Granular synthesis

The composer Iannis Xenakis is conunonly cited as one of the mentors of granular synthesis. In the 1950s, Xenakis developed important theoretical writings where he laid down the principles of the technique (Xenakis, 1971). The first computer-based granular synthesis system did not appear, however, until Curtis Roads (1988) and Barry Truax (1988) began to investigate the potentials of the technique systematically in the 1970s, in the USA and Canada respectively. 

In order to gain a better understanding of granular synthesis it is necessary to comprehend the concept of sound grains and their relationship to our auditory capacity. This notion is largely based upon a sound representation method published in 1947, in a paper by the famous British physicist Dennis Gabor (1947). Until then the representation of the inner structure of sounds was chiefly based upon Fourier s analysis technique and Helmholtz s musical acoustics (Helmholtz, 1885). 

Although it is taken for granted now that the components of a sound spectrum vary substantially during the sound emission, these variations were not considered relevant enough to be represented until recently. In his paper, Gabor proposed the basis for a representation method which combines frequency-domain and time-domain information. His point of departure was to acknowledge the fact that the ear has a time threshold for discerning 

As far as the idea of sound grains is concerned, any synthesiser producing rapid sequences of short sounds may be considered as a granular synthesiser. However, there are important issues to consider when designing a granular synthesis instrument. Three general approaches to the technique are presented as follows. 

Granular sampling employs a granulator mechanism that extracts small portions of a sampled sound and applies an envelope to them. The granulator may produce the grains in 

---

Granular Synthesis (Chant, Vosim) [Rodet et al., 1989][Kaegi et al., 1978]... 

Curtis Roads. Asynchronous Granular Synthesis. In Representations of Musical Signals, edited by G. De Poli, A. Piccialli, and C. Roads, pp. 143-185. Cambridge The MIT Press, 1991. 

The generator works by producing a sequence of dampened sinewave bursts. A single note contains a number of these bursts. Each burst has its own local envelope with either a steep or a smooth attack, and an exponential decay curve. The formant is the result of this local envelope. As the duration of each FOF burst lasts for only a few milliseconds, the envelope produces sidebands around the sinewave, as in amplitude modulation. Note that this mechanism resembles the granular synthesis technique, with the difference that the envelope of the FOF grain was especially designed to facilitate the production of formants. 

The sequential approach to granular synthesis works by synthesising sequential grain streams. The length of the grains and the intervals between them are controllable, but the grains must not overlap (Figure 5.3). 

In granular synthesis, sound complexity is achieved by the combination of large amounts of simple grains. The architecture of a granular instrument may therefore be as simple as a single oscillator and a Gaussian-like bell-shaped envelope (Figure 5.8). 

On the accompanying CD-ROM (in the folder chaosynth), there is a demonstration version of Chaosynth, a sequential granular synthesis program for Macintosh and Windows. An introduction to Chaosynth is given in Chapter 8. In this section we will study the mechanism used to control its parameters. 

Pulsar synthesis, as recently proposed by Curtis Roads (2001), shares the same fundamental principles of granular synthesis. The main difference is that the concept of sound grains is replaced by the notion of pulsars. In essence, a pulsar is a sound grain with a variable duty cycle an idea inspired by an analog synthesis effect known as pulse-width-modulation (PWM). More specifically, a pulsar consists of a waveform w (or pulsaret) of duration d followed by a silent portion s (Figure 5.12). The period p of a pulsar is therefore given by the sum of H7 + s, and d corresponds to the duration of its duty cycle. The frequency of a stream of pulsars with period p is calculated as 1/pHz. 

Figure 8.21 Granular synthesis of sounds involves the production of thousands of short sonic particles that are combined to form larger, complex sound events...

Figure 8.22 Most granular synthesis systems use stochastic methods to control the production of the sonic particles, but Chaosynth uses cellular automata. The state of the cellular automata is defined by the values of the variables at each cell. As implemented on a computer, the cells are represented as a grid of tiny rectangles whose values are indicated by different colours...

Roads, C. (1988). Introduction to granular synthesis. Computer Music Journal, 12, 2, 11-13. 

This a demo version of the author s own software for granular synthesis. 

Finally, time-based techniques approach synthesis from a time domain perspective. The parameters of time-based synthesis tend to describe sound evolution and transformation of time-related features e.g. in terms of time lapses. Examples of time modelling techniques include granular synthesis and sequential waveform composition time modelling techniques are discussed in Chapter 5. 

---