Algorithmic Composition Systems Research Articles

Affective algorithmic composition of music: A systematic review

<abstract> <p>Affective music composition systems are known to trigger emotions in humans. However, the design of such systems to stimulate users' emotions continues to be a challenge because, studies that aggregate existing literature in the domain to help advance research and knowledge is limited. This study presents a systematic literature review on affective algorithmic composition systems. Eighteen primary studies were selected from IEEE Xplore, ACM Digital Library, SpringerLink, PubMed, ScienceDirect, and Google Scholar databases following a systematic review protocol. The findings revealed that there is a lack of a unique definition that encapsulates the various types of affective algorithmic composition systems. Accordingly, a unique definition is provided. The findings also show that most affective algorithmic composition systems are designed for games to provide background music. The generative composition method was the most used compositional approach. Overall, there was rather a low amount of research in the domain. Possible reasons for these trends are the lack of a common definition for affective music composition systems and also the lack of detailed documentation of the design, implementation and evaluation of the existing systems.</p> </abstract>

Mono-objective Evolutionary Model for Affective Algorithmic Composition

The Affective Music Composition study indicates that musical features can be associated to emotions. Thus, Affective Algorithmic Composition systems implements such features in order to generate melodies that can express or induce emotions. These systems can be applied to different fields, like health and entertainment. However, the composition of melodies with unlimited duration time and diversity is still an open question. This work aims to identify strategies to perform multiple affective transformations in melodies. Therefore, emotional models and the most implemented musical features in the literature are presented, as well as a evolutionary mono-objective model to an affective transformative system.

Between the Abstract and the Concrete: A Constraint-Based Approach to Navigating Instrumental Space

Abstract This article deals with a way that algorithmic composition systems can be informed by material realities of musical performance. After a general discussion of the relation of abstract algorithms to concrete materiality, the article focuses on the idea of an instrument's space of possibilities. It briefly discusses a number of compositional approaches that seek to derive musical structure from bodily movements and from the physical properties of instruments. The last part describes a new open-source JavaScript library called OboeJS and a Web application based on this library. The system is an experimental exploration of the idea of instrumental space and an attempt to bring together abstract algorithmic processing and the concrete possibilities of a musical instrument. The system implements a flexible constraint-based search algorithm for the generation of oboe fingering sequences. This tool is presented as part of a wider approach to algorithmic composition that aims not to map data output of generative procedures to “sound generators” (e.g., performers, instruments, sound synthesis processes). Instead, I propose to derive structure from the space of possibilities of the instrument itself, which in this case is the oboe.

Towards Machine Musicians Who Have Listened to More Music Than Us

Databases of audio can form the basis for new algorithmic critic systems, applying techniques from the growing field of music information retrieval to meta-creation in algorithmic composition and interactive music systems. In this article, case studies are described where critics are derived from larger audio corpora. In the first scenario, the target music is electronic art music, and two corpuses are used to train model parameters and then compared with each other and against further controls in assessing novel electronic music composed by a separate program. In the second scenario, a “real-world” application is described, where a “jury” of three deliberately and individually biased algorithmic music critics judged the winner of a dubstep remix competition. The third scenario is a live tool for automated in-concert criticism, based on the limited situation of comparing an improvising pianists' playing to that of Keith Jarrett; the technology overlaps that described in the other systems, though now deployed in real time. Alongside description and analysis of these systems, the wider possibilities and implications are discussed.

Investigating Perceived Emotional Correlates of Rhythmic Density in Algorithmic Music Composition

Affective algorithmic composition is a growing field that combines perceptually motivated affective computing strategies with novel music generation. This article presents work toward the development of one application. The long-term goal is to develop a responsive and adaptive system for inducing affect that is both controlled and validated by biophysical measures. Literature documenting perceptual responses to music identifies a variety of musical features and possible affective correlations, but perceptual evaluations of these musical features for the purposes of inclusion in a music generation system are not readily available. A discrete feature, rhythmic density (a function of note duration in each musical bar, regardless of tempo), was selected because it was shown to be well-correlated with affective responses in existing literature. A prototype system was then designed to produce controlled degrees of variation in rhythmic density via a transformative algorithm. A two-stage perceptual evaluation of a stimulus set created by this prototype was then undertaken. First, listener responses from a pairwise scaling experiment were analyzed via Multidimensional Scaling Analysis (MDS). The statistical best-fit solution was rotated such that stimuli with the largest range of variation were placed across the horizontal plane in two dimensions. In this orientation, stimuli with deliberate variation in rhythmic density appeared farther from the source material used to generate them than from stimuli generated by random permutation. Second, the same stimulus set was then evaluated according to the order suggested in the rotated two-dimensional solution in a verbal elicitation experiment. A Verbal Protocol Analysis (VPA) found that listener perception of the stimulus set varied in at least two commonly understood emotional descriptors, which might be considered affective correlates of rhythmic density. Thus, these results further corroborate previous studies wherein musical parameters are monitored for changes in emotional expression and that some similarly parameterized control of perceived emotional content in an affective algorithmic composition system can be achieved and provide a methodology for evaluating and including further possible musical features in such a system. Some suggestions regarding the test procedure and analysis techniques are also documented here.

Investigating affect in algorithmic composition systems

There has been a significant amount of work implementing systems for algorithmic composition with the intention of targeting specific emotional responses in the listener, but a full review of this work is not currently available. This gap creates a shared obstacle to those entering the field. Our aim is thus to give an overview of progress in the area of these affectively driven systems for algorithmic composition. Performative and transformative systems are included and differentiated where appropriate, highlighting the challenges these systems now face if they are to be adapted to, or have already incorporated, some form of affective control. Possible real-time applications for such systems, utilizing affectively driven algorithmic composition and biophysical sensing to monitor and induce affective states in the listener are suggested.

Division- and Addition-Based Models of Rhythm in a Computer-Assisted Composition System

Reading descriptions of different musical traditions or of the practice of different individual composers, it is not uncommon to see a rhythmic idiom described as additive/7 meaning in simple terms that it consists of durations formed by combining, or adding together, the units provided by a rapid underlying pulse (Koetting 1970; Pratt 1987; Clayton 2000). Such a characterization distinguishes the rhythmic idiom from others involving the division of the units of some referential pulse into variable numbers of smaller parts, an important feature of tonal Western classical music (Lerdahl and Jackendoff 1983). My present concern is not the ability of additionor division-based models of rhythm to accurately describe existing musical traditions, but the value of these models as a basis for generating rhythms in an algorithmic-composition system, in which potentially novel rhythmic idioms may be sought. In particular, I will describe a series of algorithms I have used in my own work as a composer, and I will investigate some of the advantages and disadvantages each of these algorithms inherits from the particular model of rhythm on which it is based. The development of these algorithms was shaped by at least two compositional goals to create music for human performance, and to express it using common-practice Western music notation (Byrd 1994). My investigations will be made in the context of the same goals. For the sake of making the distinction between the two types of models clearer, Figure 1 presents simple illustrations of rhythms generated by purely addition-based and purely division-based means. In Figure la, the upper row of the diagram represents a stream of timespans constituting the units of some isochronous pulse. The lower row of this diagram shows these timespans grouped into durations of 3, 5, 3, 2, 5, and 2 units to constitute the resulting rhythm. (These values are labeled in the diagram, but of course we could also count unit pulses to determine them.) Although the resulting rhythm lacks the isochrony of the original unit pulse, we can treat it like another pulse and combine adjacent timespans to form larger durations. In this way, the additive process is capable of forming hierarchies of timespans deeper than the two levels illustrated here.

Making Music with Algorithms: A Case-Study System

19 Computer Music Journal, 23:2, pp. 19–30, Summer 1999 © 1999 Massachusetts Institute of Technology. Musicians, perhaps more so than any other group of artists, have always been quick to embrace technology in all its forms. From early attempts at synthesis with the Telharmonium (Roads 1996) to the latest digital-audio workstations (see, for example, Lehrman 1997; Manning 1993), musicians have looked to science to provide them with new and challenging ways of working. One composition method that has benefited greatly from advances in technology is algorithmic composition. Algorithmic composition is by no means new—formal techniques for melody composition date back to at least 1026, when Guido d’Arezzo proposed a scheme that assigned pitches to each vowel sound in a religious text (Loy 1989)—but the advent of affordable and powerful computing resources has meant that more and more people, armed with no more than a little programming knowledge and some musical ideas, have been able to realize their composition algorithms from the comfort of their own desktops. The use of computers as composition generators was pioneered in the mid-1950s by such people as Lejaren Hiller (Hiller and Isaacson 1959), whose 1956 work, The Illiac Suite for String Quartet, is recognized as being the first computer-composed composition. Iannis Xenakis (Xenakis 1960, 1971), although not the first to publish a computer-composed work, had been composing stochastically generated pieces by hand for some time before the premier of The Illiac Suite. Later, Gottfried Koenig developed the Project 1 and 2 composition programs (Koenig 1970a, b). Computer-based algorithmic composition continues the post-war trend toward formalized and systematic composition methods, such as Arnold Schoenberg’s tone-row and Anton Webern’s serialism methods (Roads 1996). Considered thus, it is apparent that the computer is merely a tool for the realization of abstract design constructs, and is best employed as a labor-saving device to free the composer from performing menial calculations by hand. Using a computer, it is possible to preview and test the musical capabilities of many different algorithmic types and judge whether they have the potential to be developed into full-fledged composition systems. This is also the view subscribed to by the authors. Computers in music, and indeed, the algorithmic composition systems themselves, are best thought of as composition tools rather than one-stop music solutions. It was with this in mind that we developed CAMUS 3D, a composition system of our own design that likens the process of music composition to that of pattern propagation. In this article, we introduce the system and show how two classes of algorithms are used to generate musical data.