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

bug fixes

pull/10/head
root 2020-01-27 15:42:56 +01:00
parent a391a7f859
commit eb63c95d97
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
utils/audio.py
View File

@ -66,12 +66,11 @@ class AudioProcessor(object):
return np.maximum(1e-10, np.dot(inv_mel_basis, mel_spec)) return np.maximum(1e-10, np.dot(inv_mel_basis, mel_spec))
def _build_mel_basis(self, ): def _build_mel_basis(self, ):
n_fft = (self.num_freq - 1) * 2
if self.mel_fmax is not None: if self.mel_fmax is not None:
assert self.mel_fmax <= self.sample_rate // 2 assert self.mel_fmax <= self.sample_rate // 2
return librosa.filters.mel( return librosa.filters.mel(
self.sample_rate, self.sample_rate,
n_fft, self.n_fft,
n_mels=self.num_mels, n_mels=self.num_mels,
fmin=self.mel_fmin, fmin=self.mel_fmin,
fmax=self.mel_fmax) fmax=self.mel_fmax)
@ -197,6 +196,7 @@ class AudioProcessor(object):
n_fft=self.n_fft, n_fft=self.n_fft,
hop_length=self.hop_length, hop_length=self.hop_length,
win_length=self.win_length, win_length=self.win_length,
pad_mode='constant'
) )
def _istft(self, y): def _istft(self, y):
@ -217,7 +217,7 @@ class AudioProcessor(object):
margin = int(self.sample_rate * 0.01) margin = int(self.sample_rate * 0.01)
wav = wav[margin:-margin] wav = wav[margin:-margin]
return librosa.effects.trim( return librosa.effects.trim(
wav, top_db=60, frame_length=self.win_length, hop_length=self.hop_length)[0] wav, top_db=40, frame_length=self.win_length, hop_length=self.hop_length)[0]
@staticmethod @staticmethod
def mulaw_encode(wav, qc): def mulaw_encode(wav, qc):
@ -247,7 +247,7 @@ class AudioProcessor(object):
print(f' [!] File cannot be trimmed for silence - {filename}') print(f' [!] File cannot be trimmed for silence - {filename}')
assert self.sample_rate == sr, "%s vs %s"%(self.sample_rate, sr) assert self.sample_rate == sr, "%s vs %s"%(self.sample_rate, sr)
if self.sound_norm: if self.sound_norm:
x = x / abs(x.max()) * 0.9 x = x / abs(x).max() * 0.9
return x return x
@staticmethod @staticmethod