Time Scale Modification

Time Scale Modification (TSM) refers to the process of speeding up or slowing down a sound without changing the pitch of any tonal components. For example, TSM of speech should sound like the speaker is talking at a slower or faster rate. When a recorded speech sample is simply played faster, the apparent ``head size'' of the speaker shrinks (the so-called ``munchkinization'' effect). This happens because speech spectra have formants which should not be moved when the rate of speech is varied. The average formant spacing in frequency is a measure of the length of the vocal tract; hence, when speech is simply played faster, the average formant spacing decreases, corresponding to a smaller head size. The illusion of size modulation is sometimes a useful effect in itself, e.g., for scaling virtual musical instruments using commuted synthesis [41,230]. However, it is also useful to be able to adjust time scales without this effect.

The dual of time scale modification is frequency shifting. In this case, we wish to shift the spectral content of a signal up or down without altering the timing of sonic events in the time domain. This effect is used, for example, to retune ``bad notes'' in the recording studio. Frequency shifting can be implemented as time-scale modification followed by sampling-rate conversion, or it can be implemented directly.

For a comprehensive tutorial review of TSM and frequency-shifting techniques, see the excellent survey article by Laroche [118]. This chapter discusses selected methods for TSM, with emphasis on frequency-domain techniques.