BoldActionMan commited on
Commit
7ce119a
1 Parent(s): c00e8d2

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +5 -4
app.py CHANGED
@@ -6,7 +6,7 @@ import torch
6
  from openvoice import se_extractor
7
  from openvoice.api import ToneColorConverter
8
  import whisper
9
- from moviepy import *
10
  from pydub import AudioSegment
11
  from df.enhance import enhance, init_df, load_audio, save_audio
12
  import translators as ts
@@ -52,7 +52,7 @@ def process_video(video_file, language_choice):
52
  audio = AudioSegment.from_file(reference_audio)
53
  resampled_audio = audio.set_frame_rate(48000)
54
  resampled_audio.export(reference_audio, format="wav")
55
- reference_audio = AudioClip(reference_audio)
56
 
57
  # Enhance the audio
58
  model, df_state, _ = init_df()
@@ -159,8 +159,9 @@ def process_video(video_file, language_choice):
159
  segment_path = os.path.join(output_dir, f'segment_{start}_{end}.wav')
160
  model.tts_to_file(translated_text, speaker_id, segment_path, speed=speed)
161
 
162
- reference_speaker = reference_audio.subclip(int(start), int(end)) # This is the voice you want to clone [int(start):int(end)]
163
- target_se, audio_name = se_extractor.get_se(reference_speaker, tone_color_converter, vad=False)
 
164
  # Run the tone color converter
165
  encode_message = "@MyShell"
166
  tone_color_converter.convert(
 
6
  from openvoice import se_extractor
7
  from openvoice.api import ToneColorConverter
8
  import whisper
9
+ from moviepy.editor import *
10
  from pydub import AudioSegment
11
  from df.enhance import enhance, init_df, load_audio, save_audio
12
  import translators as ts
 
52
  audio = AudioSegment.from_file(reference_audio)
53
  resampled_audio = audio.set_frame_rate(48000)
54
  resampled_audio.export(reference_audio, format="wav")
55
+ audio_clip = AudioFileClip(reference_audio)
56
 
57
  # Enhance the audio
58
  model, df_state, _ = init_df()
 
159
  segment_path = os.path.join(output_dir, f'segment_{start}_{end}.wav')
160
  model.tts_to_file(translated_text, speaker_id, segment_path, speed=speed)
161
 
162
+ reference_speaker = AudioFileClip.subclip(audio_clip, int(start), int(end)) # This is the voice you want to clone
163
+ reference_speaker.write_audiofile("reference_speaker.wav")
164
+ target_se, audio_name = se_extractor.get_se("reference_speaker.wav", tone_color_converter, vad=False)
165
  # Run the tone color converter
166
  encode_message = "@MyShell"
167
  tone_color_converter.convert(