fumiama/Retrieval-based-Voice-Conversion-WebUI
1
0
Fork 0
mirror of https://github.com/fumiama/Retrieval-based-Voice-Conversion-WebUI.git synced 2026-06-09 04:29:50 +08:00
Code Activity

some change precision audio processing (#94)

Browse Source 
* some change precision audio processing

* fix clipping problem in resample

resample sometimes causes signal clipping, not just librosa.resample

* fix error
This commit is contained in:
autumnmotor
2023-04-22 20:39:47 +09:00
committed by GitHub
parent c423f77a16
commit 297d92bf5d
4 changed files with 32 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
trainset_preprocess_pipeline_print.py
View File

@@ -59,19 +59,34 @@ class PreProcess:
wavfile.write(
"%s/%s_%s.wav" % (self.gt_wavs_dir, idx0, idx1),
self.sr,
(tmp_audio * 32768).astype(np.int16),
(tmp_audio * 1).astype(np.float32),
)
tmp_audio = librosa.resample(tmp_audio, orig_sr=self.sr, target_sr=16000)
# default resample type of librosa.resample is "soxr_hq".
# Quality: soxr_vhq > soxr_hq
tmp_audio = librosa.resample(tmp_audio, orig_sr=self.sr, target_sr=16000, res_type="soxr_vhq")
tmp_audio = (tmp_audio / np.abs(tmp_audio).max() * (self.max * self.alpha)) + (
1 - self.alpha
) * tmp_audio
wavfile.write(
"%s/%s_%s.wav" % (self.gt_wavs_dir, idx0, idx1),
self.sr,
(tmp_audio * 1).astype(np.float32),
)
wavfile.write(
"%s/%s_%s.wav" % (self.wavs16k_dir, idx0, idx1),
16000,
(tmp_audio * 32768).astype(np.int16),
(tmp_audio * 1).astype(np.float32),
)
def pipeline(self, path, idx0):
try:
audio = load_audio(path, self.sr)
audio = signal.filtfilt(self.bh, self.ah, audio)
# zero phased digital filter cause pre-ringing noise...
# audio = signal.filtfilt(self.bh, self.ah, audio)
audio = signal.lfilter(self.bh, self.ah, audio)
idx1 = 0
for audio in self.slicer.slice(audio):
i = 0