The implementation of pyschoacoustical signal parameters in the wavelet domain
Matt Borland, Stephen Birkett · The Journal of the Acoustical Society of America · 2012
Introducing Wavelet techniques into the psychoacoustical analysis of sound signals provides a powerful alternative to standard Fourier methods. In this paper the reformulation of existing psychoacoustical signal parameters using wavelet methods will be explored. A major motivation for this work is that traditional psychoacoustical signal analysis relies heavily on the Fourier transform to provide a frequency content representation of a time signal, but this frequency domain representation is not always accurate; especially for sounds with impactive components. These impactive events can have a more significant contribution to the calculation of psychoacoustical signal parameters by reformulating existing psychoacoustic parameters in the Wavelet domain that are dependent on the Fourier transform. To provide concrete examples of the results simulated “plucked string” sounds are analyzed with analogous Fourier and Wavelet domain signal parameters to demonstrate the difference in performance achieved using Wavelet methods for sounds which have impactive components.