Update TTS inference to validate audio duration using soundfile.

third_party/moyoyo_tts/TTS_infer_pack/TTS.py

@@ -10,6 +10,7 @@ from typing import List, Tuple, Union
 
 import ffmpeg
 import librosa
+import soundfile
 import numpy as np
 import torch
 import yaml
@@ -469,7 +470,9 @@ class TTS:
         )
         with torch.no_grad():
             wav16k, sr = librosa.load(ref_wav_path, sr=16000)
-            if (wav16k.shape[0] > 160000 or wav16k.shape[0] < 48000):
+            y, sr, = soundfile.read(ref_wav_path)
+            duration = librosa.get_duration(y=y, sr=sr)
+            if (duration < 3 or duration > 10):
                 raise OSError(i18n("参考音频在3~10秒范围外，请更换！"))
             wav16k = torch.from_numpy(wav16k)
             zero_wav_torch = torch.from_numpy(zero_wav)