Looking for a tlc for the current AstralAir no Shiroki Towa patch

[Pizza55]

I got permission from the translator to edit it but I think it would be a better idea to use a tlc first. The current patch covers the common route.

[Satsuki]

Personally I think it's pretty much hopeless to edit this patch.

Still, do your best....

[Nosebleed]

If all the translator did was machine translate the whole thing I think you need an actual translator because a TLC can only do so much and from what i've seen the scripts are in a pretty bad shape.

 

Good luck though.

[Carronade]

From what I've seen, it wasn't machine translated, it's just that the translator's English is terrible. Maybe you could get something playable with a bunch of editors... Excellent ones.

[Decay]

The translator's english is terrible and his Japanese isn't much better. At this point it would almost be a retranslation, not TLC, wouldn't it?

 

Its worse than Konosora and ask Pablo how much fun he had "TLCing" that one.

[Clephas]

Do the edit first, then have the tlc come in after.  Nothing good comes of editing after a translation-check, since he'll probably have to fix some of your changes, anyway, for quality's sake.

[Decay]

Almost all groups edit after TLC, with the idea that the person most fluent in English should have the final say on what the English text says.

 

Anyways, it seems pretty likely that besides assumptions made to the contrary, this is actually primarily sourced from machine translation and that it's beyond salvaging. 

[Clephas]

In my experience, the final result is generally better when someone who can both translate and edit better than  the translator and the editor is doing tlc after edits... that was what I did, generally speaking.  In small groups, the tlc is usually handling second-line edit as well as tlc, not to mention quality-checking in general...  Editors get so focused on making it sound good that the results inevitably lose too much accuracy, and raw translators generally don't have the mental leeway to consider both at the same time, in any case.  A good tlc can just take the lines, tell where they are wrong and how they would sound best in english while reflecting the necessary accuracy, and perform both jobs at the same time. 

 

There literally is no point in having someone of equivalent skill come in behind a translator as a tlc (machine-translation being out of the question, of course).  

 

TLC is a much higher hurdle for recruitment than an actual translator (since apparently a lot of people have misconstrued ATLAS or Google Translate as being actually of use).  To be blunt, if the tlc is wasting his time retranslating every line, the project should be dropped, period.  Either that or started over, without the first translator.  If a tlc is having to retranslate every other line, that means there was no first pass translation... there was just a haphazard effort with occasional jabs in the general direction of understanding. 

 

Incidentally, you can't use Japanese-natives as translation-checkers, they are worthless from a practical perspective.  Not enough of a vocabulary and understanding of complex usage in English to transfer their understanding of Japanese in ninety-percent of cases.  I do realize that transforming GibberEnglish to English is painful, but it is even more painful to have to deal with Japanese, GibberEnglish, and English at the same time.  

 

Edit:  A curse on anyone stupid enough to use machine-translation on a Favorite game.  For all that they are a nakige/moege company, their narration is complex enough to make machine-translations worse than worthless.  My advice is just to erase all the patch's lines and start over from scratch, if your translator really was using an 'insert Japanese then shits out something vaguely resembling English, which the "translator" messes with until he can pretend he did a good job' engine.  If your translator can't understand Japanese completely without the aid of ATLAS or one of its siblings, he shouldn't be calling himself a translator at all.  Interpreting ATLAS's output DOES NOT COUNT.

 

Edit2: Just for emphasis: Interpreting ATLAS or its siblings' output DOES NOT COUNT AS TRANSLATION UNDER ANY CIRCUMSTANCES WHATSOEVER.  I don't say this to be mean.  For interface translations in rpg hybrids machine translations can occasionally be of use (limited to single words, no particles, no colloquial turns of phrase, no unusual kanji, no unusual usage, etc).  However, they are worse than worthless when dealing with dialogue or narration.

 

Edit3: Ah... I should have just come out with it.  My way of using tlc was basically the way it was used in some 'old-hand' groups that were so obsessed with quality that they would take a month or more on a single anime episode.  It created beautiful works of art in terms of translations, ones that fit perfectly in every conceivable way... and took six months to a year to finish the series as a whole, lol.  Understand, the smaller groups generally just gave up and released after the first row of quality-checking.  However, the larger groups... it gets pretty insane when you get a lot of master-translation-artisans all going over the same ten lines in an episode for three days straight because they can't quite get it the way they want it.  Worse, each one is a 'genius', which means he basically has this ideal version he wants to 'convince' the other two is correct.  Even worse, all three are enjoying their argument immensely and have completely lost sight of the fact that the object of the group is to release, not to argue over how the perfect translation should sound for hour after hour after hour after hour on irc...

 

Edit4: Unfortunately, those arguments are why I retired... whether it was fansubbing or fantl, you get a bunch of 'geniuses' together, they'll either link perfectly, fly apart in moments, or get addicted to arguing with one another for the sake of arguing.  Since the same groups I worked with on fansubbing had members doing vn fantls (including me), you can imagine what happened when those 'geniuses' got together (I basically just got tired of listening to it and made noises while actually playing vns in the background). 

[interested translator]

Before editing and/or TLC, someone should probably turn the current script into readable sentences. I'd be happy to help with that. (I'd be cross-referencing with the Japanese game script, so don't worry about me just winging it.)

Edited December 14, 2014 by interested translator

[Pabloc]

there was just a haphazard effort with occasional jabs in the general direction of understanding.

From what I have seen, this project looks exactly like this. It's FAR worse than KonoSora - it doesn't resemble English language at all, and the TL quality is about as inaccurate as ATLAS (whether MTs were actually used or not is irrelevant, the quality is similar). Editing this thing is absolutely pointless, because if any competent translator will touch it afterwards, he/she will have to re-translate everything from scratch, so edited lines will land in the thrash.

First, it needs a translator who actually knows both English and Japanese (both, not neither, as it i in this case apparently). Then editing, and then TLC.

[Chuee]

Before editing and/or TLC, someone should probably turn the current script into readable sentences. I'd be happy to help with that. (I'd be cross-referencing with the Japanese game script, so don't worry about me just winging it.)

[Decay]

People, please, stop trying to use machine translation.

[interested translator]

From what I have seen, this project looks exactly like this. It's FAR worse than KonoSora - it doesn't resemble English language at all, and the TL quality is about as inaccurate as ATLAS (whether MTs were actually used or not is irrelevant, the quality is similar). Editing this thing is absolutely pointless, because if any competent translator will touch it afterwards, he/she will have to re-translate everything from scratch, so edited lines will land in the thrash.

First, it needs a translator who actually knows both English and Japanese (both, not neither, as it i in this case apparently). Then editing, and then TLC.

[Satsuki]

I was basically going to start from scratch, but I thought having this to look at wouldn't hurt.

You may want to contact the translator himself, since this will be a long long journey~~

[Chuee]

You may want to contact the translator himself, since this will be a long long journey~~

[Pizza55]

You can contact him if you want over at hongfire, but it's probably best to just wait until I'm done editing the common route in a few weeks. I've pretty much marked anything that looks like a mistranslation, and aside from that you'd just need to glance over the translation and the original to make sure it fits and nothing is left out. That's probably a lot better than scrapping everything he's done and starting at zero. 

[Chuee]

Are you editing it or doing a translation pass? I don't want to continue this if someone else is working on it.

[Pizza55]

Editing, I don't know anywhere near enough Japanese to pass on his translations. If you don't want to that's fine, since I'm probably going to try to finish this in 3 weeks. 

[Chuee]

Well, someone voluntereed on doing a translation pass. I suggest you wait since the translator's English isn't good enough to fix the patch with just editing.

[Decay]

Most of it is just grammatical errors which can be fixed easily. Anything other than that I've just gone to the original Japanese text to see if I can figure out what's wrong by looking up an individual word that seems mistranslated. Anything that I can't fix or check doing that I just mark and note what doesn't make sense. I'm not looking to have this be the final copy for him to update the patch with, which is why I wanted to send it to someone else once I'm finished to go over it. 

[Chuee]

It will almost certainly be both easier and better to just start from scratch. It's like when your crappy used car breaks down to the point where it's more expensive to maintain and repair it than it is to get a newer car. Using a machine translation as the base is not going to be a good idea no matter how much it's edited and TLCed. Honestly, your talents are better spent elsewhere, imo. There's no shortage of competent projects looking for talented individuals and it kind of bums me out seeing people trying to fix up a machine translation. 

[Chuee]

So, I'm about a third of the way through with this. If anybody is interested in putting in a good amount of time TLCing this (you'll probably have to retranslate some of these lines and there's plenty more that probably aren't entirely correct) just let me know.