DAFX: Digital Audio Effects, Second Edition

6.6 Conclusion

The effects described in this chapter are based on the division of the input sound into short segments. These segments are processed by simple methods such as time scaling by resampling, or amplitude multiplication by an envelope. The segment waveform is not changed, thus maintaining the characteristic of the source signal.

Two categories of effects can be obtained, depending on the strategy used to place the segments in time during the synthesis. If the order and organization of extracted segments are carefully maintained, time stretching or pitch shifting can be performed. Basic methods, SOLA and PSOLA, are presented and their characteristics are discussed. These effects aim to produce sounds that are perceived as similar to the original, but are modified in duration or pitch. As often happens with digital audio effects, the artifacts produced by these methods can be used as a method for deformation of the input sound, whilst maintaining its main characteristics. The low computational complexity of time-segment processing allows efficient real-time applications. Nevertheless, these algorithms produce artifacts that limit their scope of application. More advanced methods for time stretching and pitch shifting will be introduced in Chapters 7–11.

The second category changes the organization and the order of the segments to a great extent, and thus leads to time shuffling and granulation. In this case, the input sound can be much less recognizable in the output. The ...