- Jan 14, 2019
- 69
- 55
Others [Translation Request] The Paradise Fortress of RePure Aria 2 (Initial Version, Updates Expected)
- Thread starter Trolldier Cop
- Start date
- Jul 16, 2018
- 33
- 93
Just to inform you:
The game doesn't use the Wolf RPG engine, but a custom engine made by the game's studio and this engine makes a translation absurdly hard.
If anyone wants to translate it, you have to use DXExtract to extract the event2.dxa file and change the file name to something else like event2.backup or anything that isn't .dxa, while the extracted event2 folder has to stay where it is. This way the game automatically takes the files from that folder instead of the archive.
Inside the archive you will find .txt files that you have to open in Japanese via locale emulator (or else you will only see random gibberish instead of japanese text). You can then translate the text there and save it at the end. However, that's only where the annoying fun starts:
The game's engine is coded in such a weird way, that it uses the letter e as a line-breaker instead of displaying an e in-game, but only if it is on an even bit, else it gets displayed normally.
So for example, if you have the text:
--aaaa--Oh, what a nice daye
In game you will see the text Oh what a nice day, because the e in nice is on the 15th space (after the --aaaa--).
However, if you have something like:
fnaaaa--Then let's go to the mall.
In game you will see Then let's go to th and nothing behind that, because the first e is on the 3rd (odd), the second on the 7th (odd) and the third on the 20th (even) space.
There are two ways of dealing with that and both are complicated as shit:
1. Re-code the game with a hex editor so it uses a different character than e for these line-breaks (something like § for example). I have no clue how to do that, so I can't help you there.
Also, I know that you need to use a hex-editor to change the main menu texts, since these are written in the game's .exe file.
2. Just translate all texts and then go over all of them, counting which e's are on even spaces and replace them with a similar looking character like a double-width e: "ｅ". This turned out to work for RePure Aria 1. It looks bad in-game, but it's still better than replacing it with a capital E (which also works) or WRITING EVERYTHING IN CAPITAL LETTERS (<- see how dumb that looks?).
The game doesn't use the Wolf RPG engine, but a custom engine made by the game's studio and this engine makes a translation absurdly hard.
If anyone wants to translate it, you have to use DXExtract to extract the event2.dxa file and change the file name to something else like event2.backup or anything that isn't .dxa, while the extracted event2 folder has to stay where it is. This way the game automatically takes the files from that folder instead of the archive.
Inside the archive you will find .txt files that you have to open in Japanese via locale emulator (or else you will only see random gibberish instead of japanese text). You can then translate the text there and save it at the end. However, that's only where the annoying fun starts:
The game's engine is coded in such a weird way, that it uses the letter e as a line-breaker instead of displaying an e in-game, but only if it is on an even bit, else it gets displayed normally.
So for example, if you have the text:
--aaaa--Oh, what a nice daye
In game you will see the text Oh what a nice day, because the e in nice is on the 15th space (after the --aaaa--).
However, if you have something like:
fnaaaa--Then let's go to the mall.
In game you will see Then let's go to th and nothing behind that, because the first e is on the 3rd (odd), the second on the 7th (odd) and the third on the 20th (even) space.
There are two ways of dealing with that and both are complicated as shit:
1. Re-code the game with a hex editor so it uses a different character than e for these line-breaks (something like § for example). I have no clue how to do that, so I can't help you there.
Also, I know that you need to use a hex-editor to change the main menu texts, since these are written in the game's .exe file.
2. Just translate all texts and then go over all of them, counting which e's are on even spaces and replace them with a similar looking character like a double-width e: "ｅ". This turned out to work for RePure Aria 1. It looks bad in-game, but it's still better than replacing it with a capital E (which also works) or WRITING EVERYTHING IN CAPITAL LETTERS (<- see how dumb that looks?).
- Jan 9, 2018
- 219
- 204
- Apr 2, 2020
- 128
- 155
Heya!
I've had a little bit of exp with this kinda game. Using DXExtract, you can essentially find the bit code by file size to extract it (you can see it on the forums chat on a certain website!), but using it to recompile is a bit tricky. It pretty much needs to be reassembled using LIVEMAKER, a japanese live novel building program that has multiple uses. Anyways, you should seriously translate the docs and review them before trying it out. Haven't really checked it out all that much.
Hope this helps you! I'm mainly sticking with the rpg engines, mainly because it's easier for me.
- Dec 4, 2019
- 2
- 11
Hey, thought I'd leave my own two bits regarding the translation here.
I actually started working on translating this game back in late summer of 2019, when the game itself was only v1.02. I posted about my progress on ULMF back in the day, but then complications arose and I dropped the translation for a lack of time and motivation.
I can confirm Iceflare's concerns about the difficulty of translating this game - it's quite a hellish task without the source code, but there is oh so much more about the formatting of dialogue that they haven't mentioned - liek the 24-byte limit for lines and line break characters. Recoding the UI and in-game elements is another thing entirely, and I doubt anyone would want to mess around in a hex editor long enough to actually finish it.
That being said, during my time translating, I did make a couple things to assist me in the process, and since I don't see myself returning to this project anytime soon I thought I'd share them for somebody else to take the reins. I'll be attaching them with the message.
The first one is a Python script for handling the formatting of the text files. It has comments describing the whole process and some non-interrupting error handling, but do forgive the piss poor formatting of it all. The script works with the entire directory at once and is applied after you've replaced the lines inside all of the files with translated text.
Though do keep in ming - while it helps quite a bit, it is by no means a perfect solution, so expect it to break in certain places or on stuff that I haven't accounted for.
The second is a doc containing my attempts to decrypt the inner workings of the dialogue headers (the '--aaaa--' -like things at the start of every line of dialogue in the files). I have pinned down the basic functions of those elements, but since my translation stopped at around the beginning of Chapter 1, not all of the characters are listed.
I hope somebody can make a better use of this than I ever managed to do.
I actually started working on translating this game back in late summer of 2019, when the game itself was only v1.02. I posted about my progress on ULMF back in the day, but then complications arose and I dropped the translation for a lack of time and motivation.
I can confirm Iceflare's concerns about the difficulty of translating this game - it's quite a hellish task without the source code, but there is oh so much more about the formatting of dialogue that they haven't mentioned - liek the 24-byte limit for lines and line break characters. Recoding the UI and in-game elements is another thing entirely, and I doubt anyone would want to mess around in a hex editor long enough to actually finish it.
That being said, during my time translating, I did make a couple things to assist me in the process, and since I don't see myself returning to this project anytime soon I thought I'd share them for somebody else to take the reins. I'll be attaching them with the message.
The first one is a Python script for handling the formatting of the text files. It has comments describing the whole process and some non-interrupting error handling, but do forgive the piss poor formatting of it all. The script works with the entire directory at once and is applied after you've replaced the lines inside all of the files with translated text.
Though do keep in ming - while it helps quite a bit, it is by no means a perfect solution, so expect it to break in certain places or on stuff that I haven't accounted for.
The second is a doc containing my attempts to decrypt the inner workings of the dialogue headers (the '--aaaa--' -like things at the start of every line of dialogue in the files). I have pinned down the basic functions of those elements, but since my translation stopped at around the beginning of Chapter 1, not all of the characters are listed.
I hope somebody can make a better use of this than I ever managed to do.
Last edited:
- Nov 2, 2019
- 29
- 278
- Mar 16, 2018
- 272
- 263
- Mar 3, 2019
- 153
- 19
- Apr 2, 2020
- 128
- 155
LiveMaker system glitch: because of its particular setup, for some reason, it cannot make proper use of the the lower case "e." Causes too many glitches when compiling. Not everyone wants to replace with capital E, or Eta, the Greek letter.
Though, there "technically is" a way around this. The dialogue is also pictures rather than pure text.
- Jan 9, 2019
- 57
- 49