1KB empty subtitle files after transcribing

[Fetterbr61]

I am getting many empty subtitle files after trascribtion. I use Purfviews batch with L 3 and translate to english on. I testet a little and got the below log, can someone see what could be the error please. Thanks

---

Date: 06/13/2024 07:48:15
SE: 4.0.6.0 - Microsoft Windows NT 10.0.19045.0 - 64-bit
Message: C:\Users\crist\AppData\Roaming\Subtitle Edit\Whisper\Purfview-Whisper-Faster\whisper-faster.exe --language ja --model "large-v3" --task translate --standard_asia "D:\VLC Media folder\Javluv New Movies Latest edition\INU-052 (2013)\INU-052 (2013).wmv"

---

Date: 06/13/2024 07:48:41
SE: 4.0.6.0 - Microsoft Windows NT 10.0.19045.0 - 64-bit
Message: Calling whisper (Purfview's Faster-Whisper) with : C:\Users\crist\AppData\Roaming\Subtitle Edit\Whisper\Purfview-Whisper-Faster\whisper-faster.exe --language ja --model "large-v3" --task translate --standard_asia "D:\VLC Media folder\Javluv New Movies Latest edition\INU-052 (2013)\INU-052 (2013).wmv"
Warning: Word-level timestamps on translations may not be reliable.

Warning: 'large-v3' model may produce inferior results, try 'large-v2'!

Standalone Faster-Whisper r192.3 running on: CUDA
Starting work on: D:\VLC Media folder\Javluv New Movies Latest edition\INU-052 (2013)\INU-052 (2013).wmv
overflow (1870 > 1864) in spectral RLE, ignoring

[1288] Failed to execute script 'main' due to unhandled exception!

av.error.PermissionError: [Errno 1] Operation not permitted; last error log: [wmav2] overflow (1870 > 1864) in spectral RLE, ignoring

File "av\error.pyx", line 336, in av.error.err_check

File "av\codec\context.pyx", line 416, in av.codec.context.CodecContext._send_packet_and_recv

File "av\codec\context.pyx", line 513, in av.codec.context.CodecContext.decode

File "av\stream.pyx", line 168, in av.stream.Stream.decode

File "av\packet.pyx", line 87, in av.packet.Packet.decode

File "av\container\input.pyx", line 212, in decode

File "faster_whisper\audio.py", line 80, in _ignore_invalid_frames

File "faster_whisper\audio.py", line 90, in _group_frames

File "faster_whisper\audio.py", line 103, in _resample_frames

File "faster_whisper\audio.py", line 52, in decode_audio

File "faster_whisper\transcribe.py", line 315, in transcribe

File "D:\whisper-fast_main_.py", line 1476, in cli

File "D:\whisper-fast_main_.py", line 1600, in

Traceback (most recent call last):
Calling whisper Purfview's Faster-Whisper done in 00:00:26.0097232
Loading result from STDOUT

[niksedk]

I'm not sure what the problem is... but you try latest beta and see if it's any better?
https://github.com/SubtitleEdit/subtitleedit/releases/download/4.0.6/SubtitleEditBeta.zip

[Fetterbr61]

thanks, i have installed it and it seams to be working, not sure what did it. I have a RTX 4070 TI with a lot of cudas, what is the best setup for this and the quickest, i can see the XXL and Faster Whisper is very slow on my rig, maybe i am just doing it wrong.

[niksedk]

Thx for testing :)
Apparently Purfview Faster Whisper cannot handle wmv files. @Purfview: Correct ?
You can use a smaller model to make it go faster.
large-v2 is the recommended model as large-v3 does not seem to work quite as well...
Also, you might get better translate result in other translation engines like DeepL or Google Translate.

[Purfview]

Looks like some ffmpeg error with audio decoding, try --ff_dump.

[Fetterbr61]

Looks like some ffmpeg error with audio decoding, try --ff_dump.

How do you do a --ff dump

[Fetterbr61]

Hej og tak for dit svar. I have actually tryed and payed for the tranlation in your program but i can't get it to work. fre. 14. jun. 2024 20.40 skrev Nikolaj Olsson ***@***.***>:

…

Thx for testing :) Apparently Purfview Faster Whisper cannot handle wmv files. @Purfview <https://github.com/Purfview>: Correct ? You can use a smaller model to make it go faster. large-v2 is the recommended model as large-v3 does not seem to work quite as well... Also, you might get better translate result in other translation engines like DeepL or Google Translate. — Reply to this email directly, view it on GitHub <#8525 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BARYHED3TVH7CSCIAVNRKSTZHM2INAVCNFSM6AAAAABJHVVHYGVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TONZXG4ZDI> . You are receiving this because you authored the thread.Message ID: ***@***.*** com>

[Fetterbr61]

Hi.
It seams after a little testing that a fresh restart of the computer makes the error go away. I can transcribe without errors. Its more likely that after some time of work with different things and then making the transcribe will faile with 1 kb files. I am not sure if there could be something with the app and integration into the system, memory or other things. But as i said a restart fixes the problem for the time being.