08-22-2012 #661mamu Guest
i just finished my own extractor when you've posted this
all sound files are easily replaceable, so it won't take long to make an undub when it comes out in the us.
SolBalmung: menu stuff, like item, title and request descriptions are still missing. and most of the user guide too, but if you've played a Tales game before you'll figure it out.
08-22-2012 #662Kenshindono Guest
Mamu I know you are the best. You already looked into this jewel too. Thanks for this, That's a good news. Tales of Xillia is even better than Graces F, The character design is so beautiful
Let's stay patient and when the US & Euro version will comes out. that's gonna be a new challenge to create a great UNDUB version.
08-22-2012 #663ZerotakerZX Guest
Great almost-full-english version will be fine too.
08-22-2012 #664KaInEvIL Guest
08-22-2012 #665clouduzz Guest
I'm more partial to Tales of Destiny myself, story was decent enough on that one to me.
08-24-2012 #666Oxinar Guest
Someone could edit the script of Tales of Xillia for the quickbms extract only those files .SDBJPN
08-24-2012 #667Kenshindono Guest
I think we should wait for a better method from Mamu The dialog or text files segmentation might be bad or worse corrupted. And we need to localize a perfect file if you want to check modification after a translation test Oxinar. I'm sure the tools from Mamu will be more precise and accurate than those I posted from the other devs it was just a POC from a beginner in dev. Also that would need a way to repack those file and merge/inject the modified files into the original archive.
What we need is:
- An extractor tool for the main files
- An hex editor Shift jis (if not text files)
- A Translator crew that want to test and apply their skill on Xillia
- A tool that inject the modified files and replace the existent one in the main/original file
- Cheerleaders (girls)
After that we just have to wait for the US/EURO release of the game to come out next year and try those tools.
Or we can do like Mamu said (easiest way) get the audio files from the JP version and add/replace them with those from the US version of the game. And if the game is playable on CFW for example by decrypting the EBOOT or by replacing the original with the JP EBOOT. That would be way easier and GREAT.
I'm eager to see, now that the JP game EBOOT is fixed, if we can prepare this kind of UNDUB Project for Xillia like you did successfully guys for Graces F.
08-24-2012 #668Oxinar Guest
Thanks for the reply, but I think to translate, the WinHex 16.6 solves this because the few texts that are in the EBOOT are in Unicode UTF-8 then the rest of the text should also be with this coding. And thanks again.
08-24-2012 #669mamu Guest
are here so many modders already? extracting all the files isn't hard, but there is one problem. in the data file, there are about 150 000 unnamed files, without any directories and such. i'm trying to extract them while grouping similar files together. there are file pairs and patterns, but only a few compressed files so it won't be hard to replace them.
also yes, texts are in unicode and in SDB... files. there are actually a lot of language files (without translated text), like portugese and dutch too.
08-24-2012 #670Kenshindono Guest
no prob. Thanks for considering the translation of Xillia and for the information about the coding
Seems like Namco Bandai Prepared the localization version way before they released the game in Japan, but of course they didn't do the job for the translation. It takes at least a year to do so. Now they just need to replace those files with real translated text for their Euro/US version.
I figure out while extracting the main files that the tool prompt for replacing the same file on the target directory. The extractor tool that was created can rename every files with a number. But this way seems weird since we would need to rename them all back and inject them on the main file in the same order and at the same location.
So I thought that maybe all those segmented files can represent one or more larger files segmented in the main container file even for the text ones. Probably to improve memory cache speed and the PS3 might use (marker) to access quickly to the file by loading an adress or an index file that give direction for everything.
My opinion after what I read on Xentax it's that they used the same method with everything. At this address you could see a pics that was rebuild by the dev using each one of those file with their index. The dev have found the way to decipher the texture and to get some sample using three tool and a script. It seems that for every segmented files there is also an index files that give information about the one who follow. This way programmer don't have to make multiple path for the game and use only addresses to access the data.
First It would be really great to localize only the needed one and forget about the others. By grouping text files and ignoring the others we can manage something. TLFILE.TLDAT seems to be the data file and FILEHEADER.TOFHDB might be the main index file and there are probably also other index files inside the data file.
Sorting everything first seems to be the biggest job. Thanks Mamu for your reply and your time on this one too The great modders is you I'm just the curious one.
Hope to be understood too. My native language is French.