Rondan said:
Hi CatBus
I am a little bored and want to contribute once again 😄
So, I was thinking of doing the SDH subs in swedish, using swedish standards.
…
The version that I have is 12 of this amazing project, is there any mayor changes on the SDH subs in the latest release?
First things first – if at all possible, always base new subtitle contributions on the latest version of this project. It saves me a ton of time, and I’m always making minor adjustments that don’t even make the changelog (“subtitle #33 appears one frame later”, etc). I never simply take a subtitle from someone and put it directly in my project – I always align it with the latest version. For something like this – SDH subtitles don’t have a template like other subtitles, we will have to wing the process a bit, but it’d likely work.
In fact, because it’s been so long since the latest release, there’s a good chance you may want to use my personal current files as a basis, just to make things easier.
Assuming no changes to the actual Swedish subtitles (just adding SDH), the files you’d need are:
srt\original\SW-eng-SDH-full.srt
srt\original\ESB-eng-SDH-full.srt
srt\original\ROTJ-eng-SDH-full.srt
resources\fragments\txt\SW-eng-SDH.txt
resources\fragments\txt\ESB-eng-SDH.txt
resources\fragments\txt\ROTJ-eng-SDH.txt
…plus the regular Swedish subtitles.
The SRT files you could just scan for square brackets “[]” and colons “:” and that would show you where the English SDH cues were placed. Keep in mind that some of these SDH cues may change the timing of the line, so check that as well. As for formatting, I’m mostly just interested in making sure the SDH cues look distinct from regular subtitles. That’s why I use brackets AND all caps. Other SDH formatting styles could work, but… the formatting code in the rendering script specifically looks for square brackets and colons, so it that’s changing, there’d need to be a corresponding code change. We can use different formatting, but whatever we choose needs to be clearly identifiable by my not-very-clever Python code so that it can format the text correctly.
For speaker identification, I only ID the speaker if the speaker isn’t visible/identifiable onscreen AND it’s not obvious from the context who the speaker is (i.e. it’s not the same speaker continuing to talk but no longer visible, or there’s just nobody else who could be saying this). It’s actually not that common, but you can check the English SDH subtitles for where I decided to do it. This causes some interesting situations, where, for example, Yoda’s very first line is spoken offscreen and needs an ID, but you would spoil the reveal if you gave his name, so I just call it “OTHER VOICE” in that instance! If you feel like there are speakers that should be ID’d that I haven’t done this for, let me know – they’d need to be ID’d in any language.
The fragments are where it gets very, very weird. You’ll see that there’s a sync target called “localized” that gets a lot of “[ ALIEN LANGUAGE ]” entries. That’s because, in English, there’s a localized version of the film with burnt-in subtitles, and “[ ALIEN LANGUAGE ]” just tells the viewer that Greedo/Jabba isn’t speaking English. It’s not really applicable to Swedish since there’s no Swedish dub. You’d just add “[ ALIEN LANGUAGE ]” to the full subtitles, like I do in the English *-full.srt files. Then there’s a surprising amount of changing around SDH subtitles for the mono mixes, and so on, and that’s all handled in the fragments.
One thing is that, for the sake of simplicity, and SDH project should be ONLY SDH. If we need to make changes to the Swedish translation, it would be better to do all of that first and only after that’s done to start with the SDH work.
It would be a lot of work, but it could be done. Let me know if you’re still interested, and I’ll share the current files.
