pyACA.ToolResample Namespace Reference

Functions
	ToolResample (x, fs_out, fs_in)
	helper function: downmixes an audio signal into one channel

Function Documentation

ToolResample()

ToolResample	(		x,
			fs_out,
			fs_in )

helper function: downmixes an audio signal into one channel

Parameters

x	array with floating point audio data (dimension samples x channels)
fs_out	target sample rate of audio data
fs_in	input sample rate of audio data

Returns

x_out: resampled audio signal

t_out: corresponding time vector

Definition at line 17 of file ToolResample.py.

def ToolResample(x, fs_out, fs_in):
 
    if fs_out > fs_in:
        omega_cutoff = fs_in / fs_out
    else:
        omega_cutoff = fs_out / fs_in
    
    # compute filter coefficients
    iOrder = 4
    [b, a] = butter(iOrder, 0.9 * omega_cutoff)
 
    # time axes
    t_in = np.arange(len(x)) / float(fs_in)
    t_out = np.arange(np.round(t_in[-1] * fs_out)) / float(fs_out)
    
    if fs_out > fs_in:
        # upsample: interpolate and filter
        
        # this uses the most horrible interpolation possible
        hInterpol = interp1d(t_in, x, kind='linear')
        x_out = hInterpol(t_out)
        
        # low pass filter
        x_out = filtfilt(b, a, x_out)
    else:
        # downsample: filter and interpolate
        
        # low pass filter
        x_out = filtfilt(b, a, x)
        
        # this uses the most horrible interpolation possible
        hInterpol = interp1d(t_in, x_out, kind='linear')
        x_out = hInterpol(t_out)
 
    return x_out, t_out