Python matplotlib.pyplot.specgram() Examples

def run_phase_reconstruction_example():
    fs, d = fetch_sample_speech_tapestry()
    # actually gives however many components you say! So double what .m file
    # says
    fftsize = 512
    step = 64
    X_s = np.abs(stft(d, fftsize=fftsize, step=step, real=False,
                      compute_onesided=False))
    X_t = iterate_invert_spectrogram(X_s, fftsize, step, verbose=True)

    """
    import matplotlib.pyplot as plt
    plt.specgram(d, cmap="gray")
    plt.savefig("1.png")
    plt.close()
    plt.imshow(X_s, cmap="gray")
    plt.savefig("2.png")
    plt.close()
    """

    wavfile.write("phase_original.wav", fs, soundsc(d))
    wavfile.write("phase_reconstruction.wav", fs, soundsc(X_t)) 

def harmonics():
    synth = WaveSynth()
    freq = 1500
    num_harmonics = 6
    h_all = synth.harmonics(freq, 1, [(n, 1/n) for n in range(1, num_harmonics+1)])
    even_harmonics = [(1, 1)]  # always include fundamental tone harmonic
    even_harmonics.extend([(n, 1/n) for n in range(2, num_harmonics*2, 2)])
    h_even = synth.harmonics(freq, 1, even_harmonics)
    h_odd = synth.harmonics(freq, 1, [(n, 1/n) for n in range(1, num_harmonics*2, 2)])
    h_all.join(h_even).join(h_odd)
    import matplotlib.pyplot as plot
    plot.title("Spectrogram")
    plot.ylabel("Freq")
    plot.xlabel("Time")
    plot.specgram(h_all.get_frame_array(), Fs=synth.samplerate, noverlap=90, cmap=plot.cm.gist_heat)
    plot.show() 

def plotSpectrogram(data, fftWindow, fftSize, Fs):

    if fftSize == None:
        N = len(data)
    else:
        N = fftSize    
    
    if Fs == None:
        Fs = 2
    
    if fftWindow == "rectangular":
        plt.specgram(data, NFFT=N, Fs=Fs, 
        window=lambda data: data*np.ones(len(data)),  noverlap=int(N/10))
    elif fftWindow == "bartlett":
        plt.specgram(data, NFFT=N, Fs=Fs, 
        window=lambda data: data*np.bartlett(len(data)),  noverlap=int(N/10))
    elif args.fftWindow == "blackman":
        plt.specgram(data, NFFT=N, Fs=Fs, 
        window=lambda data: data*np.blackman(len(data)),  noverlap=int(N/10))
    elif fftWindow == "hamming":
        plt.specgram(data, NFFT=N, Fs=Fs, 
        window=lambda data: data*np.hamming(len(data)),  noverlap=int(N/10))
    elif fftWindow == "hanning":
         plt.specgram(data, NFFT=N, Fs=Fs, 
         window=lambda data: data*np.hanning(len(data)),  noverlap=int(N/10))

    plt.show() 

def run_phase_reconstruction_example():
    fs, d = fetch_sample_speech_tapestry()
    # actually gives however many components you say! So double what .m file
    # says
    fftsize = 512
    step = 64
    X_s = np.abs(stft(d, fftsize=fftsize, step=step, real=False,
                      compute_onesided=False))
    X_t = iterate_invert_spectrogram(X_s, fftsize, step, verbose=True)

    """
    import matplotlib.pyplot as plt
    plt.specgram(d, cmap="gray")
    plt.savefig("1.png")
    plt.close()
    plt.imshow(X_s, cmap="gray")
    plt.savefig("2.png")
    plt.close()
    """

    wavfile.write("phase_original.wav", fs, soundsc(d))
    wavfile.write("phase_reconstruction.wav", fs, soundsc(X_t)) 

def run_phase_reconstruction_example():
    fs, d = fetch_sample_speech_tapestry()
    # actually gives however many components you say! So double what .m file
    # says
    fftsize = 512
    step = 64
    X_s = np.abs(stft(d, fftsize=fftsize, step=step, real=False,
                      compute_onesided=False))
    X_t = iterate_invert_spectrogram(X_s, fftsize, step, verbose=True)

    """
    import matplotlib.pyplot as plt
    plt.specgram(d, cmap="gray")
    plt.savefig("1.png")
    plt.close()
    plt.imshow(X_s, cmap="gray")
    plt.savefig("2.png")
    plt.close()
    """

    wavfile.write("phase_original.wav", fs, soundsc(d))
    wavfile.write("phase_reconstruction.wav", fs, soundsc(X_t)) 

def run_phase_reconstruction_example():
    fs, d = fetch_sample_speech_tapestry()
    # actually gives however many components you say! So double what .m file
    # says
    fftsize = 512
    step = 64
    X_s = np.abs(stft(d, fftsize=fftsize, step=step, real=False,
                      compute_onesided=False))
    X_t = iterate_invert_spectrogram(X_s, fftsize, step, verbose=True)

    """
    import matplotlib.pyplot as plt
    plt.specgram(d, cmap="gray")
    plt.savefig("1.png")
    plt.close()
    plt.imshow(X_s, cmap="gray")
    plt.savefig("2.png")
    plt.close()
    """

    wavfile.write("phase_original.wav", fs, soundsc(d))
    wavfile.write("phase_reconstruction.wav", fs, soundsc(X_t)) 

def implot(arr, scale=None, title="", cmap="gray"):
    import matplotlib.pyplot as plt
    if scale is "specgram":
        # plotting part
        mag = 20. * np.log10(np.abs(arr))
        # Transpose so time is X axis, and invert y axis so
        # frequency is low at bottom
        mag = mag.T[::-1, :]
    else:
        mag = arr
    f, ax = plt.subplots()
    ax.matshow(mag, cmap=cmap)
    plt.axis("off")
    x1 = mag.shape[0]
    y1 = mag.shape[1]

    def autoaspect(x_range, y_range):
        """
        The aspect to make a plot square with ax.set_aspect in Matplotlib
        """
        mx = max(x_range, y_range)
        mn = min(x_range, y_range)
        if x_range <= y_range:
            return mx / float(mn)
        else:
            return mn / float(mx)
    asp = autoaspect(x1, y1)
    ax.set_aspect(asp)
    plt.title(title) 

def plot(self, NFFT=512, noverlap=384, **kwargs):
        '''
        Plot the spectrogram of the audio sample. 

        It takes the same keyword arguments as
        `matplotlib.pyplot.specgram <https://matplotlib.org/api/_as_gen/matplotlib.pyplot.specgram.html>`_.
        '''

        import numpy as np
        try:
            import matplotlib.pyplot as plt
        except ImportError:
            print('Warning: matplotlib is required for plotting')
            return

        # Handle single channel case
        if self.data.ndim == 1:
            data = self.data[:,None]

        nchannels = data.shape[1]

        # Try to have a square looking plot
        pcols = int(np.ceil(np.sqrt(nchannels)))
        prows = int(np.ceil(nchannels / pcols))

        for c in range(nchannels):
            plt.specgram(data[:,c], NFFT=NFFT, Fs=self.fs, noverlap=noverlap, **kwargs)
            plt.xlabel('Time [s]')
            plt.ylabel('Frequency [Hz]')
            plt.title('Channel {}'.format(c+1)) 

def implot(arr, scale=None, title="", cmap="gray"):
    import matplotlib.pyplot as plt
    if scale is "specgram":
        # plotting part
        mag = 20. * np.log10(np.abs(arr))
        # Transpose so time is X axis, and invert y axis so
        # frequency is low at bottom
        mag = mag.T[::-1, :]
    else:
        mag = arr
    f, ax = plt.subplots()
    ax.matshow(mag, cmap=cmap)
    plt.axis("off")
    x1 = mag.shape[0]
    y1 = mag.shape[1]

    def autoaspect(x_range, y_range):
        """
        The aspect to make a plot square with ax.set_aspect in Matplotlib
        """
        mx = max(x_range, y_range)
        mn = min(x_range, y_range)
        if x_range <= y_range:
            return mx / float(mn)
        else:
            return mn / float(mx)
    asp = autoaspect(x1, y1)
    ax.set_aspect(asp)
    plt.title(title) 

def graph_spectrogram(wav_file):
    rate, data = get_wav_info(wav_file)
    print (type(data), len(data))
    nfft = 256  # Length of the windowing segments
    fs = 256  # Sampling frequency
    pxx, freqs, bins, im = plt.specgram(data, nfft, fs)
    print ("pxx : ", len(pxx))
    print ("freqs : ", len(freqs))
    print ("bins : ", len(bins))
    # plt.axis('on')
    # plt.show()
    plt.axis('off')
    print (wav_file.split('.wav')[0])
    plt.savefig(wav_file.split('.wav')[0] + '.png',
                dpi=100,  # Dots per inch
                frameon='false',
                aspect='normal',
                bbox_inches='tight',
                pad_inches=0)  # Spectrogram saved as a .png
    try:
      im = Image.open(wav_file.split('.wav')[0] + '.png')
      rgb_im = im.convert('RGB')
      rgb_im.save(wav_file.split('.png')[0] + '.jpg')
    except Exception as e:
      print (e)
    if os.path.exists(wav_file.split('.wav')[0] + '.png'):
        os.system('convert '+(wav_file.split('.wav')[0] + '.png') + ' '+(wav_file.split('.wav')[0] + '.jpg'))
        os.remove(wav_file.split('.wav')[0] + '.png') 

def implot(arr, scale=None, title="", cmap="gray"):
    import matplotlib.pyplot as plt
    if scale is "specgram":
        # plotting part
        mag = 20. * np.log10(np.abs(arr))
        # Transpose so time is X axis, and invert y axis so
        # frequency is low at bottom
        mag = mag.T[::-1, :]
    else:
        mag = arr
    f, ax = plt.subplots()
    ax.matshow(mag, cmap=cmap)
    plt.axis("off")
    x1 = mag.shape[0]
    y1 = mag.shape[1]

    def autoaspect(x_range, y_range):
        """
        The aspect to make a plot square with ax.set_aspect in Matplotlib
        """
        mx = max(x_range, y_range)
        mn = min(x_range, y_range)
        if x_range <= y_range:
            return mx / float(mn)
        else:
            return mn / float(mx)
    asp = autoaspect(x1, y1)
    ax.set_aspect(asp)
    plt.title(title) 

def implot(arr, scale=None, title="", cmap="gray"):
    import matplotlib.pyplot as plt
    if scale is "specgram":
        # plotting part
        mag = 20. * np.log10(np.abs(arr))
        # Transpose so time is X axis, and invert y axis so
        # frequency is low at bottom
        mag = mag.T[::-1, :]
    else:
        mag = arr
    f, ax = plt.subplots()
    ax.matshow(mag, cmap=cmap)
    plt.axis("off")
    x1 = mag.shape[0]
    y1 = mag.shape[1]

    def autoaspect(x_range, y_range):
        """
        The aspect to make a plot square with ax.set_aspect in Matplotlib
        """
        mx = max(x_range, y_range)
        mn = min(x_range, y_range)
        if x_range <= y_range:
            return mx / float(mn)
        else:
            return mn / float(mx)
    asp = autoaspect(x1, y1)
    ax.set_aspect(asp)
    plt.title(title) 

def run_phase_vq_example():
    def _pre(list_of_data):
        # Temporal window setting is crucial! - 512 seems OK for music, 256
        # fruit perhaps due to samplerates
        n_fft = 256
        step = 32
        f_r = np.vstack([np.abs(stft(dd, n_fft, step=step, real=False,
                                compute_onesided=False))
                         for dd in list_of_data])
        return f_r, n_fft, step

    def preprocess_train(list_of_data, random_state):
        f_r, n_fft, step = _pre(list_of_data)
        clusters = copy.deepcopy(f_r)
        return clusters

    def apply_preprocess(list_of_data, clusters):
        f_r, n_fft, step = _pre(list_of_data)
        f_clust = f_r
        # Nondeterministic ?
        memberships, distances = vq(f_clust, clusters)
        vq_r = clusters[memberships]
        d_k = iterate_invert_spectrogram(vq_r, n_fft, step, verbose=True)
        return d_k

    random_state = np.random.RandomState(1999)

    fs, d = fetch_sample_speech_fruit()
    d1 = d[::9]
    d2 = d[7::8][:5]
    # make sure d1 and d2 aren't the same!
    assert [len(di) for di in d1] != [len(di) for di in d2]

    clusters = preprocess_train(d1, random_state)
    fix_d1 = np.concatenate(d1)
    fix_d2 = np.concatenate(d2)
    vq_d2 = apply_preprocess(d2, clusters)

    wavfile.write("phase_train_no_agc.wav", fs, soundsc(fix_d1))
    wavfile.write("phase_vq_test_no_agc.wav", fs, soundsc(vq_d2))

    agc_d1, freq_d1, energy_d1 = time_attack_agc(fix_d1, fs, .5, 5)
    agc_d2, freq_d2, energy_d2 = time_attack_agc(fix_d2, fs, .5, 5)
    agc_vq_d2, freq_vq_d2, energy_vq_d2 = time_attack_agc(vq_d2, fs, .5, 5)

    """
    import matplotlib.pyplot as plt
    plt.specgram(agc_vq_d2, cmap="gray")
    #plt.title("Fake")
    plt.figure()
    plt.specgram(agc_d2, cmap="gray")
    #plt.title("Real")
    plt.show()
    """

    wavfile.write("phase_train_agc.wav", fs, soundsc(agc_d1))
    wavfile.write("phase_test_agc.wav", fs, soundsc(agc_d2))
    wavfile.write("phase_vq_test_agc.wav", fs, soundsc(agc_vq_d2)) 

def fm():
    synth = WaveSynth(samplerate=8000)
    from matplotlib import pyplot as plot
    freq = 2000
    lfo1 = Sine(1, amplitude=0.4, samplerate=synth.samplerate)
    s1 = synth.sine(freq, duration=3, fm_lfo=lfo1)
    plot.title("Spectrogram")
    plot.ylabel("Freq")
    plot.xlabel("Time")
    plot.specgram(s1.get_frame_array(), Fs=synth.samplerate, noverlap=90, cmap=plot.cm.gist_heat)
    plot.show()
    with Output(nchannels=1, mixing="sequential") as out:
        synth = WaveSynth()
        freq = 440
        lfo1 = Linear(5, samplerate=synth.samplerate)
        lfo1 = EnvelopeFilter(lfo1, 1, 0.5, 0.5, 0.5, 1)
        s1 = synth.sine(freq, duration=3, fm_lfo=lfo1)
        s_all = s1.copy()
        out.play_sample(s1)
        lfo1 = Sine(1, amplitude=0.2, samplerate=synth.samplerate)
        s1 = synth.sine(freq, duration=2, fm_lfo=lfo1)
        s_all.join(s1)
        out.play_sample(s1)
        lfo1 = Sine(freq/17, amplitude=0.5, samplerate=synth.samplerate)
        s1 = synth.sine(freq, duration=2, fm_lfo=lfo1)
        s_all.join(s1)
        out.play_sample(s1)
        lfo1 = Sine(freq/6, amplitude=0.5, samplerate=synth.samplerate)
        s1 = synth.sine(freq, duration=2, fm_lfo=lfo1)
        s_all.join(s1)
        out.play_sample(s1)
        lfo1 = Sine(1, amplitude=0.4, samplerate=synth.samplerate)
        s1 = synth.triangle(freq, duration=2, fm_lfo=lfo1)
        s_all.join(s1)
        out.play_sample(s1)
        freq = 440*2
        lfo1 = Sine(freq/80, amplitude=0.4, samplerate=synth.samplerate)
        s1 = synth.triangle(freq, duration=2, fm_lfo=lfo1)
        s_all.join(s1)
        out.play_sample(s1)
        # s_all.write_wav("fmtestall.wav")
        out.wait_all_played() 

def run_phase_vq_example():
    def _pre(list_of_data):
        # Temporal window setting is crucial! - 512 seems OK for music, 256
        # fruit perhaps due to samplerates
        n_fft = 256
        step = 32
        f_r = np.vstack([np.abs(stft(dd, n_fft, step=step, real=False,
                                compute_onesided=False))
                         for dd in list_of_data])
        return f_r, n_fft, step

    def preprocess_train(list_of_data, random_state):
        f_r, n_fft, step = _pre(list_of_data)
        clusters = copy.deepcopy(f_r)
        return clusters

    def apply_preprocess(list_of_data, clusters):
        f_r, n_fft, step = _pre(list_of_data)
        f_clust = f_r
        # Nondeterministic ?
        memberships, distances = vq(f_clust, clusters)
        vq_r = clusters[memberships]
        d_k = iterate_invert_spectrogram(vq_r, n_fft, step, verbose=True)
        return d_k

    random_state = np.random.RandomState(1999)

    fs, d = fetch_sample_speech_fruit()
    d1 = d[::9]
    d2 = d[7::8][:5]
    # make sure d1 and d2 aren't the same!
    assert [len(di) for di in d1] != [len(di) for di in d2]

    clusters = preprocess_train(d1, random_state)
    fix_d1 = np.concatenate(d1)
    fix_d2 = np.concatenate(d2)
    vq_d2 = apply_preprocess(d2, clusters)

    wavfile.write("phase_train_no_agc.wav", fs, soundsc(fix_d1))
    wavfile.write("phase_vq_test_no_agc.wav", fs, soundsc(vq_d2))

    agc_d1, freq_d1, energy_d1 = time_attack_agc(fix_d1, fs, .5, 5)
    agc_d2, freq_d2, energy_d2 = time_attack_agc(fix_d2, fs, .5, 5)
    agc_vq_d2, freq_vq_d2, energy_vq_d2 = time_attack_agc(vq_d2, fs, .5, 5)

    """
    import matplotlib.pyplot as plt
    plt.specgram(agc_vq_d2, cmap="gray")
    #plt.title("Fake")
    plt.figure()
    plt.specgram(agc_d2, cmap="gray")
    #plt.title("Real")
    plt.show()
    """

    wavfile.write("phase_train_agc.wav", fs, soundsc(agc_d1))
    wavfile.write("phase_test_agc.wav", fs, soundsc(agc_d2))
    wavfile.write("phase_vq_test_agc.wav", fs, soundsc(agc_vq_d2)) 

def run_phase_vq_example():
    def _pre(list_of_data):
        # Temporal window setting is crucial! - 512 seems OK for music, 256
        # fruit perhaps due to samplerates
        n_fft = 256
        step = 32
        f_r = np.vstack([np.abs(stft(dd, fftsize=n_fft, step=step, real=False,
                                compute_onesided=False))
                         for dd in list_of_data])
        return f_r, n_fft, step

    def preprocess_train(list_of_data, random_state):
        f_r, n_fft, step = _pre(list_of_data)
        clusters = copy.deepcopy(f_r)
        return clusters

    def apply_preprocess(list_of_data, clusters):
        f_r, n_fft, step = _pre(list_of_data)
        f_clust = f_r
        # Nondeterministic ?
        memberships, distances = vq(f_clust, clusters)
        vq_r = clusters[memberships]
        d_k = iterate_invert_spectrogram(vq_r, n_fft, step, verbose=True)
        return d_k

    random_state = np.random.RandomState(1999)

    fs, d = fetch_sample_speech_fruit()
    d1 = d[::9]
    d2 = d[7::8][:5]
    # make sure d1 and d2 aren't the same!
    assert [len(di) for di in d1] != [len(di) for di in d2]

    clusters = preprocess_train(d1, random_state)
    fix_d1 = np.concatenate(d1)
    fix_d2 = np.concatenate(d2)
    vq_d2 = apply_preprocess(d2, clusters)

    wavfile.write("phase_train_no_agc.wav", fs, soundsc(fix_d1))
    wavfile.write("phase_vq_test_no_agc.wav", fs, soundsc(vq_d2))

    agc_d1, freq_d1, energy_d1 = time_attack_agc(fix_d1, fs, .5, 5)
    agc_d2, freq_d2, energy_d2 = time_attack_agc(fix_d2, fs, .5, 5)
    agc_vq_d2, freq_vq_d2, energy_vq_d2 = time_attack_agc(vq_d2, fs, .5, 5)

    """
    import matplotlib.pyplot as plt
    plt.specgram(agc_vq_d2, cmap="gray")
    #plt.title("Fake")
    plt.figure()
    plt.specgram(agc_d2, cmap="gray")
    #plt.title("Real")
    plt.show()
    """

    wavfile.write("phase_train_agc.wav", fs, soundsc(agc_d1))
    wavfile.write("phase_test_agc.wav", fs, soundsc(agc_d2))
    wavfile.write("phase_vq_test_agc.wav", fs, soundsc(agc_vq_d2))