blackteay/GPT-SoVITS
1
0
Fork 0
mirror of https://github.com/RVC-Boss/GPT-SoVITS.git synced 2025-10-07 23:48:48 +08:00
Code Issues Packages Projects Releases Wiki Activity

优化了切分音频的性能和逻辑

Browse Source
This commit is contained in:
XTer 2024-03-22 01:29:07 +08:00
parent 1979211eda
commit 8a8ddad7e8
2 changed files with 7 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
tools/srt_slicer/srt_utils.py
View File

@ -74,11 +74,16 @@ import pydub, os
def slice_audio_with_lib(audio_path, save_folder, format, subtitles, pre_preserve_time, post_preserve_time, pre_silence_time, post_silence_time, language='ZH', character='character'): def slice_audio_with_lib(audio_path, save_folder, format, subtitles, pre_preserve_time, post_preserve_time, pre_silence_time, post_silence_time, language='ZH', character='character'):
list_file = os.path.join(save_folder, 'datamapping.list') list_file = os.path.join(save_folder, 'datamapping.list')
try:
audio = pydub.AudioSegment.from_file(audio_path)
except Exception as e:
raise e
with open(list_file, 'w', encoding="utf-8") as f: with open(list_file, 'w', encoding="utf-8") as f:
for i in range(len(subtitles)): for i in range(len(subtitles)):
subtitle = subtitles[i] subtitle = subtitles[i]
start = subtitle.start.total_seconds() - pre_preserve_time start = subtitle.start.total_seconds() - pre_preserve_time
end = subtitle.end.total_seconds() + post_preserve_time end = subtitle.end.total_seconds() + post_preserve_time
if i < len(subtitles) - 1: if i < len(subtitles) - 1:
next_subtitle = subtitles[i + 1] next_subtitle = subtitles[i + 1]
end = min(end, 1.0/2*(subtitle.end.total_seconds()+next_subtitle.start.total_seconds())) end = min(end, 1.0/2*(subtitle.end.total_seconds()+next_subtitle.start.total_seconds()))
@ -86,12 +91,12 @@ def slice_audio_with_lib(audio_path, save_folder, format, subtitles, pre_preserv
prev_subtitle = subtitles[i - 1] prev_subtitle = subtitles[i - 1]
start = max(start, 1.0/2*(prev_subtitle.end.total_seconds()+subtitle.start.total_seconds())) start = max(start, 1.0/2*(prev_subtitle.end.total_seconds()+subtitle.start.total_seconds()))
try: try:
audio = pydub.AudioSegment.from_file(audio_path)
sliced_audio = audio[int(start * 1000):int(end * 1000)] sliced_audio = audio[int(start * 1000):int(end * 1000)]
file_name = f'{character}_{i + 1:03d}.{format}' file_name = f'{character}_{i + 1:03d}.{format}'
save_path = os.path.join(save_folder, file_name) save_path = os.path.join(save_folder, file_name)
sliced_audio.export(save_path, format=format) sliced_audio.export(save_path, format=format)
f.write(f"{file_name}|{character}|{language}|{subtitle.content}\n") f.write(f"{file_name}|{character}|{language}|{subtitle.content}\n")
print(f"Slice {file_name} from {start} to {end}")
except Exception as e: except Exception as e:
raise e raise e

2
tools/srt_slicer/webui.py
View File

@ -16,7 +16,7 @@ from srt_utils import (
) )
port = 8991 port = 8991
is_share = False is_share = True
if len(sys.argv) > 2: if len(sys.argv) > 2:
port = int(sys.argv[1]) port = int(sys.argv[1])
is_share = eval(sys.argv[2]) is_share = eval(sys.argv[2])