Gabriel J.L. Beckers <beckers@...>
Dear Praat users,
Does anyone know how a spectrum is calculated from a sound signal in Praat?
I presume a sound signal is not transformed as one big chunk. Or is it? Are
time-frames transformed separately and are the results then averaged? If
so, how long are the frames? Are they windowed?
Thanks in advance,
Gabriel J.L. Beckers
Section Behavioural Biology
Institute of Evolutionary and Ecological Sciences
2300 RA Leiden
Tel.: +31-71-5275015 | Fax.: +31-71-5274900
Paul Boersma <paul.boersma@...>
Gabriel Beckers wrote:
Does anyone know how a spectrum is calculated from a sound signal in Praat?It is. A single Fast Fourier Transform is applied to the samples,
after appending zeroes in order to make the number of samples
a power of two. No time frames, no windowing, no averaging.
Institute of Phonetic Sciences, University of Amsterdam
Herengracht 338, 1016CG Amsterdam, The Netherlands