1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
| # Import des bibliothèques
import matplotlib.pyplot as plot
import numpy as np
from scipy.io import wavfile
from scipy import signal
import time
def sonagram_matplotlib(Signal,Fe,L,Chvt,Pas):
tps=time.time()
for n in range(0,Chvt-3*Pas,Pas):
Signal1=Signal[n:]
spectrum,freqs,t,im = plot.specgram(Signal1,cmap=cmap, Fs=Fe,NFFT=L, noverlap=Chvt, scale='dB')
t=t+n/Fe
if n==0:
Sona_Dict=dict(zip(t,20*np.log10(abs(np.transpose(spectrum)))))
else:
Sona_Dict.update(dict(zip(t,20*np.log10(abs(np.transpose(spectrum))))))
t_r=sorted(Sona_Dict)
Sona=[Sona_Dict[n] for n in t_r]
Sona=np.transpose(Sona)
print(time.time()-tps)
return Sona,freqs,t_r
def sonagram_scipy(Signal,Fe,L,Chvt,Pas):
tps=time.time()
for n in range(0,Chvt-3*Pas,Pas):
Signal1=Signal[n:]
freqs,t,spectrum=signal.spectrogram(Signal1, Fe,noverlap=Chvt, nperseg=L,window='hanning')
t=t+n/Fe
if n==0:
Sona_Dict=dict(zip(t,20*np.log10(abs(np.transpose(spectrum)))))
else:
Sona_Dict.update(dict(zip(t,20*np.log10(abs(np.transpose(spectrum))))))
t_r=sorted(Sona_Dict)
Sona=[Sona_Dict[n] for n in t_r]
Sona=np.transpose(Sona)
print(time.time()-tps)
return Sona,freqs,t_r
def sonagram_numpy(Signal,Fe,L,Chvt,Pas):
tps=time.time()
Npt=np.size(Signal) #nbr de point du signal
Fenetre=np.hanning(L) #Fenetrage ou "Windowing" ici hanning
Sona=[]
t=[]
for n in range(Chvt+1,Npt-L-1,Pas):
SF=Signal1[n:n+L]*Fenetre #Signal "fenetré"
SFfft=np.fft.fft(SF,L)
SFfft_reduit=[20*np.log10(np.abs(SFfft[n])) for n in range(0,int(np.size(SFfft)/2))]
Sona.append(SFfft_reduit)
t.append((n+L/2)/Fe)
freqs=np.fft.fftfreq(L,1/Fe)
freq=[x for x in freqs if x>=0]
Sona= np.transpose(Sona)
print(time.time()-tps)
return Sona,freq,t
#On récupère le signal du fichier wave
Fe1, Signal1 = wavfile.read('cha_rom_aca.wav')
L=4096 #Longueur de la section d'analyse en ms (125ms)
Chvt=int(L/2) #Chevauchement ou "overlapping"
pas=int(0.001*Fe1)#nbr de point corresponant à 1ms
cmap = plot.cm.get_cmap('jet')#couleur du sonagramme
#On calcul les spectrogrammes suivant différentes méthodes
Sona_matplotlib,freqs_matplotlib,t_matplotlib = sonagram_matplotlib(Signal1,Fe1,L,Chvt,pas)
Sona_scipy,freqs_scipy,t_scipy = sonagram_scipy(Signal1,Fe1,L,Chvt,pas)
Sona_numpy,freqs_numpy,t_numpy = sonagram_numpy(Signal1,Fe1,L,Chvt,pas)
#On trace les résultats
plot.subplot(311) #Resultat matplotlib
plot.pcolormesh(t_matplotlib, freqs_matplotlib, Sona_matplotlib,cmap=cmap)
plot.yscale('log')
plot.ylim(bottom=100, top=10000)
plot.xlabel('Time')
plot.ylabel('Frequency')
plot.title('Matplotlib')
plot.colorbar()
plot.subplot(312) #Resultat scipy
plot.pcolormesh(t_scipy, freqs_scipy, Sona_scipy,cmap=cmap)
plot.yscale('log')
plot.ylim(bottom=100, top=10000)
plot.xlabel('Time')
plot.ylabel('Frequency')
plot.title('Scipy')
plot.colorbar()
plot.subplot(313) #Resultat numpy
plot.pcolormesh(t_numpy, freqs_numpy, Sona_numpy,cmap=cmap)
plot.yscale('log')
plot.ylim(bottom=100, top=10000)
plot.xlabel('Time')
plot.ylabel('Frequency')
plot.title('Numpy')
plot.colorbar()
plot.show() |
