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

Add RMS based norm in save_wav method

This commit is contained in:
Edresson Casanova 2022-04-21 08:26:27 -03:00
parent bb7a645e7a
commit 21cbfe964e
2 changed files with 7 additions and 2 deletions
TTS
server
server.py
utils
audio.py

3
TTS/server/server.py
View File

@ -111,7 +111,8 @@ synthesizer = Synthesizer(
use_cuda=args.use_cuda, use_cuda=args.use_cuda,
) )
use_multi_speaker = hasattr(synthesizer.tts_model, "num_speakers") and synthesizer.tts_model.num_speakers > 1 use_multi_speaker = hasattr(synthesizer.tts_model, "num_speakers") and (synthesizer.tts_model.num_speakers > 1 or synthesizer.tts_speakers_file is not None)
print("Multispeaker?", use_multi_speaker, synthesizer.tts_model.num_speakers)
speaker_manager = getattr(synthesizer.tts_model, "speaker_manager", None) speaker_manager = getattr(synthesizer.tts_model, "speaker_manager", None)
# TODO: set this from SpeakerManager # TODO: set this from SpeakerManager
use_gst = synthesizer.tts_config.get("use_gst", False) use_gst = synthesizer.tts_config.get("use_gst", False)

6
TTS/utils/audio.py
View File

@ -859,7 +859,11 @@ class AudioProcessor(object):
path (str): Path to a output file. path (str): Path to a output file.
sr (int, optional): Sampling rate used for saving to the file. Defaults to None. sr (int, optional): Sampling rate used for saving to the file. Defaults to None.
""" """
wav_norm = wav * (32767 / max(0.01, np.max(np.abs(wav)))) if self.do_rms_norm:
wav_norm = self.rms_volume_norm(wav, self.db_level) * 32767
else:
wav_norm = wav * (32767 / max(0.01, np.max(np.abs(wav))))
scipy.io.wavfile.write(path, sr if sr else self.sample_rate, wav_norm.astype(np.int16)) scipy.io.wavfile.write(path, sr if sr else self.sample_rate, wav_norm.astype(np.int16))
def get_duration(self, filename: str) -> float: def get_duration(self, filename: str) -> float: