a) The long-term energy distribution in different frequency bands over the 24-hour period was scaled to a 20-minute version. This was used as a guide for the spectral activity throughout the macro structure of the composition. Variations in sound activity and diversity were also included in the structure. As would be expected there were clear changes in activity at dusk and dawn, but also many interesting variations during the day and night, particularly with respect to habits of different animal species. Sound example 4 is an extract from the original recordings from 10'40am (in mono, due to the four channel recordings having little meaning in stereo). This can be compared to sound example 5, which is the opening of Viva La Selva. Although sound materials have undergone transformation, and two hours of material are packed into one minute of music (hence the more varied bird calls and the sound of the light aircraft), the spectral distribution is maintained. Sound examples 6 and 7 below illustrate a similar comparison, but this time taken from the beginning of the night period.

(b) The spatial and temporal locations of prominent callers were used to locate specific sounds in the tape work. As the tape piece was to be in stereo format, the three-dimensional co-ordinates needed to be reproduced in the stereo picture, with the illusion of the third co-ordinate via reverberation, and various filtering and convolution techniques. In many instances I would begin with original animal calls as source-sounds, and then gradually implement increasing surrogacy, or substitute non-animal sound materials. In this way the music maintained the spatial and temporal interactive attributes observed in the recordings. Sound examples 6 and 7 illustrates the increasing surrogacy and sound-source substitution method, where example 6 is the original recording (presented here in mono, original temporal information taken from three interacting “tink frogs”), and example 7 is the composition extract. Sound examples 8 and 9 illustrate how the temporal and spatial organisation of sound materials (human voice) was taken from calls of three interacting howler monkeys (example 8 original is from the original, example 9 from the composition). Finally, example 10 illustrates how this material is developed away from the 'environmental discourse', into a musical discourse.

The purpose of this work was not for its scientific value. Despite adhering to a thorough and accurate scientific data collection method, it is clear that a scientific study would require more data, comparisons between different locations, identification of vocalisers, and an understanding of animal behaviour.

Project 2: The Utility of Space

The stereo tape format used in Viva La Selva restricted the accurate application of interesting spatial data. In a stereo home listening situation (on a reasonable hi-fi), the 'space' in the music would be perceived approximately as had been intended. In concert, the need for interpretation would bring an 'unknown' aspect into the music, which although excellent for many works, in this second project was not appropriate.

In The Utility of Space, modelling, as opposed to observation, was used for obtaining numerical data. As in the first project, the data was used to control sound transformation, but was also used to create thousands of spatial locations and trajectories of sound materials in ambisonics b-format (for information on ambisonics the reader can refer to Malham, 2000). The aim was for spatialisation to describe the natural and physical sources as integral aspects of the musical structure. The final work consists of eight tracks of music, where the ambisonics b-format is decoded into a hexagonal speaker array positioned symmetrically around the audience. Approximately half the material is mixed in a conventional stereo track for normal diffusion, which is performed simultaneously with the hexagonal decoding, over the rest of the speakers in a loudspeaker orchestra.

I wanted to use a process or a pattern found in nature that was clearly evident to our perception, and valid on both macro and micro structural terms. For this purpose I used a self-organised critical system. Self-organised critical systems evolve into complex critical states, after which the size and frequency of 'disturbances' follow a 1/f type distribution (Bak 1996). Such systems can be found everywhere in our natural and social world where periods of equilibrium are punctuated by infinite scales of disturbance. For example by surveying a large sandy cliff face, there are landslides and avalanches of all sizes. In 1996, at the University of Oslo physics department, experiments were carried out on piles of rice, to test the self-organised critical behaviour of avalanches, and the approximation to 1/f pattern. After initially intending on using their observational data, it was clear that the observations were very similar to the results of simple computer simulation. The experiment and simulation are clearly described in Bak (1996). I used a small program operating on a 40x40 grid, onto which virtual grains were randomly dropped and subsequent avalanches of different sizes created.

Musical Application

The computer simulation created a large data set. The drop rate was adjusted until a suitably dynamic set of data was achieved. From a simulation which would result in 30 minutes of sound material, one third of the data was selected containing approximately 50,000 time-location entries. 90% decimation was applied, and gave 5,000 time-location entries (considering that input sounds would range in duration from 300ms to 3000ms, it was necessary to apply decimation and thus avoid a too dense build-up of sound).

A 40x40 ‘grid’ of sounds were created by transposing and filtering single input sources between two extremes containing 40 increments, (the x-axis controlling transposition, the y-axis controlling filtering). Each sound has a unique spatial location. Separate sounds, along with their respective time and spatial positions, were selected by the numerical data and mixed into an ambisonics b-format sound file with the software Vspace (Furse, 1999). Filtering and transposing each original sound source 1600 times has the advantage of producing musical variation, while maintaining a ‘frame’ within which the variation can operate. One of the source materials consisted of a short verse from the Tao te Ching, “The Thoughts and Talk of the Wise Man”, translation by Blakeny, 1955: "between the earth and sky, the space is like a bellows, empty but un-spent, when moved, its gift is copious". This was broken into 40 units, and used as 40 input sounds to create 1600 output sounds. Sound example 11 is an extract from the system using the vocal input source, taken from towards the beginning of the process. Example 12 is taken from further into the process where as larger range of 'avalanche' lengths are being created as the system builds to its critical point. (Sound examples are taken from a stereo version of the work, which is slightly shorter than the eight channel version due to the need for structural compressions, arising from the lack of complete three-dimensional trajectories. The stereo version is 13'24 long, the eight channel version 14'50).

This process was repeated with different sounds as input sources to the 40x40 sound grid, and in some instances, 40 input sounds were used. The results were decoded from b-format to the hexagonal loudspeaker array, and mixed with conventional software to give further variation, purely for musical reasons (in other words, at this stage it was necessary to hear, as opposed to guess, how the final work was going to sound!).  Example 13 illustrates a mix between different input sound sources.

After creating a lengthy stretch of music it was clear that something was missing. This missing component consisted of long, single sounds. Sound materials used in the algorithm needed to be short to avoid extreme density, and the output of the program consisted of single articulations, short gestures, and sound masses resembling granulation. Long, single sounds, by nature of their duration, have time to unfold their possible source, and present a duration over which the ear can interpret connotations. Such sounds, both of vocal (in keeping with the original text) and of other source, were used as compositional materials in two ways. First, materials were mixed to create a background, and sometimes foreground to the algorithmically generated materials. In these sounds I attempted to present connotations appropriate to the meaning behind the text (such as 'earth', 'sky', 'bellows' and 'copious'). Sound example 14 is taken from 0'56 – 1'25 in the stereo version of the composition. One layer consists of the spatialised material heard in example 11, the other layer of non-spatialised material. Example 15 is an extract from a little further into the piece.

Second, longer sounds were used as material to be spatialised, but instead of using the original self-organised critical system to locate single sounds, variations in lightly turbulent air currents were used to control sound trajectories. These trajectories were musically used to succeed lengthy avalanches, and in effect lead out of the climactic activity that had built up in the critical system. Example 16 is a section of the material spatialised and modulated by data taken from turbulance trajectories. Example 17 illustrates this material mixed into the composition, taken from 10'22 following from a significantly large 'avalanche'.


I have attempted to identify and explain some of the different methods I have used to approach a common subject. By tackling the problem with different methods, I sometimes come closer to the solution. However, this does not mean that previous methods are ignored, only that they are stored, ready to be retrieved when suitable circumstances arise. In any aesthetic process, there is rarely one best method of realising an idea.


Bak, P. 1996. How nature works. New York: Springer-Verlag.

Barrett, N. ICMC2000, proceedings.

Furrose, R. 1999. Vspace software.

Malham, D. 2000.


Social bottom