def get_cpressed_btchroma(path, compression=1):
""" to easily play with the btchromas we get """
# MAIN FUNCTION WITH COMPRESSION / STRETCH
add_loudness = False
h5 = BAF.GETTERS.open_h5_file_read(path)
chromas = BAF.GETTERS.get_segments_pitches(h5)
segstarts = BAF.GETTERS.get_segments_start(h5)
if compression >= 1:
btstarts = BAF.GETTERS.get_beats_start(h5)[::compression]
elif compression == .5:
btstarts = BAF.GETTERS.get_beats_start(h5)
interpvals = np.interp(range(1,
len(btstarts) * 2 - 1, 2),
range(0,
len(btstarts) * 2, 2), btstarts)
btstarts = np.array(zip(btstarts, interpvals)).flatten()
else:
print 'COMPRESSION OF %d NOT IMPLEMENTED YET' % compression
raise NotImplementedError
if add_loudness: # loudness specific stuff
btloudmax = BAF.get_btloudnessmax(h5)
loudnessmax = BAF.GETTERS.get_segments_loudness_max(h5)
duration = BAF.GETTERS.get_duration(h5)
h5.close()
segstarts = np.array(segstarts).flatten()
btstarts = np.array(btstarts).flatten()
# add back loudness and aligning
chromas = chromas.T
if add_loudness:
chromas = (chromas) * BAF.idB(loudnessmax)
#pass
btchroma = BAF.align_feats(chromas, segstarts, btstarts, duration)
if btchroma is None:
return None
# can we do something with these?
if add_loudness:
#btchroma = btchroma * btloudmax
# normalize
btchroma = BAF.dB(btchroma + 1e-10)
btchroma -= btchroma.min()
#assert not np.isnan(btchroma).any()
btchroma /= btchroma.max()
#btchroma = renorm_chroma(btchroma) # EXPERIMENTAL
# still not correct!
# Renormalize. Each column max is 1.
if not add_loudness:
maxs = btchroma.max(axis=0)
maxs[np.where(maxs == 0)] = 1.
btchroma = (btchroma / maxs)
return btchroma
def get_cpressed_btchroma(path, compression=1):
""" to easily play with the btchromas we get """
# MAIN FUNCTION WITH COMPRESSION / STRETCH
add_loudness = False
h5 = BAF.GETTERS.open_h5_file_read(path)
chromas = BAF.GETTERS.get_segments_pitches(h5)
segstarts = BAF.GETTERS.get_segments_start(h5)
if compression >= 1:
btstarts = BAF.GETTERS.get_beats_start(h5)[::compression]
elif compression == .5:
btstarts = BAF.GETTERS.get_beats_start(h5)
interpvals = np.interp(range(1,len(btstarts)*2-1,2),
range(0,len(btstarts)*2,2), btstarts)
btstarts = np.array(zip(btstarts,interpvals)).flatten()
else:
print 'COMPRESSION OF %d NOT IMPLEMENTED YET' % compression
raise NotImplementedError
if add_loudness: # loudness specific stuff
btloudmax = BAF.get_btloudnessmax(h5)
loudnessmax = BAF.GETTERS.get_segments_loudness_max(h5)
duration = BAF.GETTERS.get_duration(h5)
h5.close()
segstarts = np.array(segstarts).flatten()
btstarts = np.array(btstarts).flatten()
# add back loudness and aligning
chromas = chromas.T
if add_loudness:
chromas = (chromas) * BAF.idB(loudnessmax)
#pass
btchroma = BAF.align_feats(chromas, segstarts, btstarts, duration)
if btchroma is None:
return None
# can we do something with these?
if add_loudness:
#btchroma = btchroma * btloudmax
# normalize
btchroma = BAF.dB(btchroma+1e-10)
btchroma -= btchroma.min()
#assert not np.isnan(btchroma).any()
btchroma /= btchroma.max()
#btchroma = renorm_chroma(btchroma) # EXPERIMENTAL
# still not correct!
# Renormalize. Each column max is 1.
if not add_loudness:
maxs = btchroma.max(axis=0)
maxs[np.where(maxs == 0)] = 1.
btchroma = (btchroma / maxs)
return btchroma
for filename, start_time, end_time in zip(files, start_times, end_times):
# Load in MSD hdf5 file
h5 = hdf5_getters.open_h5_file_read(to_h5_path(filename))
# Load in beat times from MSD
beats = hdf5_getters.get_beats_start(h5)
# Some files have no EN analysis
if beats.size == 0:
continue
# Get indices which fall within the range of correct alignment
time_mask = np.logical_and(beats > start_time, beats < end_time)
beats = beats[time_mask]
# and beat-synchronous feature matrices, within the time range of correct alignment
chroma = beat_aligned_feats.get_btchromas(h5)[:, time_mask]
timbre = beat_aligned_feats.get_bttimbre(h5)[:, time_mask]
loudness = beat_aligned_feats.get_btloudnessmax(h5)[:, time_mask]
h5.close()
# Stack it
msd_features = np.vstack([chroma, timbre, loudness])
if np.isnan(msd_features).any():
print filename
continue
# Load in pretty midi object
pm = pretty_midi.PrettyMIDI(midi.read_midifile(to_midi_path(filename)))
# Construct piano roll, aligned to the msd beat times
piano_roll = pm.get_piano_roll(times=beats)
# Ignore notes below 36 and above 84
piano_roll = piano_roll[36:84, :]
# Write out
np.save(os.path.join(output_path, filename.replace('.mp3', '-msd.npy')), msd_features)
np.save(os.path.join(output_path, filename.replace('.mp3', '-midi.npy')), piano_roll)