Python audio: Removing noise from a signal

问题

In the code, first I'm opening wav file called output_test.wav. I then filter the noise from the signal using fftpack.

Problem: I'm trying to convert the filtered signal i.e. filtered_sig array into wav file properly. Currently when i open TestFiltered.wav I get the error: "The item was encoded into a format not supported: 0xc00d5212"

Upon further investigation it seems i'm not filtering noise correctly?

I think the error comes from the last 2 lines:

filteredwrite = np.fft.irfft(filtered_sig, axis=0)
    
wavfile.write('TestFiltered.wav', frame_rate, filteredwrite)

CODE:

import numpy as np

from scipy import fftpack

import pyaudio
import wave

from scipy.io import wavfile

def playback():
    
    CHUNK = 1024
    FORMAT = pyaudio.paInt16
    CHANNELS = 2
    RATE = 44100
    RECORD_SECONDS = 8
    WAVE_OUTPUT_FILENAME = "output.wav"


    filename = 'output_test.wav'

    # Set chunk size of 1024 samples per data frame
    chunk = 1024  

    # Open the sound file 
    wf = wave.open(filename, 'rb')

    frame_rate = wf.getframerate()


    wf_x = wf.readframes(-1)


    signal = np.frombuffer(wf_x, dtype='int16')

    #print("signalxx", signal)
    
    

    return [signal, frame_rate]




time_step = 0.5

# get the data 
data = playback()

sig = data[0]
frame_rate = data[1]

# Return discrete Fourier transform of real or complex sequence
sig_fft = fftpack.fft(sig) # tranform the sin function

# Get Amplitude ?
Amplitude = np.abs(sig_fft) # np.abs() - calculate absolute value from a complex number a + ib

Power = Amplitude**2  # create a power spectrum by power of 2 of amplitude

# Get the (angle) base spectrrum of these transform values i.e. sig_fft 
Angle = np.angle(sig_fft)    # Return the angle of the complex arguement

# For each Amplitude and Power (of each element in the array?) - there is will be a corresponding difference in xxx

# This is will return the sampling frequecy or corresponding frequency of each of the (magnitude) i.e. Power
sample_freq = fftpack.fftfreq(sig.size, d=time_step)

print(Amplitude)
print(sample_freq)


# Because we would like to remove the noise we are concerned with peak freqence that contains the peak amplitude
Amp_Freq = np.array([Amplitude, sample_freq])


# Now we try to find the peak amplitude - so we try to extract
Amp_position = Amp_Freq[0,:].argmax()

peak_freq = Amp_Freq[1, Amp_position]   # find the positions of max value position (Amplitude)

# print the position of max Amplitude
print("--", Amp_position)
# print the frequecies of those max amplitude
print(peak_freq)


high_freq_fft = sig_fft.copy()
# assign all the value the corresponding frequecies larger than the peak frequence - assign em 0 - cancel!! in the array (elements) (?)
high_freq_fft[np.abs(sample_freq) > peak_freq] = 0

print("yes:", high_freq_fft)


# Return discrete inverse Fourier transform of real or complex sequence
filtered_sig = fftpack.ifft(high_freq_fft)

# Using Fast Fourier Transform and inverse Fast Fourier Transform we can remove the noise from the frequency domain (that would be otherwise impossible to do in Time Domain) - done.
print("filtered noise: ", filtered_sig)


print("getiing frame rate $$", frame_rate)


filteredwrite =     np.fft.irfft(filtered_sig, axis=0)

print (filteredwrite)

wavfile.write('TestFiltered.wav', frame_rate, filteredwrite)

Any ideas?

来源：https://stackoverflow.com/questions/65405387/python-audio-removing-noise-from-a-signal

标签

python

scipy

fft

pyaudio

wave