Update app.py

app.py

@@ -73,7 +73,7 @@ vocos = load_vocoder()
 
 target_sample_rate = 24000
 n_mel_channels = 100
-hop_length = 256 - 6
+hop_length = 256
 target_rms = 0.1
 nfe_step = 16  # 16, 32
 cfg_strength = 2.0
@@ -379,13 +379,13 @@ def infer(ref_audio_orig, ref_text, gen_text, exp_name, remove_silence, cross_fa
     with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as f:
         aseg = AudioSegment.from_file(ref_audio_orig)
         aseg = remove_silence_edges(aseg) + AudioSegment.silent(duration=150)
-#        non_silent_segs = silence.split_on_silence(
-#            aseg, min_silence_len=1000, silence_thresh=-42, keep_silence=1000
-#        )
-#        non_silent_wave = AudioSegment.silent(duration=0)
-#        for non_silent_seg in non_silent_segs:
-#            non_silent_wave += non_silent_seg
-#        aseg = non_silent_wave
+        non_silent_segs = silence.split_on_silence(
+            aseg, min_silence_len=1000, silence_thresh=-50, keep_silence=1000
+        )
+        non_silent_wave = AudioSegment.silent(duration=0)
+        for non_silent_seg in non_silent_segs:
+            non_silent_wave += non_silent_seg
+        aseg = non_silent_wave
         
         
         audio_duration = len(aseg)