OPENGIT
/
Retrieval-based-Voice-Conversion-WebUI
mirror of https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fix: Updated librosa to version 0.10.2 

There is a bug in librosa 0.9.1.
https://github.com/librosa/librosa/pull/1594

As a result, an error occurs when executing the "Vocals/Accompaniment Separation & Reverberation Removal" function.

To address this issue, librosa has been upgraded to version 0.10.2.
Additionally, torchcrepe has been upgraded due to its dependency on librosa.
This commit is contained in:
tkyaji 2024-06-26 21:59:55 +09:00
parent 1f1755fe3d
commit 330bdd9692
7 changed files with 45 additions and 41 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
infer/lib/uvr5_pack/lib_v5/spec_utils.py
View File

@ -43,8 +43,8 @@ def wave_to_spectrogram(
wave_left = np.asfortranarray(wave[0]) wave_left = np.asfortranarray(wave[0])
wave_right = np.asfortranarray(wave[1]) wave_right = np.asfortranarray(wave[1])
spec_left = librosa.stft(wave_left, 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, 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]) spec = np.asfortranarray([spec_left, spec_right])
@ -78,7 +78,7 @@ def wave_to_spectrogram_mt(
kwargs={"y": wave_left, "n_fft": n_fft, "hop_length": hop_length}, kwargs={"y": wave_left, "n_fft": n_fft, "hop_length": hop_length},
) )
thread.start() thread.start()
spec_right = librosa.stft(wave_right, n_fft, hop_length=hop_length) spec_right = librosa.stft(wave_right, n_fft=n_fft, hop_length=hop_length)
thread.join() thread.join()
spec = np.asfortranarray([spec_left, spec_right]) spec = np.asfortranarray([spec_left, spec_right])
@ -230,26 +230,30 @@ def cache_or_load(mix_path, inst_path, mp):
if d == len(mp.param["band"]): # high-end band if d == len(mp.param["band"]): # high-end band
X_wave[d], _ = librosa.load( X_wave[d], _ = librosa.load(
mix_path, bp["sr"], False, dtype=np.float32, res_type=bp["res_type"] mix_path,
sr=bp["sr"],
mono=False,
dtype=np.float32,
res_type=bp["res_type"]
) )
y_wave[d], _ = librosa.load( y_wave[d], _ = librosa.load(
inst_path, inst_path,
bp["sr"], sr=bp["sr"],
False, mono=False,
dtype=np.float32, dtype=np.float32,
res_type=bp["res_type"], res_type=bp["res_type"],
) )
else: # lower bands else: # lower bands
X_wave[d] = librosa.resample( X_wave[d] = librosa.resample(
X_wave[d + 1], X_wave[d + 1],
mp.param["band"][d + 1]["sr"], orig_sr=mp.param["band"][d + 1]["sr"],
bp["sr"], target_sr=bp["sr"],
res_type=bp["res_type"], res_type=bp["res_type"],
) )
y_wave[d] = librosa.resample( y_wave[d] = librosa.resample(
y_wave[d + 1], y_wave[d + 1],
mp.param["band"][d + 1]["sr"], orig_sr=mp.param["band"][d + 1]["sr"],
bp["sr"], target_sr=bp["sr"],
res_type=bp["res_type"], res_type=bp["res_type"],
) )
@ -401,8 +405,8 @@ def cmb_spectrogram_to_wave(spec_m, mp, extra_bins_h=None, extra_bins=None):
mp.param["mid_side_b2"], mp.param["mid_side_b2"],
mp.param["reverse"], mp.param["reverse"],
), ),
bp["sr"], orig_sr=bp["sr"],
sr, target_sr=sr,
res_type="sinc_fastest", res_type="sinc_fastest",
) )
else: # mid else: # mid
@ -419,7 +423,7 @@ def cmb_spectrogram_to_wave(spec_m, mp, extra_bins_h=None, extra_bins=None):
), ),
) )
# wave = librosa.core.resample(wave2, bp['sr'], sr, res_type="sinc_fastest") # wave = librosa.core.resample(wave2, bp['sr'], sr, res_type="sinc_fastest")
wave = librosa.core.resample(wave2, bp["sr"], sr, res_type="scipy") wave = librosa.resample(wave2, orig_sr=bp["sr"], target_sr=sr, res_type="scipy")
return wave.T return wave.T
@ -506,8 +510,8 @@ def ensembling(a, specs):
def stft(wave, nfft, hl): def stft(wave, nfft, hl):
wave_left = np.asfortranarray(wave[0]) wave_left = np.asfortranarray(wave[0])
wave_right = np.asfortranarray(wave[1]) wave_right = np.asfortranarray(wave[1])
spec_left = librosa.stft(wave_left, nfft, hop_length=hl) spec_left = librosa.stft(wave_left, n_fft=nfft, hop_length=hl)
spec_right = librosa.stft(wave_right, nfft, hop_length=hl) spec_right = librosa.stft(wave_right, n_fft=nfft, hop_length=hl)
spec = np.asfortranarray([spec_left, spec_right]) spec = np.asfortranarray([spec_left, spec_right])
return spec return spec
@ -569,8 +573,8 @@ if __name__ == "__main__":
if d == len(mp.param["band"]): # high-end band if d == len(mp.param["band"]): # high-end band
wave[d], _ = librosa.load( wave[d], _ = librosa.load(
args.input[i], args.input[i],
bp["sr"], sr=bp["sr"],
False, mono=False,
dtype=np.float32, dtype=np.float32,
res_type=bp["res_type"], res_type=bp["res_type"],
) )
@ -580,8 +584,8 @@ if __name__ == "__main__":
else: # lower bands else: # lower bands
wave[d] = librosa.resample( wave[d] = librosa.resample(
wave[d + 1], wave[d + 1],
mp.param["band"][d + 1]["sr"], orig_sr=mp.param["band"][d + 1]["sr"],
bp["sr"], target_sr=bp["sr"],
res_type=bp["res_type"], res_type=bp["res_type"],
) )

24
infer/modules/uvr5/vr.py
View File

@ -60,20 +60,20 @@ class AudioPre:
( (
X_wave[d], X_wave[d],
_, _,
) = librosa.core.load( # 理论上librosa读取可能对某些音频有bug，应该上ffmpeg读取，但是太麻烦了弃坑 ) = librosa.load( # 理论上librosa读取可能对某些音频有bug，应该上ffmpeg读取，但是太麻烦了弃坑
music_file, music_file,
bp["sr"], sr=bp["sr"],
False, mono=False,
dtype=np.float32, dtype=np.float32,
res_type=bp["res_type"], res_type=bp["res_type"],
) )
if X_wave[d].ndim == 1: if X_wave[d].ndim == 1:
X_wave[d] = np.asfortranarray([X_wave[d], X_wave[d]]) X_wave[d] = np.asfortranarray([X_wave[d], X_wave[d]])
else: # lower bands else: # lower bands
X_wave[d] = librosa.core.resample( X_wave[d] = librosa.resample(
X_wave[d + 1], X_wave[d + 1],
self.mp.param["band"][d + 1]["sr"], orig_sr=self.mp.param["band"][d + 1]["sr"],
bp["sr"], target_sr=bp["sr"],
res_type=bp["res_type"], res_type=bp["res_type"],
) )
# Stft of wave source # Stft of wave source
@ -241,20 +241,20 @@ class AudioPreDeEcho:
( (
X_wave[d], X_wave[d],
_, _,
) = librosa.core.load( # 理论上librosa读取可能对某些音频有bug，应该上ffmpeg读取，但是太麻烦了弃坑 ) = librosa.load( # 理论上librosa读取可能对某些音频有bug，应该上ffmpeg读取，但是太麻烦了弃坑
music_file, music_file,
bp["sr"], sr=bp["sr"],
False, mono=False,
dtype=np.float32, dtype=np.float32,
res_type=bp["res_type"], res_type=bp["res_type"],
) )
if X_wave[d].ndim == 1: if X_wave[d].ndim == 1:
X_wave[d] = np.asfortranarray([X_wave[d], X_wave[d]]) X_wave[d] = np.asfortranarray([X_wave[d], X_wave[d]])
else: # lower bands else: # lower bands
X_wave[d] = librosa.core.resample( X_wave[d] = librosa.resample(
X_wave[d + 1], X_wave[d + 1],
self.mp.param["band"][d + 1]["sr"], orig_sr=self.mp.param["band"][d + 1]["sr"],
bp["sr"], target_sr=bp["sr"],
res_type=bp["res_type"], res_type=bp["res_type"],
) )
# Stft of wave source # Stft of wave source

4
requirements-amd.txt
View File

@ -3,7 +3,7 @@ joblib>=1.1.0
numba==0.56.4 numba==0.56.4
numpy==1.23.5 numpy==1.23.5
scipy scipy
librosa==0.9.1 librosa==0.10.2
llvmlite==0.39.0 llvmlite==0.39.0
fairseq==0.12.2 fairseq==0.12.2
faiss-cpu==1.7.3 faiss-cpu==1.7.3
@ -41,7 +41,7 @@ pyworld==0.3.2
httpx httpx
onnxruntime onnxruntime
onnxruntime-gpu onnxruntime-gpu
torchcrepe==0.0.20 torchcrepe==0.0.23
fastapi==0.88 fastapi==0.88
ffmpy==0.3.1 ffmpy==0.3.1
python-dotenv>=1.0.0 python-dotenv>=1.0.0

4
requirements-dml.txt
View File

@ -2,7 +2,7 @@ joblib>=1.1.0
numba==0.56.4 numba==0.56.4
numpy==1.23.5 numpy==1.23.5
scipy scipy
librosa==0.9.1 librosa==0.10.2
llvmlite==0.39.0 llvmlite==0.39.0
fairseq==0.12.2 fairseq==0.12.2
faiss-cpu==1.7.3 faiss-cpu==1.7.3
@ -39,7 +39,7 @@ colorama>=0.4.5
pyworld==0.3.2 pyworld==0.3.2
httpx httpx
onnxruntime-directml onnxruntime-directml
torchcrepe==0.0.20 torchcrepe==0.0.23
fastapi==0.88 fastapi==0.88
ffmpy==0.3.1 ffmpy==0.3.1
python-dotenv>=1.0.0 python-dotenv>=1.0.0

4
requirements-ipex.txt
View File

@ -7,7 +7,7 @@ joblib>=1.1.0
numba==0.56.4 numba==0.56.4
numpy==1.23.5 numpy==1.23.5
scipy scipy
librosa==0.9.1 librosa==0.10.2
llvmlite==0.39.0 llvmlite==0.39.0
fairseq==0.12.2 fairseq==0.12.2
faiss-cpu==1.7.3 faiss-cpu==1.7.3
@ -45,7 +45,7 @@ pyworld==0.3.2
httpx httpx
onnxruntime; sys_platform == 'darwin' onnxruntime; sys_platform == 'darwin'
onnxruntime-gpu; sys_platform != 'darwin' onnxruntime-gpu; sys_platform != 'darwin'
torchcrepe==0.0.20 torchcrepe==0.0.23
fastapi==0.88 fastapi==0.88
ffmpy==0.3.1 ffmpy==0.3.1
python-dotenv>=1.0.0 python-dotenv>=1.0.0

4
requirements-py311.txt
View File

@ -2,7 +2,7 @@ joblib>=1.1.0
numba numba
numpy numpy
scipy scipy
librosa==0.9.1 librosa==0.10.2
llvmlite llvmlite
fairseq @ git+https://github.com/One-sixth/fairseq.git fairseq @ git+https://github.com/One-sixth/fairseq.git
faiss-cpu faiss-cpu
@ -40,7 +40,7 @@ pyworld==0.3.2
httpx httpx
onnxruntime; sys_platform == 'darwin' onnxruntime; sys_platform == 'darwin'
onnxruntime-gpu; sys_platform != 'darwin' onnxruntime-gpu; sys_platform != 'darwin'
torchcrepe==0.0.20 torchcrepe==0.0.23
fastapi==0.88 fastapi==0.88
torchfcpe torchfcpe
ffmpy==0.3.1 ffmpy==0.3.1

4
requirements.txt
View File

@ -2,7 +2,7 @@ joblib>=1.1.0
numba numba
numpy numpy
scipy scipy
librosa==0.9.1 librosa==0.10.2
llvmlite llvmlite
fairseq fairseq
faiss-cpu faiss-cpu
@ -40,7 +40,7 @@ pyworld==0.3.2
httpx httpx
onnxruntime; sys_platform == 'darwin' onnxruntime; sys_platform == 'darwin'
onnxruntime-gpu; sys_platform != 'darwin' onnxruntime-gpu; sys_platform != 'darwin'
torchcrepe==0.0.20 torchcrepe==0.0.23
fastapi==0.88 fastapi==0.88
torchfcpe torchfcpe
ffmpy==0.3.1 ffmpy==0.3.1