The main theme of this issue of Organised Sound is algorithmic compostion, one which we have featured before, and yet the flavour of the current issue is very different. It perhaps says something of the richness and the variety of the topic that it can elicit such different responses. Perhaps many of us are still not clear exactly what algorithmic composition is, and the vagueness of the definition itself contributes to a broad range of responses.

To begin with, I would like to cite the influence, both direct and indirect, of the following disciplines on my personal: research: philosophical reflection, linguistics and physics. I will do this briefly sketching the fusion of these disciplines, not strictly with my music and the role it plays as the foundation of a personal aesthetic.

This paper outlines some of the technical and aesthetic issues arising from my works which involve computer technology directly in live performance. A process of experimentation in algorithmic composition is described, particularly in relation to aspects of form and ways of automating musical processes at various structural levels. Three works are then described: Kendhang is a work in which a dancer influences the output of a simple algorithmic music system based on interpolation; Or is an instrumental trio which uses realtime sound granulation; and Vug extends this realtime granulation and applies it to a solo clarinet line. An exploration of both the distinctions and similarities between these works is used to develop a general approach to the integration of computer-based algorithmic systems with live performance. This approach emphasises the need for efficient use of technological resources in order to convey expressive content accurately.

This paper describes the ideas and methods that led to the writing of an algorithmic composition NTrope Suite. This piece, for solo recorders and voice, was generated by ‘mixing’ works by different composers from different eras. The idea behind the work was to examine random generation procedures that could maintain stylistic properties typical to the reference works. The interesting property of this method is that it implements a sort of ‘statistical learning’, that optimally preserves the properties of the reference pieces and also properly ‘generalises’ them so as to create a new valid work. The musical result is very coherent, maintaining both stylistic resemblance to the reference music and exhibiting some surprising originality as well. Theoretically, the resulting piece is closest to the reference works in terms of mutual entropy. The algorithm and its theoretical significance are discussed in the paper.

Palatal Sound is a model of vocal tract articulation influenced by physiologic and acoustic analysis of the voice. Specifically, the term articulation refers to all movement within the vocal tract that results in open, filter-like sonorities, as well as in turbulent to absolute airflow modification. This model presents a complete mapping of place within the vocal tract that features flexibility across different vocal tract sizes and proportions. The principles behind this comprehensive mapping of acoustic and physical sound production techniques should not be foreign to those persons who create, combine, design, model or research sound. Therefore, this model might suggest avenues of sound exploration regardless of media or application. This text first presents a brief overview of the current trends of oral modification using vowels, followed by an introduction to and acoustic analyses of the comprehensive vocal tract model as applied to open-like sonorities. This model is then expanded through the presentation of other methods of open-like behaviours. Following the discussion of open sonorities, turbulent-like behaviours are discussed by first identifying the use of language-based fricatives and stops. After this (re-)exposition, the comprehensive model is applied to turbulent structures through examples and acoustic analyses. Finally, these turbulent methods are completed by additional, complementary methods of vocal tract turbulence. The intentions of this paper are: (i) to document this model clearly, (ii) to identify differences between speech and song articulatory behaviour and that of this comprehensive model with the aid of selected acoustic analyses, (iii) to suggest that this model renders valuable scientific information about the limits of vocal tract physiology, and (iv) to propose the practical use of this model by composers and performers.

KYKLOS, an algorithmic composition program, is presented here. It generalises musical scales for use in composition as well as in performance. The sonic output of the system is referred to as polymodal music since it consists of four independent voices playing ‘synthetic modes’. KYKLOS is suitable for computer-assisted composition because it generates MIDI files which can be altered later by the composer. It can equally well be used in live performance for dynamic modification of parameters enabling realtime musical control.

For a system to be able to generate realtime accompaniment to previously unknown songs, it must predict their harmonic development, i.e. the chords to be played. We claim that such a system must combine long-term experience, to identify typical chord sequences (e.g. II–V and II–V–I), with ‘on-the-fly’ adaptation to track-recurrent structures (e.g. choruses and refrains) of the particular song being played. We have implemented a prediction system using a neural network model that encompasses prior knowledge about typical chord sequences. The results achieved are very encouraging, and rather better than those reported in the literature. However, our predictor could not adapt its behaviour to the idiosyncrasies of each song, since online learning is difficult in neural networks. In this paper, we propose an extension to our previous work by the inclusion of a rule-based sequence tracker, which detects recurrent chord sequences while the song is being performed. We show that this hybrid model, which combines a neural network predictor with a rule-based sequence tracker, improves the system's performance.

This paper presents an end-user interface that allows realtime parametric control of sound events resulting in an interactive environment, in which evolutionary computation is applied to algorithmic composition. The resulting system, Vox Populi, uses genetic algorithms to generate and evaluate a sequence of chords played as MIDI data. Harmonic, tonal and voice range fitness are used to control musical features. Based on the ordering of consonance of musical intervals, the notion of approximating a sequence of notes to its harmonically compatible note or tonal centre is used. This method employs fuzzy formalism and is posited as an optimisation approach based on factors relevant to hearing music.

A computer-assisted composition tool for investigating the application of evolutionary techniques in the composition of music is presented. The nature of such an application is examined in terms of defining the possible mechanisms that provide a means for automated creativity. These mechanisms take inspiration from processes found in Darwinian-based evolution theory, genetic algorithm theory and similar aesthetically based uses of a genetic search heuristic in the visual arts. A formal model of ‘musical evolution’ is proposed, with particular emphasis placed on the ways in which a genetic algorithm can be used to effectively manipulate a variety of compositional structures within a hierarchical and generative grammar-based model of musical composition. The result is a prototype Windows MIDI application called GeNotator that allows composers to experiment with a range of musical structures by interactively ‘evolving’ them through a familiar and comprehensive graphical user interface.