def get_btchromas_loudness(h5):
"""
Similar to btchroma, but adds the loudness back.
We use the segments_loudness_max
There is no max value constraint, simply no negative values.
"""
# if string, open and get chromas, if h5, get chromas
if type(h5).__name__ == 'str':
h5 = GETTERS.open_h5_file_read(h5)
chromas = GETTERS.get_segments_pitches(h5)
segstarts = GETTERS.get_segments_start(h5)
btstarts = GETTERS.get_beats_start(h5)
duration = GETTERS.get_duration(h5)
loudnessmax = GETTERS.get_segments_loudness_max(h5)
h5.close()
else:
chromas = GETTERS.get_segments_pitches(h5)
segstarts = GETTERS.get_segments_start(h5)
btstarts = GETTERS.get_beats_start(h5)
duration = GETTERS.get_duration(h5)
loudnessmax = GETTERS.get_segments_loudness_max(h5)
# get the series of starts for segments and beats
segstarts = np.array(segstarts).flatten()
btstarts = np.array(btstarts).flatten()
# add back loudness
chromas = chromas.T * idB(loudnessmax)
# aligned features
btchroma = align_feats(chromas, segstarts, btstarts, duration)
if btchroma is None:
return None
# done (no renormalization)
return btchroma
def get_btloudnessmax(h5):
"""
Get beat-aligned loudness max from a song file of the Million Song Dataset
INPUT:
h5 - filename or open h5 file
RETURN:
btloudnessmax - beat-aligned loudness max, one beat per column
or None if something went wrong (e.g. no beats)
"""
# if string, open and get max loudness, if h5, get max loudness
if type(h5).__name__ == 'str':
h5 = GETTERS.open_h5_file_read(h5)
loudnessmax = GETTERS.get_segments_loudness_max(h5)
segstarts = GETTERS.get_segments_start(h5)
btstarts = GETTERS.get_beats_start(h5)
duration = GETTERS.get_duration(h5)
h5.close()
else:
loudnessmax = GETTERS.get_segments_loudness_max(h5)
segstarts = GETTERS.get_segments_start(h5)
btstarts = GETTERS.get_beats_start(h5)
duration = GETTERS.get_duration(h5)
# get the series of starts for segments and beats
# NOTE: MAYBE USELESS?
# result for track: 'TR0002Q11C3FA8332D'
# segstarts.shape = (708,)
# btstarts.shape = (304,)
segstarts = np.array(segstarts).flatten()
btstarts = np.array(btstarts).flatten()
# reverse dB
loudnessmax = idB(loudnessmax)
# aligned features
btloudnessmax = align_feats(loudnessmax.reshape(1,
loudnessmax.shape[0]),
segstarts, btstarts, duration)
if btloudnessmax is None:
return None
# set it back to dB
btloudnessmax = dB(btloudnessmax + 1e-10)
# done (no renormalization)
return btloudnessmax