Post your JAV subtitle files here - JAV Subtitle Repository (JSP)★NOT A SUB REQUEST THREAD★

[Taako]

You gotta be careful since extracting(aka demuxing or demultiplexing) audio isn't the same as converting(aka re-encoding) audio to a specific format. Both will give you an audio file on its own as a result(so you can technically say re-encoding it is extracting it too) so the difference might not be obvious to those unfamiliar with how these things work.

The former doesn't change the audio in any way and is basically instantaneous or takes a couples secs to do, it just takes the audio part out of the file container that holds both video and audio and the latter takes that audio part and completely re-encodes it to a different format or even the same to create an entirely new file and can take a few seconds or minutes depending on your pc.

I think VLC can do both of those things though, but the first one is better since re-encoding means quality loss(unless it's to a lossless format like wav, but those files are huge).

This is very true.
Audacity will as well. It's very rare. You just never know.

 

[soloporhoy666]

After Jux-45 just sat there for hours then timed out yesterday, decided to try a different file. Crapped out again, but differently and with different error messages. I'll try a couple other files but I highly doubt this will ever work. This software should not be up on the web because it isn't done yet. There are too many failure triggers in this code.

View attachment 3129755

There it says that you are missing an essential audio codec to work. ffmpeg

 

[soloporhoy666]

DASS-070 Akari Mitani

One on the things I have always liked about JAV is that they often commit to telling a real story with genuine acting. This video is a prime example of that. In fact, I actually found myself fast-forwarding through the sex scenes to see where the plot was going. Hope others enjoy it.

View attachment 3129788

Akari Mitani's movies are mostly incredible, many have a good story, she and Nao Jinguji share my No 1.

 

[soloporhoy666]

After Jux-45 just sat there for hours then timed out yesterday, decided to try a different file. Crapped out again, but differently and with different error messages. I'll try a couple other files but I highly doubt this will ever work. This software should not be up on the web because it isn't done yet. There are too many failure triggers in this code.

View attachment 3129755

It is failing in the most important part of Whisper to encode the audio

 

[soloporhoy666]

After Jux-45 just sat there for hours then timed out yesterday, decided to try a different file. Crapped out again, but differently and with different error messages. I'll try a couple other files but I highly doubt this will ever work. This software should not be up on the web because it isn't done yet. There are too many failure triggers in this code.

View attachment 3129755

I have read that many other colleagues did not have to install anything on their computers, but each one of us has different versions of computers, in my case, the first time I tried to use the collab, it gave me precisely a codec error, ffmpeg I had to install it manually on my computer (I'm sorry to report that it was quite difficult for me to install). but that solved the error that you mentioned in your files.

 

[SamKook]

There it says that you are missing an essential audio codec to work. ffmpeg

I have read that many other colleagues did not have to install anything on their computers, but each one of us has different versions of computers, in my case, the first time I tried to use the collab, it gave me precisely a codec error, ffmpeg I had to install it manually on my computer (I'm sorry to report that it was quite difficult for me to install). but that solved the error that you mentioned in your files.

That doesn't make any sense, it's doing the processing entirely off of your computer once you upload the file so installing anything locally wouldn't help in any way.

What that error means is that ffmpeg had an error while running, not that it's missing. More specifically, it's unable to create the temporary uncompressed wav file for processing.

If you look at the Setup Whisper code, you'll clearly see the line where it installs ffmpeg on the remote system used for the processing: "pip install deepl srt ffmpeg-python spleeter".

 

[soloporhoy666]

That doesn't make any sense, it's doing the processing entirely off of your computer once you upload the file so installing anything locally wouldn't help in any way.

What that error means is that ffmpeg had an error while running, not that it's missing. More specifically, it's unable to create the temporary uncompressed wav file for processing.

If you look at the Setup Whisper code, you'll clearly see the line where it installs ffmpeg on the remote system used for the processing: "pip install deepl srt ffmpeg-python spleeter".

I only mentioned what worked in my case, watching several videos, because I don't know about codes, but some youtuber that I see from time to time, mentioned that the codec was necessary, but I don't know anything about that, I just followed the instructions to be able to use the collab that I got to use before.
By the way once I installed the codec, boom like magic I was able to use Whisper.

 

[SamKook]

I only mentioned what worked in my case, watching several videos, because I don't know about codes, but some youtuber that I see from time to time, mentioned that the codec was necessary, but I don't know anything about that, I just followed the instructions to be able to use the collab that I got to use before.

What I'm saying is that you doing that did nothing, you must have done something differently after installing it that fixed it(or your audio file was corrupt before).

It's true ffmpeg is required, but only on the pc whisper is running on.

 

[Taako]

What I'm saying is that you doing that did nothing, you must have done something differently after installing it that fixed it(or your audio file was corrupt before).

It's true ffmpeg is required, but only on the pc whisper is running on.

Indeed I didn't install anything, I don't think...

Anyways, if it can work for me and I'm not using it, I hope it will for others.
It just seems buggy. I had to work with it more times than I wanted.

 

[Taako]

I have read that many other colleagues did not have to install anything on their computers, but each one of us has different versions of computers, in my case, the first time I tried to use the collab, it gave me precisely a codec error, ffmpeg I had to install it manually on my computer (I'm sorry to report that it was quite difficult for me to install). but that solved the error that you mentioned in your fileView attachment 3129813s.

Yes. All our computers are different. I force myself to do it on mines just in case others needed help.
Not sure why it was so hard to get started on my computer but... its buggy...
Oh well... back to cleaning my subs.

 

[SamKook]

Yes. All our computers are different. I forced myself to do it on mine just in case others needed help.
Not sure why it was so hard to get started on my computer but... its buggy...
Oh well... back to cleaning my subs.

You're making me feel bad now since all I had to do the first time I installed it was to copy paste the command and run it, lol.

 

[soloporhoy666]

You're making me feel bad now since all I had to do the first time I installed it was to copy paste the command and run it, lol.

But they are deviating from the main topic, I only see that they correct what worked for me, but what solution can they give to the partner, simply do nothing and it is solved by itself? I don't know ...

 

[SamKook]

That's the same error I got when I uploaded a corrupted audio file. I suspect the issue is that you're uploading video files instead of only the audio file. whisper can handle those fine but considering the colab has extra restrictions, it might not be able to use the extracted audio because you're telling it the input is the video file. The interface they made for it might not be smart enough to find the audio file after it gets demuxed by ffmpeg.

^ That is very likely the solution, uploading an audio file(that you tested to make sure is working first).

What that ffmpeg error means is the input file isn't detected as audio basically so either the uploaded file got corrupted, was corrupt to begin with or the colab version can't handle extracting the audio from a video file(haven't tested using a video so I don't know).

 

[Prinsipe]

Is there a significant difference in extracting the mp3 rather than converting when running the audio file to whisper?

Anyway, Whisper is not built for the web. It is a python program that will run standalone without an internet connection as long as you have a powerful GPU.

 

[Taako]

You're making me feel bad now since all I had to do the first time I installed it was to copy paste the command and run it, lol.

hahaha no way. I'm so glad it was easy for you.
But my computer is kinda old I do wonder what graphic card will make it run better on install?

And from your previous statement, I agree, if everyone uses the web version, it's gonna slow the servers sooner or later.

 

[maelstrom9999]

^ That is very likely the solution, uploading an audio file(that you tested to make sure is working first).

What that ffmpeg error means is the input file isn't detected as audio basically so either the uploaded file got corrupted, was corrupt to begin with or the colab version can't handle extracting the audio from a video file(haven't tested using a video so I don't know).

Is there a way to take a video file, then make an audio only file from it? I mean, if it won't accept video files, that would be the only solution, right?

 

[Taako]

Is there a way to take a video file, then make an audio only file from it? I mean, if it won't accept video files, that would be the only solution, right?

Yes MkvToolNix GUI is your friend

 

[SamKook]

Is there a significant difference in extracting the mp3 rather than converting when running the audio file to whisper?

The difference is not big, likely unnoticeable in most cases. but it could hurt recognition and almost certainly won't help if you convert instead of extract.

hahaha no way. I'm so glad it was easy for you.
But my computer is kinda old I do wonder what graphic card will make it run better on install?

A very expensive nvidia gamer gpu(RTX 3080 in my case) is what most people would have that can work well enough for it, but just to do ai stuff, you'd buy a datacenter one like the 3 they mention on the colab page.

 

[SamKook]

Is there a way to take a video file, then make an audio only file from it? I mean, if it won't accept video files, that would be the only solution, right?

Yes MkvToolNix GUI is your friend

From the screenshot, his video is a mp4 so I don't think mkvtoolnix works for that, at least it didn't before, only for mkv.
Multiple people mentionned some so just pick one. i use mp4box with a separate gui but the new version complicates things so not the easiest to install.

Edit: Googling "demux mp4" will get you plenty of options, but here's a tutorial to do it with vlc which you likely already have: https://bartman88.blogspot.com/2018/04/how-to-extract-or-demux-audio-or-video.html

Edit2: Oh, you can do it with mkvtoolnix, nice. Can't put anything inside an mp4 but you can take it out. Well, not quite take it out but put it inside an mka which is a mkv with only audio so should do the trick.

 

[maelstrom9999]

Well I decided to keep trying to make the video files work for awhile before bothering with that and holy shit, it finally worked. I can only get it to work by using the embedded "choose files" in the upload audio section, finding the file in its path, then waiting for the whole file to upload, which took roughly 2 hours in this case. At which point, with this method, you don't need to copy the filepath. It is automatically populated into the file path field under play whisper the moment the upload is complete. Then (but not before) hit the "play whisper" arrow.

Seems to have taken roughly an hour to throw an output file into my download folder. Haven't checked syncing yet, but so far as dialogue, I'd rate it as an above average autosub, as compared to say raw autosubs made from Chinese softsubs. It's definitely in cleanable shape.

The title is midd-847, but I won't upload until I've cleaned it at least somewhat.

Now I'm going back to check the files that failed and try doing them this way instead. If this can work on all or even most of my video files, it's going to be great. I still have 36 titles in my faves folders with no sub files. Gonna get all of them. Will post all here of course.

Thanks for all the help provided above.