Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

refactor(audio.processor): use load_wav from numpy_transforms

This commit is contained in:
Enno Hermann 2023-11-14 13:46:57 +01:00
parent 9a43eafd60
commit 13e640f17e
1 changed files with 4 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
TTS/utils/audio/processor.py
View File

@ -5,7 +5,6 @@ import librosa
import numpy as np
import scipy.io.wavfile
import scipy.signal
import soundfile as sf
from TTS.tts.utils.helpers import StandardScaler
from TTS.utils.audio.numpy_transforms import (
@ -16,6 +15,7 @@ from TTS.utils.audio.numpy_transforms import (
deemphasis,
find_endpoint,
griffin_lim,
load_wav,
mel_to_spec,
millisec_to_length,
preemphasis,
@ -587,15 +587,10 @@ class AudioProcessor(object):
Returns:
np.ndarray: Loaded waveform.
"""
if self.resample:
# loading with resampling. It is significantly slower.
x, sr = librosa.load(filename, sr=self.sample_rate)
elif sr is None:
# SF is faster than librosa for loading files
x, sr = sf.read(filename)
assert self.sample_rate == sr, "%s vs %s" % (self.sample_rate, sr)
if sr is not None:
x = load_wav(filename=filename, sample_rate=sr, resample=True)
else:
x, sr = librosa.load(filename, sr=sr)
x = load_wav(filename=filename, sample_rate=self.sample_rate, resample=self.resample)
if self.do_trim_silence:
try:
x = self.trim_silence(x)