readbeyond / aeneas

aeneas is a Python/C library and a set of tools to automagically synchronize audio and text (aka forced alignment)
http://www.readbeyond.it/aeneas/
GNU Affero General Public License v3.0
2.44k stars 218 forks source link

aeneas crashes on large files, error 1 #279

Open YaBoiBreezy opened 2 years ago

YaBoiBreezy commented 2 years ago

I got aeneas, it works in english and french on 5-minute mp3's but when I give it a full 12 hour one (600 MB mp3, 700 KB text file), it crashes with basically no info. I did -v -l, it sets up the whole text file with TextFilter: Applying regex: '[' for each line, then gives this:

[DEBU] 2021-11-17 08:55:28.925569 TextFile: Parsed 3194 fragments [DEBU] 2021-11-17 08:55:28.925569 TextFile: Setting language: 'fr' [DEBU] 2021-11-17 08:55:28.928566 Task: Populate text file... done [INFO] 2021-11-17 08:55:28.928566 CLI: Creating task... done [INFO] 2021-11-17 08:55:28.928566 CLI: Executing task... [DEBU] 2021-11-17 08:55:28.928566 ExecuteTask: Executing task... [DEBU] 2021-11-17 08:55:28.938566 ExecuteTask: Both audio and text input file are present [DEBU] 2021-11-17 08:55:28.938566 ExecuteTask: Executing single level task... [DEBU] 2021-11-17 08:55:28.938566 ExecuteTask: STEP 1 BEGIN (extract MFCC real wave) [DEBU] 2021-11-17 08:55:28.938566 AudioFileMFCC: Initializing MFCCs... [DEBU] 2021-11-17 08:55:28.938566 AudioFile: Loading audio data... [DEBU] 2021-11-17 08:55:28.939566 AudioFile: self.file_format is None or not good => converting self.file_path [DEBU] 2021-11-17 08:55:28.940568 AudioFile: Temporary PCM16 mono WAVE file: 'C:\Users\BreezA\AppData\Local\Temp\tmp71xnoktg.wav' [DEBU] 2021-11-17 08:55:28.940568 AudioFile: Converting audio file to mono... [DEBU] 2021-11-17 08:55:28.943568 FFMPEGWrapper: Calling with arguments '['ffmpeg', '-i', 'stream69French.mp3', '-ac', '1', '-ar', '16000', '-y', '-map_metadata', '-1', '-flags', '+bitexact', '-f', 'wav', 'C:\Users\BreezA\AppData\Local\Temp\tmp71xnoktg.wav']' [DEBU] 2021-11-17 08:56:58.146430 FFMPEGWrapper: Call completed [DEBU] 2021-11-17 08:56:58.146430 FFMPEGWrapper: Returning output file path 'C:\Users\BreezA\AppData\Local\Temp\tmp71xnoktg.wav' [DEBU] 2021-11-17 08:56:58.147432 AudioFile: Converting audio file to mono... done [CRIT] 2021-11-17 08:56:58.149421 ExecuteTask: STEP 1 (extract MFCC real wave) FAILURE [CRIT] 2021-11-17 08:56:58.149421 ExecuteTask: Unexpected error while executing task [CRIT] 2021-11-17 08:56:58.150435 ExecuteTask: [ERRO] 2021-11-17 08:56:58.150435 CLI: An unexpected error occurred while executing the task: [ERRO] 2021-11-17 08:56:58.150435 CLI: Unexpected error while executing task : [DEBU] 2021-11-17 08:56:58.150435 CLI: Execution completed with code 1

It breaks on english and french, I watched in task manager and nothing goes to 100%, ffmpeg.exe takes 30% cpu, that's it.

yasntrk commented 2 years ago

Can you share your text file?

i990049 commented 1 year ago

try split mp3 into many files.each file no more than 2 hours

sharinganthief commented 1 year ago

i get errors after 20 minutes

changyr66 commented 1 year ago

I got errors after 30min. Using Python library package instead of the command line might help.