def analysis(inputFile1='../../sounds/violin-B3.wav',
window1='blackman',
M1=1001,
N1=1024,
t1=-100,
minSineDur1=0.05,
nH=60,
minf01=200,
maxf01=300,
f0et1=10,
harmDevSlope1=0.01,
stocf=0.1,
inputFile2='../../sounds/soprano-E4.wav',
window2='blackman',
M2=901,
N2=1024,
t2=-100,
minSineDur2=0.05,
minf02=250,
maxf02=500,
f0et2=10,
harmDevSlope2=0.01):
"""
Analyze two sounds with the harmonic plus stochastic model
inputFile: input sound file (monophonic with sampling rate of 44100)
window: analysis window type (rectangular, hanning, hamming, blackman, blackmanharris)
M: analysis window size
N: fft size (power of two, bigger or equal than M)
t: magnitude threshold of spectral peaks
minSineDur: minimum duration of sinusoidal tracks
nH: maximum number of harmonics
minf0: minimum fundamental frequency in sound
maxf0: maximum fundamental frequency in sound
f0et: maximum error accepted in f0 detection algorithm
harmDevSlope: allowed deviation of harmonic tracks, higher harmonics have higher allowed deviation
stocf: decimation factor used for the stochastic approximation
returns inputFile: input file name; fs: sampling rate of input file,
hfreq, hmag: harmonic frequencies, magnitude; stocEnv: stochastic residual
"""
# size of fft used in synthesis
Ns = 512
# hop size (has to be 1/4 of Ns)
H = 128
# read input sounds
(fs1, x1) = UF.wavread(inputFile1)
(fs2, x2) = UF.wavread(inputFile2)
# compute analysis windows
w1 = get_window(window1, M1)
w2 = get_window(window2, M2)
# compute the harmonic plus stochastic models
hfreq1, hmag1, hphase1, stocEnv1 = HPS.hpsModelAnal(
x1, fs1, w1, N1, H, t1, nH, minf01, maxf01, f0et1, harmDevSlope1,
minSineDur1, Ns, stocf)
hfreq2, hmag2, hphase2, stocEnv2 = HPS.hpsModelAnal(
x2, fs2, w2, N2, H, t2, nH, minf02, maxf02, f0et2, harmDevSlope2,
minSineDur2, Ns, stocf)
return inputFile1, fs1, hfreq1, hmag1, stocEnv1, inputFile2, hfreq2, hmag2, stocEnv2
def main(inputFile='../../sounds/sax-phrase-short.wav', window='blackman', M=601, N=1024, t=-100, minSineDur=0.1, nH=100, minf0=350, maxf0=700, f0et=5, harmDevSlope=0.01, stocf=0.1): """ inputFile: input sound file (monophonic with sampling rate of 44100) window: analysis window type (rectangular, hanning, hamming, blackman, blackmanharris) M: analysis window size; N: fft size (power of two, bigger or equal than M) t: magnitude threshold of spectral peaks; minSineDur: minimum duration of sinusoidal tracks nH: maximum number of harmonics; minf0: minimum fundamental frequency in sound maxf0: maximum fundamental frequency in sound; f0et: maximum error accepted in f0 detection algorithm harmDevSlope: allowed deviation of harmonic tracks, higher harmonics have higher allowed deviation stocf: decimation factor used for the stochastic approximation """ # size of fft used in synthesis Ns = 512 # hop size (has to be 1/4 of Ns) H = 128 # read input sound (fs, x) = UF.wavread(inputFile) # compute analysis window w = get_window(window, M) # compute the harmonic plus stochastic model of the whole sound hfreq, hmag, hphase, stocEnv = HPS.hpsModelAnal(x, fs, w, N, H, t, nH, minf0, maxf0, f0et, harmDevSlope, minSineDur, Ns, stocf) # synthesize a sound from the harmonic plus stochastic representation y, yh, yst = HPS.hpsModelSynth(hfreq, hmag, hphase, stocEnv, Ns, H, fs) # output sound file (monophonic with sampling rate of 44100) outputFileSines = 'output_sounds/' + os.path.basename(inputFile)[:-4] + '_hpsModel_sines.wav' outputFileStochastic = 'output_sounds/' + os.path.basename(inputFile)[:-4] + '_hpsModel_stochastic.wav' outputFile = 'output_sounds/' + os.path.basename(inputFile)[:-4] + '_hpsModel.wav' # write sounds files for harmonics, stochastic, and the sum UF.wavwrite(yh, fs, outputFileSines) UF.wavwrite(yst, fs, outputFileStochastic) UF.wavwrite(y, fs, outputFile) return x, fs, hfreq, stocEnv, y
def analysis(inputFile='../../sounds/sax-phrase-short.wav',
window='blackman',
M=601,
N=1024,
t=-100,
minSineDur=0.1,
nH=100,
minf0=350,
maxf0=700,
f0et=5,
harmDevSlope=0.01,
stocf=0.1):
"""
Analyze a sound with the harmonic plus stochastic model
inputFile: input sound file (monophonic with sampling rate of 44100)
window: analysis window type (rectangular, hanning, hamming, blackman, blackmanharris)
M: analysis window size
N: fft size (power of two, bigger or equal than M)
t: magnitude threshold of spectral peaks
minSineDur: minimum duration of sinusoidal tracks
nH: maximum number of harmonics
minf0: minimum fundamental frequency in sound
maxf0: maximum fundamental frequency in sound
f0et: maximum error accepted in f0 detection algorithm
harmDevSlope: allowed deviation of harmonic tracks, higher harmonics have higher allowed deviation
stocf: decimation factor used for the stochastic approximation
returns inputFile: input file name; fs: sampling rate of input file,
hfreq, hmag: harmonic frequencies, magnitude; mYst: stochastic residual
"""
# size of fft used in synthesis
Ns = 512
# hop size (has to be 1/4 of Ns)
H = 128
# read input sound
(fs, x) = UF.wavread(inputFile)
# compute analysis window
w = get_window(window, M)
# compute the harmonic plus stochastic model of the whole sound
hfreq, hmag, hphase, mYst = HPS.hpsModelAnal(x, fs, w, N, H, t, nH, minf0,
maxf0, f0et, harmDevSlope,
minSineDur, Ns, stocf)
# synthesize the harmonic plus stochastic model without original phases
y, yh, yst = HPS.hpsModelSynth(hfreq, hmag, np.array([]), mYst, Ns, H, fs)
# write output sound
outputFile = 'output_sounds/' + os.path.basename(
inputFile)[:-4] + '_hpsModel.wav'
UF.wavwrite(y, fs, outputFile)
# create figure to plot
plt.figure(figsize=(12, 9))
# frequency range to plot
maxplotfreq = 15000.0
# plot the input sound
plt.subplot(3, 1, 1)
plt.plot(np.arange(x.size) / float(fs), x)
plt.axis([0, x.size / float(fs), min(x), max(x)])
plt.ylabel('amplitude')
plt.xlabel('time (sec)')
plt.title('input sound: x')
# plot spectrogram stochastic compoment
plt.subplot(3, 1, 2)
numFrames = int(mYst[:, 0].size)
sizeEnv = int(mYst[0, :].size)
frmTime = H * np.arange(numFrames) / float(fs)
binFreq = (.5 * fs) * np.arange(sizeEnv * maxplotfreq /
(.5 * fs)) / sizeEnv
plt.pcolormesh(
frmTime, binFreq,
np.transpose(mYst[:, :int(sizeEnv * maxplotfreq / (.5 * fs)) + 1]))
plt.autoscale(tight=True)
# plot harmonic on top of stochastic spectrogram
if (hfreq.shape[1] > 0):
harms = hfreq * np.less(hfreq, maxplotfreq)
harms[harms == 0] = np.nan
numFrames = int(harms[:, 0].size)
frmTime = H * np.arange(numFrames) / float(fs)
plt.plot(frmTime, harms, color='k', ms=3, alpha=1)
plt.xlabel('time (sec)')
plt.ylabel('frequency (Hz)')
plt.autoscale(tight=True)
plt.title('harmonics + stochastic spectrogram')
# plot the output sound
plt.subplot(3, 1, 3)
plt.plot(np.arange(y.size) / float(fs), y)
plt.axis([0, y.size / float(fs), min(y), max(y)])
plt.ylabel('amplitude')
plt.xlabel('time (sec)')
plt.title('output sound: y')
plt.tight_layout()
plt.show(block=False)
return inputFile, fs, hfreq, hmag, mYst
def analysis(inputFile='../../sounds/sax-phrase-short.wav', window='blackman', M=601, N=1024, t=-100,
minSineDur=0.1, nH=100, minf0=350, maxf0=700, f0et=5, harmDevSlope=0.01, stocf=0.1):
"""
Analyze a sound with the harmonic plus stochastic model
inputFile: input sound file (monophonic with sampling rate of 44100)
window: analysis window type (rectangular, hanning, hamming, blackman, blackmanharris)
M: analysis window size
N: fft size (power of two, bigger or equal than M)
t: magnitude threshold of spectral peaks
minSineDur: minimum duration of sinusoidal tracks
nH: maximum number of harmonics
minf0: minimum fundamental frequency in sound
maxf0: maximum fundamental frequency in sound
f0et: maximum error accepted in f0 detection algorithm
harmDevSlope: allowed deviation of harmonic tracks, higher harmonics have higher allowed deviation
stocf: decimation factor used for the stochastic approximation
returns inputFile: input file name; fs: sampling rate of input file,
hfreq, hmag: harmonic frequencies, magnitude; mYst: stochastic residual
"""
# size of fft used in synthesis
Ns = 512
# hop size (has to be 1/4 of Ns)
H = 128
# read input sound
(fs, x) = UF.wavread(inputFile)
# compute analysis window
w = get_window(window, M)
# compute the harmonic plus stochastic model of the whole sound
hfreq, hmag, hphase, mYst = HPS.hpsModelAnal(x, fs, w, N, H, t, nH, minf0, maxf0, f0et, harmDevSlope, minSineDur, Ns, stocf)
# synthesize the harmonic plus stochastic model without original phases
y, yh, yst = HPS.hpsModelSynth(hfreq, hmag, np.array([]), mYst, Ns, H, fs)
# write output sound
outputFile = 'output_sounds/' + os.path.basename(inputFile)[:-4] + '_hpsModel.wav'
UF.wavwrite(y,fs, outputFile)
# create figure to plot
plt.figure(figsize=(12, 9))
# frequency range to plot
maxplotfreq = 15000.0
# plot the input sound
plt.subplot(3,1,1)
plt.plot(np.arange(x.size)/float(fs), x)
plt.axis([0, x.size/float(fs), min(x), max(x)])
plt.ylabel('amplitude')
plt.xlabel('time (sec)')
plt.title('input sound: x')
# plot spectrogram stochastic compoment
plt.subplot(3,1,2)
numFrames = int(mYst[:,0].size)
sizeEnv = int(mYst[0,:].size)
frmTime = H*np.arange(numFrames)/float(fs)
binFreq = (.5*fs)*np.arange(sizeEnv*maxplotfreq/(.5*fs))/sizeEnv
plt.pcolormesh(frmTime, binFreq, np.transpose(mYst[:,:sizeEnv*maxplotfreq/(.5*fs)+1]))
plt.autoscale(tight=True)
# plot harmonic on top of stochastic spectrogram
if (hfreq.shape[1] > 0):
harms = hfreq*np.less(hfreq,maxplotfreq)
harms[harms==0] = np.nan
numFrames = int(harms[:,0].size)
frmTime = H*np.arange(numFrames)/float(fs)
plt.plot(frmTime, harms, color='k', ms=3, alpha=1)
plt.xlabel('time (sec)')
plt.ylabel('frequency (Hz)')
plt.autoscale(tight=True)
plt.title('harmonics + stochastic spectrogram')
# plot the output sound
plt.subplot(3,1,3)
plt.plot(np.arange(y.size)/float(fs), y)
plt.axis([0, y.size/float(fs), min(y), max(y)])
plt.ylabel('amplitude')
plt.xlabel('time (sec)')
plt.title('output sound: y')
plt.tight_layout()
plt.show(block=False)
return inputFile, fs, hfreq, hmag, mYst
maxplotfreq = 5000.0
numFrames = int(mXr[:, 0].size)
frmTime = H * np.arange(numFrames) / float(fs)
binFreq = fs * np.arange(N * maxplotfreq / fs) / N
plt.pcolormesh(frmTime, binFreq, np.transpose(mXr[:, : N * maxplotfreq / fs + 1]))
plt.xlabel("time (sec)")
plt.ylabel("frequency (Hz)")
plt.title("spectrogram of residual")
plt.autoscale(tight=True)
plt.tight_layout()
plt.savefig("cello-phrase-residual.png")
# compute the harmonic plus stochastic model
stocf = 0.5
hfreq, hmag, hphase, mYst = HPS.hpsModelAnal(
x, fs, w, N, H, t, nH, minf0, maxf0, f0et, harmDevSlope, minSineDur, Ns, stocf
)
y, yh, yst = HPS.hpsModelSynth(hfreq, hmag, np.array([]), mYst, Ns, H, fs)
UF.wavwrite(yst, fs, "cello-phrase-stochastic.wav")
UF.wavwrite(yh, fs, "cello-phrase-harmonic.wav")
UF.wavwrite(y, fs, "cello-phrase-synthesis.wav")
# plot the spectrogram of the stochastic component
plt.figure(6, figsize=(16, 4.5))
maxplotfreq = 5000.0
numFrames = int(mYst[:, 0].size)
Nst = 2 * int(mYst[0, :].size)
lastbin = int(Nst * maxplotfreq / fs)
maxplotfreq = fs * (lastbin - 1) / Nst
frmTime = H * np.arange(numFrames) / float(fs)
binFreq = fs * np.arange(lastbin) / Nst
N2=1024 t2=-100 minSineDur2=0.05 minf02=250 maxf02=500 f0et2=10 harmDevSlope2=0.01 Ns = 512 H = 128 (fs1, x1) = UF.wavread(inputFile1) (fs2, x2) = UF.wavread(inputFile2) w1 = get_window(window1, M1) w2 = get_window(window2, M2) hfreq1, hmag1, hphase1, stocEnv1 = HPS.hpsModelAnal(x1, fs1, w1, N1, H, t1, nH, minf01, maxf01, f0et1, harmDevSlope1, minSineDur1, Ns, stocf) hfreq2, hmag2, hphase2, stocEnv2 = HPS.hpsModelAnal(x2, fs2, w2, N2, H, t2, nH, minf02, maxf02, f0et2, harmDevSlope2, minSineDur2, Ns, stocf) hfreqIntp = np.array([0, 0, .1, 0, .9, 1, 1, 1]) hmagIntp = np.array([0, 0, .1, 0, .9, 1, 1, 1]) stocIntp = np.array([0, 0, .1, 0, .9, 1, 1, 1]) yhfreq, yhmag, ystocEnv = HPST.hpsMorph(hfreq1, hmag1, stocEnv1, hfreq2, hmag2, stocEnv2, hfreqIntp, hmagIntp, stocIntp) y, yh, yst = HPS.hpsModelSynth(yhfreq, yhmag, np.array([]), ystocEnv, Ns, H, fs1) UF.wavwrite(y,fs1, 'hps-morph-total.wav') plt.figure(figsize=(12, 9)) # frequency range to plot maxplotfreq = 15000.0
def analysis(inputFile1='../../sounds/violin-B3.wav',
window1='blackman',
M1=1001,
N1=1024,
t1=-100,
minSineDur1=0.05,
nH=60,
minf01=200,
maxf01=300,
f0et1=10,
harmDevSlope1=0.01,
stocf=0.1,
inputFile2='../../sounds/soprano-E4.wav',
window2='blackman',
M2=901,
N2=1024,
t2=-100,
minSineDur2=0.05,
minf02=250,
maxf02=500,
f0et2=10,
harmDevSlope2=0.01):
"""
Analyze two sounds with the harmonic plus stochastic model
inputFile: input sound file (monophonic with sampling rate of 44100)
window: analysis window type (rectangular, hanning, hamming, blackman, blackmanharris)
M: analysis window size
N: fft size (power of two, bigger or equal than M)
t: magnitude threshold of spectral peaks
minSineDur: minimum duration of sinusoidal tracks
nH: maximum number of harmonics
minf0: minimum fundamental frequency in sound
maxf0: maximum fundamental frequency in sound
f0et: maximum error accepted in f0 detection algorithm
harmDevSlope: allowed deviation of harmonic tracks, higher harmonics have higher allowed deviation
stocf: decimation factor used for the stochastic approximation
returns inputFile: input file name; fs: sampling rate of input file,
hfreq, hmag: harmonic frequencies, magnitude; stocEnv: stochastic residual
"""
# size of fft used in synthesis
Ns = 512
# hop size (has to be 1/4 of Ns)
H = 128
# read input sounds
(fs1, x1) = UF.wavread(inputFile1)
(fs2, x2) = UF.wavread(inputFile2)
# compute analysis windows
w1 = get_window(window1, M1)
w2 = get_window(window2, M2)
# compute the harmonic plus stochastic models
hfreq1, hmag1, hphase1, stocEnv1 = HPS.hpsModelAnal(
x1, fs1, w1, N1, H, t1, nH, minf01, maxf01, f0et1, harmDevSlope1,
minSineDur1, Ns, stocf)
hfreq2, hmag2, hphase2, stocEnv2 = HPS.hpsModelAnal(
x2, fs2, w2, N2, H, t2, nH, minf02, maxf02, f0et2, harmDevSlope2,
minSineDur2, Ns, stocf)
# create figure to plot
plt.figure(figsize=(9, 6))
# frequency range to plot
maxplotfreq = 15000.0
# plot spectrogram stochastic component of sound 1
plt.subplot(2, 1, 1)
numFrames = int(stocEnv1[:, 0].size)
sizeEnv = int(stocEnv1[0, :].size)
frmTime = H * np.arange(numFrames) / float(fs1)
binFreq = (.5 * fs1) * np.arange(sizeEnv * maxplotfreq /
(.5 * fs1)) / sizeEnv
plt.pcolormesh(
frmTime, binFreq,
np.transpose(stocEnv1[:, :int(sizeEnv * maxplotfreq / (.5 * fs1)) +
1]))
plt.autoscale(tight=True)
# plot harmonic on top of stochastic spectrogram of sound 1
if (hfreq1.shape[1] > 0):
harms = np.copy(hfreq1)
harms = harms * np.less(harms, maxplotfreq)
harms[harms == 0] = np.nan
numFrames = int(harms[:, 0].size)
frmTime = H * np.arange(numFrames) / float(fs1)
plt.plot(frmTime, harms, color='k', ms=3, alpha=1)
plt.xlabel('time (sec)')
plt.ylabel('frequency (Hz)')
plt.autoscale(tight=True)
plt.title('harmonics + stochastic spectrogram of sound 1')
# plot spectrogram stochastic component of sound 2
plt.subplot(2, 1, 2)
numFrames = int(stocEnv2[:, 0].size)
sizeEnv = int(stocEnv2[0, :].size)
frmTime = H * np.arange(numFrames) / float(fs2)
binFreq = (.5 * fs2) * np.arange(sizeEnv * maxplotfreq /
(.5 * fs2)) / sizeEnv
plt.pcolormesh(
frmTime, binFreq,
np.transpose(stocEnv2[:, :int(sizeEnv * maxplotfreq / (.5 * fs2)) +
1]))
plt.autoscale(tight=True)
# plot harmonic on top of stochastic spectrogram of sound 2
if (hfreq2.shape[1] > 0):
harms = np.copy(hfreq2)
harms = harms * np.less(harms, maxplotfreq)
harms[harms == 0] = np.nan
numFrames = int(harms[:, 0].size)
frmTime = H * np.arange(numFrames) / float(fs2)
plt.plot(frmTime, harms, color='k', ms=3, alpha=1)
plt.xlabel('time (sec)')
plt.ylabel('frequency (Hz)')
plt.autoscale(tight=True)
plt.title('harmonics + stochastic spectrogram of sound 2')
plt.tight_layout()
plt.show(block=False)
return inputFile1, fs1, hfreq1, hmag1, stocEnv1, inputFile2, hfreq2, hmag2, stocEnv2
def analysis(inputFile1='../../sounds/violin-B3.wav', window1='blackman', M1=1001, N1=1024, t1=-100,
minSineDur1=0.05, nH=60, minf01=200, maxf01=300, f0et1=10, harmDevSlope1=0.01, stocf=0.1,
inputFile2='../../sounds/soprano-E4.wav', window2='blackman', M2=901, N2=1024, t2=-100,
minSineDur2=0.05, minf02=250, maxf02=500, f0et2=10, harmDevSlope2=0.01):
"""
Analyze two sounds with the harmonic plus stochastic model
inputFile: input sound file (monophonic with sampling rate of 44100)
window: analysis window type (rectangular, hanning, hamming, blackman, blackmanharris)
M: analysis window size
N: fft size (power of two, bigger or equal than M)
t: magnitude threshold of spectral peaks
minSineDur: minimum duration of sinusoidal tracks
nH: maximum number of harmonics
minf0: minimum fundamental frequency in sound
maxf0: maximum fundamental frequency in sound
f0et: maximum error accepted in f0 detection algorithm
harmDevSlope: allowed deviation of harmonic tracks, higher harmonics have higher allowed deviation
stocf: decimation factor used for the stochastic approximation
returns inputFile: input file name; fs: sampling rate of input file,
hfreq, hmag: harmonic frequencies, magnitude; stocEnv: stochastic residual
"""
# size of fft used in synthesis
Ns = 512
# hop size (has to be 1/4 of Ns)
H = 128
# read input sounds
(fs1, x1) = UF.wavread(inputFile1)
(fs2, x2) = UF.wavread(inputFile2)
# compute analysis windows
w1 = get_window(window1, M1)
w2 = get_window(window2, M2)
# compute the harmonic plus stochastic models
hfreq1, hmag1, hphase1, stocEnv1 = HPS.hpsModelAnal(x1, fs1, w1, N1, H, t1, nH, minf01, maxf01, f0et1, harmDevSlope1, minSineDur1, Ns, stocf)
hfreq2, hmag2, hphase2, stocEnv2 = HPS.hpsModelAnal(x2, fs2, w2, N2, H, t2, nH, minf02, maxf02, f0et2, harmDevSlope2, minSineDur2, Ns, stocf)
# create figure to plot
plt.figure(figsize=(12, 9))
# frequency range to plot
maxplotfreq = 15000.0
# plot spectrogram stochastic component of sound 1
plt.subplot(2,1,1)
numFrames = int(stocEnv1[:,0].size)
sizeEnv = int(stocEnv1[0,:].size)
frmTime = H*np.arange(numFrames)/float(fs1)
binFreq = (.5*fs1)*np.arange(sizeEnv*maxplotfreq/(.5*fs1))/sizeEnv
plt.pcolormesh(frmTime, binFreq, np.transpose(stocEnv1[:,:sizeEnv*maxplotfreq/(.5*fs1)+1]))
plt.autoscale(tight=True)
# plot harmonic on top of stochastic spectrogram of sound 1
if (hfreq1.shape[1] > 0):
harms = np.copy(hfreq1)
harms = harms*np.less(harms,maxplotfreq)
harms[harms==0] = np.nan
numFrames = int(harms[:,0].size)
frmTime = H*np.arange(numFrames)/float(fs1)
plt.plot(frmTime, harms, color='k', ms=3, alpha=1)
plt.xlabel('time (sec)')
plt.ylabel('frequency (Hz)')
plt.autoscale(tight=True)
plt.title('harmonics + stochastic spectrogram of sound 1')
# plot spectrogram stochastic component of sound 2
plt.subplot(2,1,2)
numFrames = int(stocEnv2[:,0].size)
sizeEnv = int(stocEnv2[0,:].size)
frmTime = H*np.arange(numFrames)/float(fs2)
binFreq = (.5*fs2)*np.arange(sizeEnv*maxplotfreq/(.5*fs2))/sizeEnv
plt.pcolormesh(frmTime, binFreq, np.transpose(stocEnv2[:,:sizeEnv*maxplotfreq/(.5*fs2)+1]))
plt.autoscale(tight=True)
# plot harmonic on top of stochastic spectrogram of sound 2
if (hfreq2.shape[1] > 0):
harms = np.copy(hfreq2)
harms = harms*np.less(harms,maxplotfreq)
harms[harms==0] = np.nan
numFrames = int(harms[:,0].size)
frmTime = H*np.arange(numFrames)/float(fs2)
plt.plot(frmTime, harms, color='k', ms=3, alpha=1)
plt.xlabel('time (sec)')
plt.ylabel('frequency (Hz)')
plt.autoscale(tight=True)
plt.title('harmonics + stochastic spectrogram of sound 2')
plt.tight_layout()
plt.show(block=False)
return inputFile1, fs1, hfreq1, hmag1, stocEnv1, inputFile2, hfreq2, hmag2, stocEnv2
os.path.join(os.path.dirname(os.path.realpath(__file__)),
'../../../sounds/sax-phrase-short.wav'))
w = np.blackman(601)
N = 1024
t = -100
nH = 100
minf0 = 350
maxf0 = 700
f0et = 5
minSineDur = .1
harmDevSlope = 0.01
Ns = 512
H = Ns / 4
stocf = .2
hfreq, hmag, hphase, mYst = HPS.hpsModelAnal(x, fs, w, N, H, t, nH, minf0,
maxf0, f0et, harmDevSlope,
minSineDur, Ns, stocf)
y, yh, yst = HPS.hpsModelSynth(hfreq, hmag, hphase, mYst, Ns, H, fs)
maxplotfreq = 10000.0
plt.figure(1, figsize=(9, 7))
plt.subplot(311)
plt.plot(np.arange(x.size) / float(fs), x, 'b')
plt.autoscale(tight=True)
plt.title('x (sax-phrase-short.wav)')
plt.subplot(312)
numFrames = int(mYst[:, 0].size)
sizeEnv = int(mYst[0, :].size)
frmTime = H * np.arange(numFrames) / float(fs)
def main(
inputFile="../../sounds/sax-phrase-short.wav",
window="blackman",
M=601,
N=1024,
t=-100,
minSineDur=0.1,
nH=100,
minf0=350,
maxf0=700,
f0et=5,
harmDevSlope=0.01,
stocf=0.1,
):
"""
inputFile: input sound file (monophonic with sampling rate of 44100)
window: analysis window type (rectangular, hanning, hamming, blackman, blackmanharris)
M: analysis window size; N: fft size (power of two, bigger or equal than M)
t: magnitude threshold of spectral peaks; minSineDur: minimum duration of sinusoidal tracks
nH: maximum number of harmonics; minf0: minimum fundamental frequency in sound
maxf0: maximum fundamental frequency in sound; f0et: maximum error accepted in f0 detection algorithm
harmDevSlope: allowed deviation of harmonic tracks, higher harmonics have higher allowed deviation
stocf: decimation factor used for the stochastic approximation
"""
# size of fft used in synthesis
Ns = 512
# hop size (has to be 1/4 of Ns)
H = 128
# read input sound
(fs, x) = UF.wavread(inputFile)
# compute analysis window
w = get_window(window, M)
# compute the harmonic plus stochastic model of the whole sound
hfreq, hmag, hphase, stocEnv = HPS.hpsModelAnal(
x, fs, w, N, H, t, nH, minf0, maxf0, f0et, harmDevSlope, minSineDur, Ns, stocf
)
# synthesize a sound from the harmonic plus stochastic representation
y, yh, yst = HPS.hpsModelSynth(hfreq, hmag, hphase, stocEnv, Ns, H, fs)
# output sound file (monophonic with sampling rate of 44100)
outputFileSines = "output_sounds/" + os.path.basename(inputFile)[:-4] + "_hpsModel_sines.wav"
outputFileStochastic = "output_sounds/" + os.path.basename(inputFile)[:-4] + "_hpsModel_stochastic.wav"
outputFile = "output_sounds/" + os.path.basename(inputFile)[:-4] + "_hpsModel.wav"
# write sounds files for harmonics, stochastic, and the sum
UF.wavwrite(yh, fs, outputFileSines)
UF.wavwrite(yst, fs, outputFileStochastic)
UF.wavwrite(y, fs, outputFile)
# create figure to plot
plt.figure(figsize=(12, 9))
# frequency range to plot
maxplotfreq = 15000.0
# plot the input sound
plt.subplot(3, 1, 1)
plt.plot(np.arange(x.size) / float(fs), x)
plt.axis([0, x.size / float(fs), min(x), max(x)])
plt.ylabel("amplitude")
plt.xlabel("time (sec)")
plt.title("input sound: x")
# plot spectrogram stochastic component
plt.subplot(3, 1, 2)
numFrames = int(stocEnv[:, 0].size)
sizeEnv = int(stocEnv[0, :].size)
frmTime = H * np.arange(numFrames) / float(fs)
binFreq = (0.5 * fs) * np.arange(sizeEnv * maxplotfreq / (0.5 * fs)) / sizeEnv
plt.pcolormesh(frmTime, binFreq, np.transpose(stocEnv[:, : sizeEnv * maxplotfreq / (0.5 * fs) + 1]))
plt.autoscale(tight=True)
# plot harmonic on top of stochastic spectrogram
if hfreq.shape[1] > 0:
harms = hfreq * np.less(hfreq, maxplotfreq)
harms[harms == 0] = np.nan
numFrames = harms.shape[0]
frmTime = H * np.arange(numFrames) / float(fs)
plt.plot(frmTime, harms, color="k", ms=3, alpha=1)
plt.xlabel("time (sec)")
plt.ylabel("frequency (Hz)")
plt.autoscale(tight=True)
plt.title("harmonics + stochastic spectrogram")
# plot the output sound
plt.subplot(3, 1, 3)
plt.plot(np.arange(y.size) / float(fs), y)
plt.axis([0, y.size / float(fs), min(y), max(y)])
plt.ylabel("amplitude")
plt.xlabel("time (sec)")
plt.title("output sound: y")
plt.tight_layout()
plt.show(block=False)
