Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Fix build and librosa numpy dependency errors #1311

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
+13 −14
Open

Fix build and librosa numpy dependency errors #1311

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
369ee0a
update librosa to 0.10.1
evezers Apr 27, 2024
7228bb8
remove unneeded Dora dependency
evezers Apr 27, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
14 changes: 7 additions & 7 deletions lib_v5/spec_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -271,8 +271,8 @@ def wave_to_spectrogram(wave, hop_length, n_fft, mp, band, is_v51_model=False):
wave_left = np.asfortranarray(wave[0])
wave_right = np.asfortranarray(wave[1])

spec_left = librosa.stft(wave_left, n_fft, hop_length=hop_length)
spec_right = librosa.stft(wave_right, n_fft, hop_length=hop_length)
spec_left = librosa.stft(wave_left, n_fft=n_fft, hop_length=hop_length)
spec_right = librosa.stft(wave_right, n_fft=n_fft, hop_length=hop_length)

spec = np.asfortranarray([spec_left, spec_right])

Expand Down Expand Up @@ -337,7 +337,7 @@ def cmb_spectrogram_to_wave(spec_m, mp, extra_bins_h=None, extra_bins=None, is_v
spec_s *= get_lp_filter_mask(spec_s.shape[1], bp['lpf_start'], bp['lpf_stop'])
else:
spec_s = fft_lp_filter(spec_s, bp['lpf_start'], bp['lpf_stop'])
wave = librosa.resample(spectrogram_to_wave(spec_s, bp['hl'], mp, d, is_v51_model), bp['sr'], sr, res_type=wav_resolution)
wave = librosa.resample(spectrogram_to_wave(spec_s, bp['hl'], mp, d, is_v51_model), orig_sr=bp['sr'], target_sr=sr, res_type=wav_resolution)
else: # mid
if is_v51_model:
spec_s *= get_hp_filter_mask(spec_s.shape[1], bp['hpf_start'], bp['hpf_stop'] - 1)
Expand All @@ -347,7 +347,7 @@ def cmb_spectrogram_to_wave(spec_m, mp, extra_bins_h=None, extra_bins=None, is_v
spec_s = fft_lp_filter(spec_s, bp['lpf_start'], bp['lpf_stop'])

wave2 = np.add(wave, spectrogram_to_wave(spec_s, bp['hl'], mp, d, is_v51_model))
wave = librosa.resample(wave2, bp['sr'], sr, res_type=wav_resolution)
wave = librosa.resample(wave2, orig_sr=bp['sr'], target_sr=sr, res_type=wav_resolution)

return wave

Expand Down Expand Up @@ -406,8 +406,8 @@ def wave_to_spectrogram_old(wave, hop_length, n_fft):
wave_left = np.asfortranarray(wave[0])
wave_right = np.asfortranarray(wave[1])

spec_left = librosa.stft(wave_left, n_fft, hop_length=hop_length)
spec_right = librosa.stft(wave_right, n_fft, hop_length=hop_length)
spec_left = librosa.stft(wave_left, n_fft=n_fft, hop_length=hop_length)
spec_right = librosa.stft(wave_right, n_fft=n_fft, hop_length=hop_length)

spec = np.asfortranarray([spec_left, spec_right])

Expand Down Expand Up @@ -716,7 +716,7 @@ def change_pitch_semitones(y, sr, semitone_shift):
factor = 2 ** (semitone_shift / 12) # Convert semitone shift to factor for resampling
y_pitch_tuned = []
for y_channel in y:
y_pitch_tuned.append(librosa.resample(y_channel, sr, sr*factor, res_type=wav_resolution_float_resampling))
y_pitch_tuned.append(librosa.resample(y_channel, orig_sr=sr, target_sr=sr*factor, res_type=wav_resolution_float_resampling))
y_pitch_tuned = np.array(y_pitch_tuned)
new_sr = sr * factor
return y_pitch_tuned, new_sr
Expand Down
7 changes: 3 additions & 4 deletions requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@ einops==0.6.0
future==0.18.3
julius==0.2.7
kthread==0.2.3
librosa==0.9.2
librosa==0.10.1
llvmlite
matchering==2.0.6
ml_collections==0.1.1
Expand Down Expand Up @@ -36,7 +36,6 @@ onnx
onnxruntime
onnxruntime-gpu
onnx2pytorch
SoundFile==0.11.0; sys_platform != 'darwin'
soundfile==0.12.1; sys_platform != 'darwin'
PySoundFile==0.9.0.post1; sys_platform == 'darwin'
Dora==0.0.3
numpy==1.23.5
numpy==1.24.2
6 changes: 3 additions & 3 deletions separate.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1104,7 +1104,7 @@ def loading_mix(self):
wav_resolution = bp['res_type']

if d == bands_n: # high-end band
X_wave[d], _ = librosa.load(audio_file, bp['sr'], False, dtype=np.float32, res_type=wav_resolution)
X_wave[d], _ = librosa.load(audio_file, sr=bp['sr'], mono=False, dtype=np.float32, res_type=wav_resolution)
X_spec_s[d] = spec_utils.wave_to_spectrogram(X_wave[d], bp['hl'], bp['n_fft'], self.mp, band=d, is_v51_model=self.is_vr_51_model)

if not np.any(X_wave[d]) and is_mp3:
Expand All @@ -1113,7 +1113,7 @@ def loading_mix(self):
if X_wave[d].ndim == 1:
X_wave[d] = np.asarray([X_wave[d], X_wave[d]])
else: # lower bands
X_wave[d] = librosa.resample(X_wave[d+1], self.mp.param['band'][d+1]['sr'], bp['sr'], res_type=wav_resolution)
X_wave[d] = librosa.resample(X_wave[d+1], orig_sr=self.mp.param['band'][d+1]['sr'], target_sr=bp['sr'], res_type=wav_resolution)
X_spec_s[d] = spec_utils.wave_to_spectrogram(X_wave[d], bp['hl'], bp['n_fft'], self.mp, band=d, is_v51_model=self.is_vr_51_model)

if d == bands_n and self.high_end_process != 'none':
Expand Down Expand Up @@ -1444,7 +1444,7 @@ def loading_mix(X, mp):
X_wave[d] = X

else: # lower bands
X_wave[d] = librosa.resample(X_wave[d+1], mp.param['band'][d+1]['sr'], bp['sr'], res_type=wav_resolution)
X_wave[d] = librosa.resample(X_wave[d+1], orig_sr=mp.param['band'][d+1]['sr'], target_sr=bp['sr'], res_type=wav_resolution)

X_spec_s[d] = spec_utils.wave_to_spectrogram(X_wave[d], bp['hl'], bp['n_fft'], mp, band=d, is_v51_model=True)

Expand Down