Turn on Audio Sync for audio extraction process

morpheus65535 · Sep 2, 2024 · 5ee7cde · 5ee7cde
1 parent 97ea3a8
commit 5ee7cde
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/custom_libs/subliminal_patch/providers/whisperai.py b/custom_libs/subliminal_patch/providers/whisperai.py
@@ -143,7 +143,7 @@ def encode_audio_stream(path, ffmpeg_path, audio_stream_language=None):
             logger.debug(f"Whisper will only use the {audio_stream_language} audio stream for {path}")
             inp = inp[f'a:m:language:{audio_stream_language}']
 
-        out, _ = inp.output("-", format="s16le", acodec="pcm_s16le", ac=1, ar=16000) \
+        out, _ = inp.output("-", format="s16le", acodec="pcm_s16le", ac=1, ar=16000, af="aresample=async=1") \
                     .run(cmd=[ffmpeg_path, "-nostdin"], capture_stdout=True, capture_stderr=True)
 
     except ffmpeg.Error as e:

diff --git a/libs/ffsubsync/speech_transformers.py b/libs/ffsubsync/speech_transformers.py
@@ -367,6 +367,8 @@ def fit(self, fname: str, *_) -> "VideoSpeechTransformer":
                 "1",
                 "-acodec",
                 "pcm_s16le",
+                "-af",
+                "aresample=async=1",
                 "-ar",
                 str(self.frame_rate),
                 "-",