It seems that you're using an outdated browser. Some things may not work as they should (or don't work at all).
We suggest you upgrade newer and better browser like: Chrome, Firefox, Internet Explorer or Opera

×
Hello.

Any recommendation for good FLOSS/FOSS programs for creating subtitles?
Any free audio to text Linux programs will do.
No posts in this topic were marked as the solution yet. If you can help, add your reply
Louis Rossman had Futo created, which is a privacy aware voice keyboard input device tool for android.

https://futo.org/

I'm not quite sure on other apps as i don't usually look/go for these tools.
avatar
.Keys: Hello.

Any recommendation for good FLOSS/FOSS programs for creating subtitles?
Any free audio to text Linux programs will do.
Is this of use to you https://github.com/abb128/LiveCaptions ?

Didn't try it myself due to time constraints, but want to, since it seems quite nice.
avatar
McFirson: Didn't try it myself due to time constraints, but want to, since it seems quite nice.
The release is source code. When it isn't trivial to compile i find it a pain to try for source code.
Hmm. Speech to text is tricky. But possible.

You could do some piping or output routing; for example Pocketsphynx > some kind of program to display text on an OSD...



(Other programs exist, such as Julius), or you could use a notification daemon to pipe the text to a notification caption...
Post edited July 24, 2024 by dnovraD
Not sure if this is helpful or not but kdenlive now has a built in auto subtitle generator although I have not tried it yet and the archwiki also list subtitle editors, some of which support audio to text.
Thanks for the answers everyone.
Surprisingly, Pitney_b answer was exactly what I was looking for because I was also using Kdenlive to try to add subtitles automatically.

The setup is quite automatic and you can use VOSK Python script inside Kdenlive or let it auto download dependencies and install language models.

The only thing that can be considered annoying is that apparently it uses some OpenAI models (?) for speech-to-text?
(Need to confirm that.)
So be warned if you want to try that and are against OpenAI.