An ASIC for Digital Additive Sine-Wave Synthesis
Andy D. Houghton, Andy J. Fisher, Thierry F. Malet · Computer Music Journal · 1995
In 1822 Frangois M. C. Fourier described how any signal may be broken down into an infinite series of appropriately weighted pure sine waves (Fourier 1955). Today this notion is central to many signalprocessing applications. It is the aim of the project described in this article to create a flexible and visual tool for analyzing, manipulating, and recreating using Fourier analysis coupled with additive sine-wave synthesis or ASWS. By applying Fourier theory, a sample may separated into a set of sine waves with time-function envelopes, and in this format many aspects of the can be readily visualized and thus intelligently manipulated. Figure 1 illustrates a typical sound generated in this way. This landscape has time, frequency, and amplitude on orthogonal axes, which is a very convenient format for describing a sound, since it bypasses many of the difficulties associated with raw sampled sound. For example, the sound's frequency can be modified quite independently of its time, and vice versa, without resorting to complex resampling algorithms. Filtering the requires nothing more than weighting the enveloped partial components, based on their frequency, and, perhaps, time. While the idea of creating by adding together sine waves is not in itself new, obstacles (encountered particularly in analysis) have hitherto prevented ASWS from providing a generalpurpose method for synthesis. Advancement of analysis techniques, coupled with hardware improvements, however, are starting to make ASWS an attractive alternative to more traditional soundsynthesis approaches (Brown 1991; Gambardella 1979). This article describes some of the design aspects that were considered during the development of the authors' ASWS engine, based on a custom application-specific integrated circuit (ASIC). The ASWS Engine