TL Project Shin Koihime Musou Translation Project

[Vorthod Wiler]

couldn't text wrapping automatically be done by inserting a carriage return character at the end of the line n where n is the max number of characters that can fit on a single line within the dialog box? Single byte ASCII characters, of course.

While that is possible in Japane

se, there are some issues that a

rise when attempting such a thi

ng in English

 

That being said, this game actually does do that already despite the fact that they also hard coded in every single line break in the scripts. It was incredibly confusing when I put together my first script and tried it in the game but would end up with a single letter overflowing onto the next line right before I put in my own \n character. There could be some potential to do it in a program by only breaking to the next line when you see the spacebar character, but there also is the problem of where do you split the text into a separate text box to make it sound as seamless as possible if the English line goes longer than four lines? It's certainly possible to get something decent with code, but I think a human eye is going to be best for that kind of job.

[noobsaleh]

I've seen it done seamlessly in another translation project, but that one didn't have an automatic word wrap like you mentioned.

Basically if the nth character is a letter and not a space, a carriage return is inserted at the beginning of the word containing said nth letter.

Once again though, that might pose issues in an engine with automatic word wrap.

[Vorthod Wiler]

Basically if the nth character is a letter and not a space, a carriage return is inserted at the beginning of the word containing said nth letter.

 

I did bring this up in the paragraph after the example I posted, but I'll expand that here. While theoretically that method should work, there is also the problem of what happens when a piece of dialogue exceeds four lines (Which is very common when large English words force early line breaks or when the Japanese text goes over 3 lines). A human can easily tell what a good point to split into a separate text box could be (imagine a four-sentence line. Sentences 3 and 4 might go together very well and it could be better to send them both over to the next box to make the transition more seamless), but since a program can't understand the contents of a paragraph without some very advanced coding, it might choose a poor place to start the split. Not to mention that simply changing a single word to make the line fit into one box in the first place would also be a valid strategy that is essentially impossible to program since any code that can do that without changing the accuracy of the translation is probably a highly-qualified translation program which, as Google has taught us, probably doesn't exist for Japanese. Basically, the work on the code would either be at the expense of quality which I would rather avoid for a game as highly-anticipated as this or it would be far too much work to justify the time it would take away from the translations overall.

 

God, there were a LOT of run-on sentences in there. I apologize.

[Kelebek1]

I should've probably asked this before, but does anyone have a problem with the Japanese font overlapping? ASCII doesn't overlap, just Japanese. I've seen others have the same problem but haven't found a solution to it. Does anyone working on the project know how to?

[Vorthod Wiler]

I don't know why, but for whatever reason, they never normalized the spacing on a ton of the characters, and some things like the ellipses character will destroy the spacing for the entire line. This seems to be a general error in their programming, but I haven't found a solution. The good news is that roman letters seem to work just fine, so in the long run it shouldn't become much of a hindrance. Luckily, it doesn't seem to overlap badly enough to make it unreadable, but that could just be because I've gotten used to it.

[Kelebek1]

Not everyone has the problem though, so it can't be the game itself, otherwise everyone would have it, and I don't think they'd let that go out. Look at all the screenshots on VNDB for instance, they're all fine.

[xellos266]

I only hope the battle titles are in both languages,I mean jap is nice but most battles I know them in chinese.So a double language there would be very nice,just meant it as a sugestion 2 make this game even more perfect 4 all hardcore fans.

[Vorthod Wiler]

Not everyone has the problem though, so it can't be the game itself, otherwise everyone would have it, and I don't think they'd let that go out. Look at all the screenshots on VNDB for instance, they're all fine.

Huh, I've seen so many people complaining about it, I just assumed it was a general problem, but you seem to be right about those screenshots. I have to guess that it's a problem with operating systems since I can't imagine what other unsupported programs it may be loading, but I don't have the time to start trying different compatibility modes right now.

[Vorthod Wiler]

I only hope the battle titles are in both languages,I mean jap is nice but most battles I know them in chinese.So a double language there would be very nice,just meant it as a sugestion 2 make this game even more perfect 4 all hardcore fans.

 

While I believe the current plan is to maintain consistency with the original (which seemed to mostly go with Japanese names), there is likely to be a separate document with Translator notes available to tell you things like Sekiheki=Chi Bi=Red Cliff

[Decay]

Honestly I think the original translation made a mistake in using the Japanese pronunciation for names while romanizing them. I'm disappointed that this project will as well, but I understand the need for consistency.

[Vorthod Wiler]

Personally, I think hearing the voices say one name and seeing the text box say another would be INCREDIBLY distracting. It might work for things that can get full re-dubs like Dynasty Warriors, but for something like this, that simply wouldn't work well even with the whole "mana"/"True name" thing.

[Decay]

Personally, I think I would hardly notice it.

 

edit: To clarify, I probably wouldn't even be able to tell the point at which their name is said in the dialog when listening to it, let alone detect that it's different. I perhaps could have gone through the entire first game without realizing. I imagine most people who have very little to no understanding of Japanese fall into the same boat. 

[MKDude]

As you said, we do have to keep consistency between SKM and the original Koihime Musou.

 

As I'm sure just about everyone knows, we didn't work on Koihime Musou at all. We're keeping consistency with a game done by an entirely different team. Furthermore, with MangaGamer publishing this, we do have to keep consistency with their products. We can only take but so many liberties.

 

Personally, I don't know much Japanese either, but I can very quickly tell when the dialogue doesn't match up with the text. That may just be because I've played a lot of visual novels though.

[Vorthod Wiler]

I apologize to any psychology majors in the audience as this is going to be nothing but speculation based on personal experience rather than anything I've actually studied.

 

So perhaps this is a downside of speaking Japanese well enough to translate but not to the level where I can feel comfortable avoiding English translations, or perhaps it's a downside to trying to learn new words whenever I hear ANYTHING in a foreign language since I seemed to have noticed naming issues even before I took Japanese class, but if I play through a 60+ hour game and never hear the voices say the same names that I am seeing in the text boxes, I'm going to be wondering how good the translation actually is because they clearly have either done something weird with the character names or failed to explain something clearly central to the game. Once that kind of doubt or confusion sets in, people will latch onto every single mistake they can find in the game to further their theory that this is a poor localization of the game even if the mistakes are few enough to consider the release to be rather high-quality. This is going to be increasingly confusing with the "True Name" issue since there is no historical equivalent of those and therefore would remain in their Japanese pronunciations allowing anyone paying attention to voices (as well as many people who are not) to notice them and then ask themselves why only half of the names can be recognized.

Personally, I would rather have a few hardcore fans have to draw their own connections in a story that completely forgoes many portions of the original source material in the first place rather than have anybody going through the kind of experience above as I've had some experiences ruined quite profoundly by doubt directed at the translators.

[Kelebek1]

There are historical equivalents, it's as it says, their true name (called 名). Men in China were given a courtesy name (called 字) when they came of age, and then there were a bunch of complicated rules for their usages. Cao Cao's Zi is Mengde (孟德), Liu Bei's is Xuande (玄德), Sun Jian's is Wentai (文臺) etc etc. The game uses them at some points as well, I definitely remember Cao Cao's Zi being used in a bunch of places. There's a great explanation of them here: http://www.chinaknowledge.de/History/Terms/titles.html - They obvious made up new names so it's not identical to history, but it's the same type of system and the whole two-name thing wasn't intentionally made up by BaseSon.

 

Anyway I'm sure they just use different names because they're all female, and so male names are just weird, and they also want Japanese names rather than Chinese ones. In that respect, the first point does still make things weird, using Cao Cao for a girl doesn't seem quite right even in English (all the more so if you know the real characters).

 

Wouldn't break anything in the novel though, around all those scenes where Kazuto calls them by the name he hears (the mana) and then they get all mad. You could do the same thing you'd just have to make sure you used Cao Cao/Mengde everywhere instead.

 

I'm not too bothered on the character names issue I don't think, but the battle names should definitely be in English.

[Decay]

Anyway I'm sure they just use different names because they're all female, and so male names are just weird, and they also want Japanese names rather than Chinese ones. In that respect, the first point does still make things weird, using Cao Cao for a girl doesn't seem quite right even in English (all the more so if you know the real characters).

 

I think it's important specifically because it's weird. I mean, think about it, this is something that comes up in the original Japanese version for Japanese players. I would say that having these characters be intrinsically connected to the historical male characters and the weirdness that comes with it is a key part of the experience. In the same way that Napoleon Bonaparte is a blond tsundere girl in Eiyuu Senki. If that was something that was part of the original experience, translators should endeavor to preserve it, I feel. 

 

But I'll also accept the counter-arguments about the disconnect between spoken and written dialog, although I personally don't think it would be much of an issue for myself.

[Vorthod Wiler]

There are historical equivalents, it's as it says, their true name (called 名). Men in China were given a courtesy name (called 字) when they came of age, and then there were a bunch of complicated rules for their usages. Cao Cao's Zi is Mengde (孟德), Liu Bei's is Xuande (玄德), Sun Jian's is Wentai (文臺) etc etc. The game uses them at some points as well, I definitely remember Cao Cao's Zi being used in a bunch of places. There's a great explanation of them here: http://www.chinaknowledge.de/History/Terms/titles.html - They obvious made up new names so it's not identical to history, but it's the same type of system and the whole two-name thing wasn't intentionally made up by BaseSon.

 

In koihime, they have their family name, their personal name (the 名), their courtesy name (the 字), and their true name (called 真名). To use Aisha as an example: Family name is Kan (Guan), personal name is U (Yu), courtesy name is Unchou (Yunchang), and true name is Aisha. All four of these names are mentioned in the game, but the True name doesn't seem to actually have a historical equivalent (despite most wikis giving a chinese reading when listing them). So the two-name thing might not be baseson's, but I'm sure the three-name one is. It was probably just to give more girly names to use instead of the historical male names (Even Kazuto himself needs to be taught about true names despite knowing a lot about this period's history and culture)

[Thorgald]

I did bring this up in the paragraph after the example I posted, but I'll expand that here. While theoretically that method should work, there is also the problem of what happens when a piece of dialogue exceeds four lines (Which is very common when large English words force early line breaks or when the Japanese text goes over 3 lines). 

[Vorthod Wiler]

While that would be an excellent solution, I have absolutely no idea how to hack into the actual .exe of the game (we've been going into archive files to change data so far). I would love to get a smaller font size since English doesn't need as high of a resolution to read as Japanese, but I have no idea how to do that at all.

[Kelebek1]

Editing the exe would be one way, but often scripts will support changing font size. As I remember, there are some bigger messages throughout the game aren't there? If you can find one then you should be able to find the opcode to change the size.

[Vorthod Wiler]

Hmm, good point, but I don't remember actually seeing any of those in the game. Even Shunran who shouts quite loudly at certain points of the game will only have her word stretched over the entire four lines as opposed to actually changing the font size. Shunran is actually the most likely candidate I can think of for something like this, so I'm not sure something like that happens in this game.

[MKDude]

Just a reminder that we are still looking for TL Checker. If someone wants to line up script too, we could teach ya how. Those are the two positions currently open, though feel free to volunteer for other positions if you're confident in your skills.

[lotrfan1224]

so I don't need any experience at all? because I don't have any but im willing to lean

[Vorthod Wiler]

For text lining, we don't really need experience, just some extra hands on deck.however, if you feel you can begin to recognize the character names in Japanese after seeing them multiple times that would be a huge help since our current english files don't have the character names included for voiced lines. (also, occasionally a simple line (like a scream) won't be translated so it will be slightly more involved than just copy-pasting english lines over to the japanese script. Recognizing speaker names could help you get things back in line easier after such a piece of dialogue). I would also be perfectly willing to provide assistance to any people interested in this position with the aforementioned naming stuff. I could compile a list of names pretty easily and you could use that if needed.

 

 

If you want to check translations, all we ask is that you can speak japanese well enough that you think you might be able to translate some raw scripts to a readable state if we send you that direction. That level of fluency should be able to catch some of the easier errors in our scripts and will save time when we get a big-guns translator to do their check. My current plan as head translator is more of a quantity approach since quality wasn't working very well (we've been searching for a perfectly fluent translator and that search has failed spectacularly); the more eyes we can have catching errors around the board, the easier it will be when I (or anyone else who pops in by that time) go around and do a final sweep of things.

NOTE: when I say quantity over quality, that does NOT mean I intend to compromise the quality of the final work; I just want as many people searching for errors as I can get even if individually they can't catch them all. To explain my reasoning a bit, I had a reasonably proficient speaker of the language who didn't quite pass the old (much harder) TL checker test but they did manage to figure out a censored reference to a real-world place that other, more generally-skilled checkers were unable to get. Therefore I don't feel bad asking middle-skilled people to join up, they might still be able to catch something that others can't.

[Kelebek1]

Why don't you have the names on there? They're right there in the script.