Skip to content

Commit

Permalink
consolidated spectrogram helper in feature
Browse files Browse the repository at this point in the history
  • Loading branch information
@bmcfee
bmcfee committed Jan 19, 2015
1 parent 4405851 commit 5fcaac3
Show file tree
Hide file tree
Showing 2 changed files with 32 additions and 60 deletions.
88 changes: 30 additions & 58 deletions librosa/feature/spectral.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,19 @@
from ..core.pitch import estimate_tuning


def __get_spec(y=None, sr=22050, S=None, n_fft=2048, hop_length=512, power=1):
'''Helper function to retrieve a magnitude spectrogram.'''

if S is not None:
# Infer n_fft from spectrogram shape
n_fft = 2 * (S.shape[0] - 1)
else:
# Otherwise, compute a magnitude spectrogram from input
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))**power

return S, n_fft


# -- Spectral features -- #
@cache
def spectral_centroid(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
Expand Down Expand Up @@ -80,13 +93,7 @@ def spectral_centroid(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
Instantaneous-frequency spectrogram
'''

# If we don't have a spectrogram, build one
if S is None:
# By default, use a magnitude spectrogram
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))
else:
# Infer n_fft from spectrogram shape
n_fft = (S.shape[0] - 1) * 2
S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length)

if not np.isrealobj(S):
raise ValueError('Spectral centroid is only defined '
Expand Down Expand Up @@ -170,13 +177,8 @@ def spectral_bandwidth(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
bandwidth : np.ndarray [shape=(1, t)]
frequency bandwidth for each frame
'''
# If we don't have a spectrogram, build one
if S is None:
# By default, use a magnitude spectrogram
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))
else:
# Infer n_fft from spectrogram shape
n_fft = (S.shape[0] - 1) * 2

S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length)

if not np.isrealobj(S):
raise ValueError('Spectral bandwidth is only defined '
Expand Down Expand Up @@ -245,13 +247,8 @@ def spectral_contrast(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
each row of spectral contrast values corresponds to a given
octave-based frequency
'''
# If we don't have a spectrogram, build one
if S is None:
# By default, use a magnitude spectrogram
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))
else:
# Infer n_fft from spectrogram shape
n_fft = (S.shape[0] - 1) * 2

S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length)

# Compute the center frequencies of each bin
if freq is None:
Expand Down Expand Up @@ -341,13 +338,7 @@ def spectral_rolloff(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
roll-off frequency for each frame
'''

# If we don't have a spectrogram, build one
if S is None:
# By default, use a magnitude spectrogram
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))
else:
# Infer n_fft from spectrogram shape
n_fft = (S.shape[0] - 1) * 2
S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length)

if not np.isrealobj(S):
raise ValueError('Spectral centroid is only defined '
Expand Down Expand Up @@ -404,10 +395,7 @@ def rms(y=None, S=None, n_fft=2048, hop_length=512):
RMS value for each frame
'''

# If we don't have a spectrogram, build one
if S is None:
# By default, use a magnitude spectrogram
S = stft(y, n_fft=n_fft, hop_length=hop_length)
S, _ = __get_spec(y=y, S=S, n_fft=n_fft, hop_length=hop_length)

return np.sqrt(np.mean(np.abs(S)**2, axis=0, keepdims=True))

Expand Down Expand Up @@ -464,13 +452,8 @@ def poly_features(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
coefficients : np.ndarray [shape=(order+1, t)]
polynomial coefficients for each frame
'''
# If we don't have a spectrogram, build one
if S is None:
# By default, use a magnitude spectrogram
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))
else:
# Infer n_fft from spectrogram shape
n_fft = (S.shape[0] - 1) * 2

S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length)

# Compute the center frequencies of each bin
if freq is None:
Expand Down Expand Up @@ -610,13 +593,8 @@ def logfsgram(y=None, sr=22050, S=None, n_fft=4096, hop_length=512, **kwargs):
`P[f, t]` contains the energy at pitch bin `f`, frame `t`.
'''

# If we don't have a spectrogram, build one
if S is None:
# By default, use a power spectrogram
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))**2

else:
n_fft = (S.shape[0] - 1) * 2
S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length,
power=2)

# If we don't have tuning already, grab it from S
if 'tuning' not in kwargs:
Expand Down Expand Up @@ -700,13 +678,10 @@ def chromagram(y=None, sr=22050, S=None, norm=np.inf, n_fft=2048,
Vector normalization
"""

n_chroma = kwargs.get('n_chroma', 12)
S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length,
power=2)

# Build the power spectrogram if unspecified
if S is None:
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))**2
else:
n_fft = (S.shape[0] - 1) * 2
n_chroma = kwargs.get('n_chroma', 12)

if tuning is None:
tuning = estimate_tuning(S=S, sr=sr, bins_per_octave=n_chroma)
Expand Down Expand Up @@ -826,7 +801,7 @@ def melspectrogram(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
sampling rate of `y`
S : np.ndarray [shape=(d, t)]
magnitude or power spectrogram
power spectrogram
n_fft : int > 0 [scalar]
length of the FFT window
Expand All @@ -853,11 +828,8 @@ def melspectrogram(y=None, sr=22050, S=None, n_fft=2048, hop_length=512,
Short-time Fourier Transform
"""

# Compute the STFT
if S is None:
S = np.abs(stft(y, n_fft=n_fft, hop_length=hop_length))**2
else:
n_fft = 2 * (S.shape[0] - 1)
S, n_fft = __get_spec(y=y, sr=sr, S=S, n_fft=n_fft, hop_length=hop_length,
power=2)

# Build a Mel filter
mel_basis = filters.mel(sr, n_fft, **kwargs)
Expand Down
4 changes: 2 additions & 2 deletions librosa/segment.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@


@cache
def __band_infinite(d, width, v_in=0.0, v_out=np.inf, dtype=np.float32):
def __band_infinite(n, width, v_in=0.0, v_out=np.inf, dtype=np.float32):
'''Construct a square, banded matrix `X` where
`X[i, j] == v_in` if `|i - j| <= width`
`X[i, j] == v_out` if `|i - j| > width`
Expand All @@ -24,7 +24,7 @@ def __band_infinite(d, width, v_in=0.0, v_out=np.inf, dtype=np.float32):
'''

# Instantiate the matrix
band = np.empty((d, d), dtype=dtype)
band = np.empty((n, n), dtype=dtype)

# Fill the out-of-band values
band.fill(v_out)
Expand Down

0 comments on commit 5fcaac3

Please sign in to comment.