Several suggestions

[williameyelash]

Loving the tool, so I thought I'd make some suggestions to make it a little better.

1. I decided to give the Accurate mode a go on a 90-minute movie (1500 lines) for the sake of comparison, and it got 'stuck' at 94%. It was doing something, but after about 15 hours since the job started I decided to stop it. Tried it on a 1-minute clip of the same movie, it was finished in about 10 minutes. So I guess I just had to wait more, but there was no way of telling if the program was really doing something as it was on 94% for 15 hours. Now I know the process is that long, but maybe it would be possible to make it more transparent, e.g. by adding some progress stats (lines/images processed or something like that) so the user has a better understanding of the current progress. Recognizing subtitle content, please wait... is not that informative. It doesn't have to be in real-time, just an update on the progress once every minute or something like that to keep the user posted.
2. You can change the settings (working area, recognition mode, etc.) during an active OCR job, which is a bit confusing. I guess 'freezing' these options for the time being would make sense.
3. The program always starts in the default size window mode. Would be cool if it remembered the last window size/expanded mode.
4. Would be neat to be able to 'trim' the video before starting the job. Could be useful in case you know the opening and end credits of a movie do not contain any subtitles and you don't want the tool to waste precious time on processing these sections (especially in the Accurate mode). The subtitle timecodes, however, would have to take into account the trimmed sections, of course, so there is no unneeded time shift in the end result.
5. Navigation bar is too thin. You have to left-click precisely on the bar to jump to the next/previous frame of the video. A wider bar would make it easier.
6. Timecodes on the playback bar would also be quite useful.
7. Line breaks are often ignored for some reason.
8. Consecutive lines with the same text are merged into one, which is not always desirable. Would be neat to have an option disabling such merges.
9. After installing 2.2.0 and running it on a movie, the project folder was not deleted from the 'output' folder (Keep Cache Data was disabled). After the next jobs were completed, the folder was deleted just fine. Probably a first-run glitch.
10. [Finished] Subtitle file generated successfully 71.73s would probably be easier to understand if it looked like [Finished] Subtitle file generated successfully. The process took 71.73s

Thanks for this wonderful tool!

[kidsryan0321]

I’m having the same issue as you. The progress gets stuck at 94% and never moves. I checked the computer’s background processes, and the CPU usage had clearly dropped, unlike during the earlier extraction stage when the CPU was almost fully utilized. This makes me suspect that the process has already stopped running, yet it still remains at 94%, and no text file is generated.

I tried extracting subtitles from two long videos (both over 1 hour and 30 minutes), and coincidentally both of them got stuck at exactly 94% and completely stopped responding. So I think this issue should be reported. It’s probably not just a coincidence — there seems to be a real problem.

[PJCisMe]

me too