def get_chromagram(y, sr, chroma):
"""
returns chromagram
Parameters
----------
y : number > 0 [scalar]
audio
sr: number > 0 [scalar]
target sampling rate
chroma: str
chroma-samplerate-framesize-overlap
Returns
-------
list of chromagrams
"""
params = get_parameters_chroma(chroma)
chroma = params["chroma"]
doce_bins_tuned_chroma = None
if chroma == 'nnls':
doce_bins_tuned_chroma = get_nnls(y, params["sr"], params["fr"], params["off"])
elif chroma == 'cqt':
win = get_window('blackmanharris', params["fr"])
doce_bins_tuned_chroma = chroma_cqt(y=y, sr=params["sr"],
C=None,
hop_length=params["off"],
norm=None,
# threshold=0.0,
window=win,
fmin=110,
n_chroma=12,
n_octaves=4 if params["chroma"] == "cqt" and params["sr"] == 5525 else 5,
bins_per_octave=36)
elif chroma == 'cens':
win = get_window('blackmanharris', params["fr"])
doce_bins_tuned_chroma = chroma_cens(y=y, sr=params["sr"],
C=None,
hop_length=params["off"],
norm=None,
window=win,
fmin=110,
n_chroma=12,
n_octaves=5,
bins_per_octave=36)
elif chroma == 'stft':
win = get_window('blackmanharris', params["fr"])
doce_bins_tuned_chroma = chroma_stft(y=y, sr=params["sr"], hop_length=params["off"], norm=None, window=win,
n_chroma=12)
return doce_bins_tuned_chroma
def chroma_cens(args):
sig = get_sig(args)
fs, nfft, noverlap = unroll_args(args, ['fs', 'nfft', 'noverlap'])
hopsize = nfft - noverlap
return rosaft.chroma_cens(y=sig, sr=fs, hop_length=hopsize)
def extract_features(soundwave,sampling_rate,sound_name="test",feature_list=[]):
"""
extracts features with help of librosa
:param soundwave: extracted soundwave from file
:param sampling_rate: sampling rate
:param feature_list: list of features to compute
:param sound_name: type of sound, i.e. dog
:return: np.array of all features for the soundwave
"""
print("Computing features for ",sound_name)
if len(feature_list)==0:
feature_list=["chroma_stft","chroma_cqt","chroma_cens","melspectrogram",
"mfcc","rmse","spectral_centroid","spectral_bandwidth",
"spectral_contrast","spectral_flatness","spectral_rolloff",
"poly_features","tonnetz","zero_crossing_rate"]
features=[]
#feature_len
#"chroma_stft":12
if "chroma_stft" in feature_list:
features.append(feat.chroma_stft(soundwave, sampling_rate))
#"chroma_cqt":12
if "chroma_cqt" in feature_list:
features.append(feat.chroma_cqt(soundwave, sampling_rate))
#"chroma_cens":12
if "chroma_cens" in feature_list:
features.append(feat.chroma_cens(soundwave, sampling_rate))
#"malspectrogram":128
if "melspectrogram" in feature_list:
features.append(feat.melspectrogram(soundwave, sampling_rate))
#"mfcc":20
if "mfcc" in feature_list:
features.append(feat.mfcc(soundwave, sampling_rate))
#"rmse":1
if "rmse" in feature_list:
features.append(feat.rmse(soundwave))
#"spectral_centroid":1
if "spectral_centroid" in feature_list:
features.append(feat.spectral_centroid(soundwave, sampling_rate))
#"spectral_bandwidth":1
if "spectral_bandwidth" in feature_list:
features.append(feat.spectral_bandwidth(soundwave, sampling_rate))
#"spectral_contrast":7
if "spectral_contrast" in feature_list:
features.append(feat.spectral_contrast(soundwave, sampling_rate))
#"spectral_flatness":1
if "spectral_flatness" in feature_list:
features.append(feat.spectral_flatness(soundwave))
#"spectral_rolloff":1
if "spectral_rolloff" in feature_list:
features.append(feat.spectral_rolloff(soundwave, sampling_rate))
#"poly_features":2
if "poly_features" in feature_list:
features.append(feat.poly_features(soundwave, sampling_rate))
#"tonnetz":6
if "tonnetz" in feature_list:
features.append(feat.tonnetz(soundwave, sampling_rate))
#"zero_crossing_rate":1
if "zero_crossing_rate" in feature_list:
features.append(feat.zero_crossing_rate(soundwave))
return np.concatenate(features)
counter = 0
train_data = np.array([])
for chunk in train_data_reader:
#print(chunk)
chunk1 = np.array(chunk)
for thing in chunk1:
print(counter)
thing1 = np.array(thing)
#print(thing1)
row = np.array([])
cstft = np.mean(lf.chroma_stft(thing1[:-1]).T, axis=0)
row = np.concatenate((row, cstft))
cqt = np.mean(lf.chroma_cqt(thing1[:-1]).T, axis=0)
row = np.concatenate((row, cqt))
sens = np.mean(lf.chroma_cens(thing1[:-1]).T, axis=0)
row = np.concatenate((row, sens))
spcent = np.mean(lf.spectral_centroid(thing1[:-1]).T, axis=0)
row = np.concatenate((row, spcent))
flatness = np.mean(lf.spectral_flatness(thing1[:-1]).T, axis=0)
row = np.concatenate((row, flatness))
rolloff = np.mean(lf.spectral_rolloff(thing1[:-1]).T, axis=0)
row = np.concatenate((row, rolloff))
mspec = np.mean(lf.melspectrogram(thing1[:-1]).T, axis=0)
row = np.concatenate((row, mspec))
mfcc = np.mean(lf.mfcc(thing1[:-1], n_mfcc=30).T, axis=0)
row = np.concatenate((row, mfcc))
tonnetz = np.mean(lf.tonnetz(thing1[:-1]).T, axis=0)
row = np.concatenate((row, tonnetz))
rmse = np.mean(lf.rmse(thing1[:-1]).T, axis=0)
row = np.concatenate((row, rmse))
def plotstft_chroma(audiopath,
binsize=2**10,
plotpath=None,
colormap="jet",
pitch_vector=None):
samples, samplerate = librosa.core.load(audiopath,
sr=16000,
mono=True,
duration=14)
# decimo per evitare robaccia inutile, SOLO PER FIGURA ESPERIMENTO 417!!
print 'Se vuoi decimare (esperimento 417) qui dentro a plotstft_chroma lo puoi fare togliendo un commento qui sotto'
#samples = SS.decimate(samples, 4)
#samplerate = samplerate / 4
#pitch_vector = pitch_vector / 2 # devo fare lo stesso con ivalori di pitch
print 'Decimo per dimezzare altezza img'
samples = SS.decimate(samples, 2)
samplerate = samplerate / 2
#pitch_vector = pitch_vector / 2 # devo fare lo stesso con ivalori di pitch
s = stft(samples, binsize)
sshow, freq = logscale_spec(s, factor=1.0, sr=samplerate)
ims = 20. * np.log10(np.abs(sshow) / 10e-6) # amplitude to decibel
timebins, freqbins = np.shape(ims)
plt.figure(figsize=(10, 3))
plt.imshow(np.transpose(ims),
origin="lower",
aspect="auto",
cmap=colormap,
interpolation="none",
clim=(np.min(ims) + 70, np.max(ims)))
#plt.colorbar()
plt.xlabel("Time [s]", fontsize=24)
plt.ylabel("Frequency [Hz]", fontsize=24)
plt.xlim([0, timebins - 1])
plt.ylim([0, freqbins])
plt.tick_params(axis='both', which='major', labelsize=20)
xlocs = np.float32(np.linspace(0, timebins - 1, 4))
plt.xticks(xlocs, [
"%.0f" % l for l in ((xlocs * len(samples) / timebins) +
(0.5 * binsize)) / samplerate
])
ylocs = np.int16(np.round(np.linspace(0, freqbins - 1, 6)))
plt.yticks(ylocs, ["%.0f" % freq[i] for i in ylocs])
if pitch_vector != None:
plt.plot(pitch_vector, 'r')
if plotpath:
plt.savefig(plotpath, format='pdf', bbox_inches="tight")
else:
plt.show()
plt.clf()
### CHROMA
samples, samplerate = librosa.core.load(
audiopath, sr=16000, mono=True,
duration=14) # reload per annullare processing above
chroma = lbf.chroma_cens(y=samples, sr=samplerate, hop_length=128)
plt.figure(figsize=(10, 3))
lbd.specshow(chroma, y_axis='chroma', x_axis='time', cmap='gray')
plt.xlabel("Time [s]", fontsize=24)
plt.ylabel("Pitch Class", fontsize=24, color='w')
plt.tick_params(axis='y',
which='major',
labelsize=20,
labelcolor='w',
length=6,
width=3)
plt.tick_params(axis='x', which='major', labelsize=20)
plt.savefig("chromagram-" + plotpath, format='pdf', bbox_inches="tight")
