df2#

class diffsptk.SecondOrderDigitalFilter(sample_rate, pole_frequency=None, pole_bandwidth=None, zero_frequency=None, zero_bandwidth=None, ir_length=None)[source]#

See this page for details.

Parameters:

sample_rateint >= 1 [scalar]: Sample rate in Hz.
pole_frequencyfloat > 0 [scalar]: Pole frequency in Hz.
pole_bandwidthfloat > 0 [scalar]: Pole bandwidth in Hz.
zero_frequencyfloat > 0 [scalar]: Zero frequency in Hz.
zero_bandwidthfloat > 0 [scalar]: Zero bandwidth in Hz.
ir_lengthint >= 1 [scalar]: Length of impulse response.

forward(x)[source]#

Apply a second order digital filter.

Parameters:

xTensor [shape=(B, 1, T) or (B, T) or (T,)]: Input waveform.

Returns:

yTensor [shape=(B, 1, T) or (B, T) or (T,)]: Filterd waveform.

Examples

>>> x = diffsptk.ramp(4)
>>> df2 = diffsptk.SecondOrderDigitalFilter(16000, 100, 200, 1000, 50, 5)
>>> y = df2(x)
>>> y
tensor([0.0000, 1.0000, 2.0918, 3.4161, 5.1021])