stft#

diffsptk.STFT#: alias of ShortTermFourierTransform

class diffsptk.ShortTermFourierTransform(frame_length, frame_period, fft_length, *, center=True, zmean=False, norm='power', window='blackman', out_format='power', eps=1e-09, relative_floor=None)[source]#

This module is a simple cascade of framing, windowing, and spectrum calculation.

Parameters:

frame_lengthint >= 1 [scalar]: Frame length, \(L\).
frame_peirodint >= 1 [scalar]: Frame period, \(P\).
fft_lengthint >= L [scalar]: Number of FFT bins, \(N\).
centerbool [scalar]: If True, assume that the center of data is the center of frame, otherwise assume that the center of data is the left edge of frame.
zmeanbool [scalar]: If True, perform mean subtraction on each frame.
norm[‘none’, ‘power’, ‘magnitude’]: Normalization type of window.
window[‘blackman’, ‘hamming’, ‘hanning’, ‘bartlett’, ‘trapezoidal’, ‘rectangular’]: Window type.
out_format[‘db’, ‘log-magnitude’, ‘magnitude’, ‘power’, ‘complex’]: Output format.
epsfloat >= 0 [scalar]: A small value added to power spectrum.
relative_floorfloat < 0 [scalar]: Relative floor in decibels.

forward(x)[source]#

Compute short-term Fourier transform.

Parameters:

xTensor [shape=(…, T)]: Waveform.

Returns:

yTensor [shape=(…, T/P, N/2+1)]: Spectrum.

Examples

>>> x = diffsptk.ramp(1, 3)
>>> x
tensor([1., 2., 3.])
>>> stft = diffsptk.STFT(frame_length=3, frame_period=1, fft_length=8)
>>> y = stft(x)
>>> y
tensor([[1.0000, 1.0000, 1.0000, 1.0000, 1.0000],
        [4.0000, 4.0000, 4.0000, 4.0000, 4.0000],
        [9.0000, 9.0000, 9.0000, 9.0000, 9.0000]])