48 lines
1.3 KiB
Python
48 lines
1.3 KiB
Python
|
import streamlit as st
|
||
|
|
import numpy as np
|
||
|
|
import matplotlib.pyplot as plt
|
||
|
|
import librosa
|
||
|
|
import soundfile as sf
|
||
|
|
import wave
|
||
|
|
|
||
|
|
|
||
|
|
|
||
|
|
st.set_option('deprecation.showPyplotGlobalUse', False)
|
||
|
|
|
||
|
|
with st.sidebar :
|
||
|
|
p=st.slider("Hearing loss percentage",min_value=0,max_value=100,step=1)
|
||
|
|
p*=0.01
|
||
|
|
filename = "file.wav"
|
||
|
|
f=open(filename,'rb')
|
||
|
|
y,sr = librosa.load(filename)
|
||
|
|
valider = st.button("Valider")
|
||
|
|
|
||
|
|
|
||
|
|
if valider:
|
||
|
|
fft_signal = np.fft.fft(y)
|
||
|
|
amp = np.abs(fft_signal)
|
||
|
|
frequencies = np.fft.fftfreq(len(y), 1/sr)
|
||
|
|
a=np.random.choice(len(fft_signal), size=int(len(fft_signal)*p), replace=False)
|
||
|
|
fft_signal[a]=0
|
||
|
|
amp2 = np.abs(fft_signal)
|
||
|
|
y_filtered = np.fft.ifft(fft_signal).real
|
||
|
|
plt.plot(fft_signal)
|
||
|
|
plt.show()
|
||
|
|
y_norm = librosa.util.normalize(y_filtered)
|
||
|
|
sf.write('output_file.wav',y_norm, sr,'PCM_24')
|
||
|
|
audio_file = open('output_file.wav','rb')
|
||
|
|
st.write("Original file")
|
||
|
|
st.audio(f)
|
||
|
|
st.write("Transformed file")
|
||
|
|
st.audio(audio_file)
|
||
|
|
col1,col2=st.columns(2)
|
||
|
|
with col1:
|
||
|
|
st.header('Original audio')
|
||
|
|
fig, ax = plt.subplots()
|
||
|
|
ax.plot(frequencies,amp)
|
||
|
|
st.pyplot(fig)
|
||
|
|
with col2:
|
||
|
|
st.header('Transformed audio')
|
||
|
|
fig2, ax2 = plt.subplots()
|
||
|
|
ax2.plot(frequencies,amp2)
|
||
|
|
st.pyplot(fig2)
|