Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

Very happy with your program, thanks again! Even though it's working well most of the time, I have a few videos where I just get a long list of [speaking in French] or even [speaking on foreign language] even though it's all French, and a very clear one with no accent (documentaries videos). A bit hard to understand these fails! I also tried another documentary with mixed languages used (with interviews for instance), and I can confirm that German is not being picked up for some reason. Also it doesn't look like it's able to pick up different languages in 1 single file - it was probably not meant to be used that way but it could be super nice if it was possible!

Hi jairoaf,

Yes, the automatic detection uses the start of the audio and sometimes it might guess incorrectly.

I am preparing a new version that will allow the user to select the language manually as well, so it should fix this issue. 

Hi jairoaf,

I've just released v2.5.1, which adds the ability to manually set which language the audio is in. This should help you when the automatic language detection doesn't work properly. Note that it does only work with one language per input file.

(1 edit)

Awesome - just tested it and it seems to solve the issue indeed. One thing you might want to know though, is the fact the drag and drop functionality doesn't work anymore (a "forbidden/stop" icon replaces the mouse cursor). I'm on windows 10.

Also, I still have 1 issue with 1 specific video where the automatic transcript stops (it starts repeating the same line 15 times then nothing, despite of 20mins of video left - settings change with the accuracy slider didn't help). But I think I identified a pattern - the previous video where French/no dialog was picked up at all started with a rather strong background music. But it's now picked up properly when French is forced in the settings. That other video I just mentioned picks up the French properly but stop at a moment when again, the video has a stronger music background before the talking resumes. Could the model struggle when unusual music type/noise level is heard? Nothing that I really need to be fixed to be honest (I solved the issue by editing the bit with the music out of the video) but it's just food for thoughts if you ever see someone else is having similar issues and you fancy working on a solution.

Thanks though! I'm happy with the update!