In lieu of an abstract, here is a brief excerpt of the content:

  • About This Issue

Certain installments of Computer Music Journal are planned as special issues on a given theme, sometimes by guest editors who issue calls for manuscript submissions, and sometimes by conference organizers who solicit extended versions of highly rated conference papers. However, for regular issues, such as the present one, the topic appearing on the cover is instead arrived at by considering the contents of one or more independently submitted articles that have passed peer review and happen to be ready for publication around the same time. This issue’s theme is “Emulative Algorithms and Creative Algorithms.” All but one of the issue’s articles discuss algorithms that either simulate traditional, non-computer-based musical techniques or else explore original, creative approaches. In these pages, the dichotomy of imitative versus original creation, of standard versus nonstandard musical practice, shows up in three areas: sound synthesis, musical composition, and musical performance. For each of these areas, we present an emulative approach first, followed by a non-emulative one. (For the areas of synthesis and performance, the two types of approach are documented by a corresponding pair of articles; but in the case of musical composition, they are both examined within a single article.)

Imitative digital sound synthesis has played a central role in the history of computer music. For example, Max Mathews and his colleagues organized the first digital audio synthesis languages around traditional musical metaphors, composer Jean-Claude Risset’s 1971 An Introductory Catalogue of Computer Synthesized Sounds included some 15 recipes for simulating acoustic musical instruments, and so on. All imitative synthesis algorithms need to approximate the time-varying spectral characteristics of the original sound as it reaches the ear. Physical-modeling algorithms go further and attempt to approximate the way that the vibrations are excited and propagated within the musical instrument itself. Mimicking conventional instrumental timbres can fulfill a commercial demand, of course; but going beyond that prosaic motivation (which was, in any case, less germane in the years before real-time digital synthesis became affordable), researchers harnessing such techniques have pointed out that important discoveries in musical acoustics and sound synthesis have been made by comparing synthetic output to a familiar aural reference point. Indeed, even beyond the realm of sound synthesis, emulative algorithms in general entail analysis and tend to serve goals related to science, engineering, and music theory, whereas creative algorithms, which need not analyze any pre-existing model, tend to serve more purely artistic goals.

The present issue’s first article, by Niklas Lindroos, Henri Penttinen, and Vesa Välimäki, describes a recent example from the lineage of emulative sound synthesis: a new model of the electric guitar. The authors base their work on earlier research in waveguide synthesis of plucked stringed instruments. Digital waveguide synthesis has constituted an important subset of physical modeling synthesis, and the plucked string appeared as an early example of a physically modeled sound source. The technique introduced here includes several novel alterations. The model’s excitation component accounts for the scrape of the plectrum along the string, a variable plectrum angle, and the first reflection of the pluck pulse traveling along the string. The string component is time-varying to permit pitch-bending and to emulate the two-stage decay of an electric guitar tone. Because pickups drastically color the guitar sound, a component that models the magnetic pickup is included; it uses a cascade of all-pass filters to capture the pickup’s inharmonicity.

Front cover. Clockwise front upper left: Mozart’s musical game, K. 516f;a graphical display from Oliver Bown’s live algorithm that uses a decision tree; microtonal music by Harriet Padberg; and a virtual timpanist, from software by Alexandre Bouënard and colleagues.

Back cover. Block diagrams of the electric guitar model by Niklas Lindroos, Henri Penttinen, and Vesa Välimäki. The overall diagram appears at the top. Below it are detailed diagrams of various sections: the plucking-noise and excitation-pulse model, the plucking-point filter, the string model, adn the pickup model.

In this issue’s second article, Luc Döbereiner throws down the gauntlet in defense of “a body of sound-synthesis techniques that is often misrepresented...

pdf

Share