Package lib :: Module voice_misc

Module voice_misc

Classes
	cont_kernel_c An instance of this class produces a window with a specified number of nodes.

Functions

window(n, k, norm=2)
Make a window of length n with k nodes.

source code

test_cont_kernel()

source code

cont_kernel(n, k, norm=2, kernel=<function window at 0x42add70>)
Make a window with width n, except that it changes continuously when n is a floating point number.

source code

xexp(a)
exp(a), except protected from underflows.

source code

start_end(t, dt, window, len_d)
Find start and end points for a window.

source code

near_win_size(near=None, tol=None, min=None, max=None, real=False)
Find a window size near the specified size where the FFT algorithm is fastest.

source code

frommid_generator(s, e)
Generate numbers in [s,e] starting at the midpoint.

source code

near_win_size_real(near=None, tol=None, min=None, max=None)

source code

test_win_size()

source code

test_print_win()

source code

bandpass_first_order(d, cutoff_hp, cutoff_lp)
A combinatio of a first-order high-pass filter and a low-pass filter.

source code

test_hpf()

source code

fft_freq(n, d=1.0)
Returns an array of indices whose frequencies are between f0 and f1 after you've called fft().

source code

fft_indices(f0, f1, n, d=1.0)
Returns an array of indices whose frequencies are between f0 and f1 after you've called real_fft().

source code

numpy.ndarray

real_fft_freq(n, d=1.0)
Returns the frequencies associated with an index, after you've called real_fft().

source code

real_fft_indices(f0, f1, n, d=1.0)
Returns the indices whose frequencies are between f0 and f1 after you've called real_fft().

source code

dataprep_flat_real(data, dt, edgewidth, pad=None)

source code

dataprep_flat(data, dt, edgewidth, pad=None)

source code

dataprep_flat_generic(data, dt, edgewidth, pad, isreal)
Pad the data, subtract the average, and round the edges of the window.

source code

dataprep2_flat_generic(data, dt, edgewidth, pad, isreal)
Pad the data, subtract the average, and round the edges of the window.

source code

lowpass_sym_butterworth(d, cutoff_freq, order=4)
A time-symmetric filter with a magnitude response that is the same as the Butterworth filter.

source code

lowpass_sym_Gaussian(d, cutoff_freq)
A time-symmetric filter with a Gaussian impulse response.

source code

lowpass_sym_gaussian(d, cutoff_freq)
A time-symmetric filter with a Gaussian impulse response.

source code

hipass_sym_butterworth(d, cutoff_freq, order=4)
A time-symmetric filter with a magnitude response that is the same as the Butterworth filter.

source code

test_lpb()

source code

test_hpb()

source code

pink_noise(n)
Create pink noise where the power spectral density goes as 1/f (except right at f=0).

source code

Variables
	__package__ = `'lib'`

Imports: math, numpy, ortho_poly, die, gpkmisc, GT, hipass_first_order, lopass_first_order

Function Details

window(n, k, norm=2)

source code

Make a window of length n with k nodes. The window is normalized so that numpy.sum(w**norm)==1.

cont_kernel(n, k, norm=2, kernel=<function window at 0x42add70>)

source code

Make a window with width n, except that it changes continuously when n is a floating point number. The resulting window is always an odd length.

start_end(t, dt, window, len_d)

source code

Find start and end points for a window.

Parameters:

t (float) - desired window center (seconds).
window (int) - window width (in samples).
dt (float) - sampling interval (seconds).
len_d (int @return (starting sample, end) where end=sample beyond the last @rtype (int, int)) - length of data array (samples).

bandpass_first_order(d, cutoff_hp, cutoff_lp)

source code

A combinatio of a first-order high-pass filter and a low-pass filter. Cutoff freq is measured in cycles per point.

fft_freq(n, d=1.0)

source code

Returns an array of indices whose frequencies are between f0 and f1 after you've called fft(). This will produce negative values where appropriate.

real_fft_freq(n, d=1.0)

source code

Returns the frequencies associated with an index, after you've called real_fft().

Parameters:

d (float) - The time interval between samples.
n (int) - the size of the transformed array (i.e. frequency domain, complex)

Returns: numpy.ndarray

the frequency associated with each element in a Fourier Transform

lowpass_sym_butterworth(d, cutoff_freq, order=4)

source code

A time-symmetric filter with a magnitude response that is the same as the Butterworth filter. Cutoff freq is measured in cycles per point.

lowpass_sym_Gaussian(d, cutoff_freq)

source code

A time-symmetric filter with a Gaussian impulse response. Cutoff freq is measured in cycles per point.

lowpass_sym_gaussian(d, cutoff_freq)

source code

A time-symmetric filter with a Gaussian impulse response. Cutoff freq is measured in cycles per point.

hipass_sym_butterworth(d, cutoff_freq, order=4)

source code

A time-symmetric filter with a magnitude response that is the same as the Butterworth filter. Cutoff freq is measured in cycles per point.