I would like to introduce our new site - www.subsync.net.
You can use it to sync a bad-synced subtitle in your language based on a good-synced subtitle in ANY language.
Check out this super easy guide:
+ Reply to Thread
Results 1 to 3 of 3
Just out of curiosity:
Does this also work if the number of subtitle entries inside the .srt files differ?
Some subtitles do not translate as much as others do. What will happen in such cases, is there some language/translation matching done or some other 'clever' heuristic, or does the site just copy the time codes from one file to the other?
Hey Selur, thank you for asking.
SubSyncs’ algorithm can handle all kinds of situations: different versions of subtitles, uneven number of lines, shift in times, different frame rates, added lines (such as copyrights and other things translators add) and a mix of the situations above.
The algorithm doesn't perform any kind of translation and doesn't just copy the time codes from one file to the other.
It actually does a 'clever' heuristic analysis.
Unfortunately, we cannot promise a success 100% of the times, but from the intensive tests that we have conducted, the algorithm we wrote has a very high success rate.
In case the process fails, you can always try it again with another pair of subtitles.
A failure can occur if the two subtitles are way too different from one another, e.g. 500 lines in one file and 1000 lines in the other (yes we encountered such cases, true story). Another example for differences: too many added lines between the original lines, for example 4 added lines every line.
We are constantly improving the algorithm and the more subtitles we run into, the more edge cases we encounter and are able to try to fix.
Feel free to post any additional questions, we will be glad to answer.
The SubSync.net crew.