[Blank_User]

Hi every one!
I'm just new here...
sorry to disturb you all but i need help. I've been trying to translate a ps1 game and so far i still don't have enough progress. i tried to translate it using the hex editor. when i load the game in the psx emulator it worked but the the text does not display properly. they are like gibberish characters. i don't understand even a single word. pls help me to fix this.


thank you

[Isao Kronos]

The Newbie Package of REQUIRED Material

ROMHacking.net FAQ: You ask, we answer!
ROMHacking.net Getting Started Section: Newbies Go HERE!
ROMHacking.net Documents Section!
How to ask questions the smart way.
On the Essence of ROM Hacking
Talk with experienced people in our IRC chat and ask specific questions there.

[Blank_User]

thank you for this things
but i still can't figure out how to fix the gibberish txt on the ps1 iso. i did not find any information on fixing a "cavespeak" in a ps1 iso. i hope you could help me. thanks again

[henke37]

You have the character encoding and/or character set wrong. Check what the game actually uses. You must be able to read the text before editing it.

[Blank_User]

thanks for the answer but can you pls teach me how to do it. can you tell how to check the character encoding on ps1 iso.

[henke37]

Blind guessing usually works for standard formats. Otherwise you need to figure out the custom character table. It is quite similar to one alphabet cryptography, but you have both the cryptotext and plain text, making it very easy.

[VicVergil]

Usually ASCII or Shift-JIS for more recent games...
But more often than not for earlier titles, letters do not have standard values like that. You'll have to use a relative search tool like Monkeymoore and search for a given word..
The principle is that it searches a given word in English "egg" in the script, and it searches for a word with has letter 1 with value X, letter 2 with value X+2, and letter 3 with value X+2, (assuming the font is organized "e f g".. there are crazy developers out there who would not do something as obvious, like Capcom and their Ducktales: when in doubt, check the font... with your favorite tile editor) and assumes a set of values for letters and shows you a text with those values. If it's readable, then you got your initial table (well, "egg" is a bad example, you need mixing upper-case and lower-case for better results, and longer words).

You got a TBL file! You can use it with  a hex editor to make the text readable and EDITABLE, and you may notice gaps. Those are missing symbols/special characters you need to complete the tbl file with (while looking to the in-game text to see what they do) using... notepad!
If you want to get dangerous, you could also have a second copy of the game to butcher and experiment on by replacing a given piece of dialogue with some of the values you're still unsure of, just to test them in-game in emulators and see what the symbols are.

Said method works too with Japanese.. (and Korean, for the matter.)

You might want to try searching for a pure hiragana word, then a pure katakana word, but you'll sure avoid kanji like the pleague (at least for now). Wonder what the hell is hiragana etc? Well you should at least have some idea about the language you're attempting.
Hiragana more often than not is organized like "a i u e o ka ki ku ke ko " sometimes it has the ga family immediately after the ka one, or separately in the end and skipping to the sa family.. sometimes it's more like "a small_a i small_i ..."
Looking at the font sure helps.

Why no mixing hiragana and katakana, you say? Well, because some games use the same values for both with a switch before.

So you did figure out the kana? Now try to identify the kanji one by one by filling the gaps again, as in above for the special characters.

And there's the fact that a compressed file in the game binary most likely won't yield any results with this method, so you need to figure out the compression first or you're a sitting duck. Emulator RAM almost always have invariably the decompressed text data, for obvious reasons.

Chinese>English translations are scarce for a reason. It's extremely hard and painful to try and distinguish the actual text from all of the surrounding gibberish, due to the fact it looks exactly the same (no offense to Chinese folks).

[Blank_User]

thank you for all the replies. I'm still having problems in making translations the truth is, this is the first ps1 game that I will translate. The name of the game is Black Matrix 00. i really want to play it in English.
I am bit confuse in making a table for game but i have read some threads that if you can view japanese characters on your computer there is no need to make a table. I'm also having with problems the compression of the games cause i don't know how to figure them out. I tried to extract the files in the .bin file of the game by using daemon tools. all I got are .xa files, .dat files and .cnf files. I can't extract the .str files. can you pls tell me how to extract them. there are other people say that i need to draw my fonts using  tile editor.
anyways thank you very much and i appreciate all the replies

[BRPXQZME]

Chinese>English translations are scarce for a reason. It's extremely hard and painful to try and distinguish the actual text from all of the surrounding gibberish, due to the fact it looks exactly the same (no offense to Chinese folks).

Well, not only is the linguistic point not in any sense true, there is a much more practical reason: there aren’t a whole lot of Chinese video games to translate via hack in the first place.

[VicVergil]

Well, not only is the linguistic point not in any sense true, there is a much more practical reason: there aren’t a whole lot of Chinese video games to translate via hack in the first place.

I did see someone trying to locate the text from Joyland games (Ys Seven PC, Trails trilogy PC) before giving up in frustration because Shift-JIS encoding does make the chinese appear, but with it loads and loads of garbage data appearing, also, as Chinese characters. So my point is more a technical one than a linguistic one.
And making a TBL for a kanji-only text with relative search is a PAIN. Especially if the kanji are in a custom order.

thank you for all the replies. I'm still having problems in making translations the truth is, this is the first ps1 game that I will translate. The name of the game is Black Matrix 00. i really want to play it in English.
I am bit confuse in making a table for game but i have read some threads that if you can view japanese characters on your computer there is no need to make a table.

Actually, that means if you see the text clearly with a text editor supporting Ascii or Shift-JIS (yes, opening a game file in a word processor, as silly as it sounds... I won't advise you to edit it, since word processors ERASE certain bytes that may be crucial to the game, if they're not printable characters)...
that means the game is using the standard hex values for the letters (the same as web navigators).
That means you'll need a TBL with Shift-JIS values and won't need to research the values yourself. (in fact, it's here: http://www.romhacking.net/documents/179/)

But you'll still need a TBL file for the hex editors.
And to add any custom values the game may be using to change the color / do a line break / end the text / Play a music etc, like that for example:
CE 06 = [TextColorBlue]

I'm also having with problems the compression of the games cause i don't know how to figure them out. I tried to extract the files in the .bin file of the game by using daemon tools. all I got are .xa files, .dat files and .cnf files. I can't extract the .str files. can you pls tell me how to extract them. there are other people say that i need to draw my fonts using  tile editor.
anyways thank you very much and i appreciate all the replies

Tile Molester is a powerful tile editor tool for unknown formats..
The font you're searching for is likely an image with two colors (often).
If the font from the Japanese game is incomplete and/or missing English letters, you'll need to do that.

That aside, you'll likely run into size limitations. English language is almost twice space-consuming as the equivalent Japanese text, and other latin-based languages are even more wordy almost always.
You'll need to modify pointers so that you can fit the text and move things around, but until then (and even then actually), hex editing is replacing something with something of equal length... or else it WILL corrupt the game.

How did you know it's compressed, by the way?
Have you tried relative search and/or viewing it with a Shift-JIS table? Are you sure that file is the one with the text?

I suggest actually to do another project instead of this one, just so that you grasp the notion of TBL files, pointers and tile editing in practice with something easier. Magical Hat for the Megadrive, a NES game for example.... Once you feel at home with those notions (possibly even some ASM hacking won't hurt), it should be time to begin your hack effort for this one.

[Zoinkity]

Alrighty Blank_User, the trouble is that more often than not text will be present in the binaries themselves and not alone in its own little bank.
There's a couple ways to go about this.

• You could search each file for a string you expect to be in one of them.  That only requires patience.
• You could narrow down the file by watching which are loaded at what times in an emulator with debug features.  The process is a bit complicated, and would be somewhat different for each debugger, and not all debuggers may be capable of it.

If you aren't even sure how they write strings (as in binary values->displayed text) you may want to start by using an emulator with a debugger and use its memory editor.  Things like name entry screens are great, since you can add and delete a char.  In all but the stupidest titles the name entered will be a string you're adding to, and the deleted char will be replaced by zero.
In a way this isn't much different from cheat searches.  You search for zeroes first, then add a char and filter the search for things changed.  Delete that char, then do another filtered search for the same values changed to 0.  Eventually you can find the string, and with it all the possible chars that be entered into it.

For any standard encoding you can find tables online that will list the value for a given char.  Just compare the values you find against the values in the tables until one matches.  There may be a few special chars added in here and there (like buttons, etc).

Hopefully it uses a standard encoding.  Otherwise, well, you have to work that out too.  Some jerk games will use the order the chars appear in the font itself.  Worst case you'll have to learn enough about ASM to read what the char writer is doing.

...not only is the linguistic point not in any sense true...

True from the linguistic perspective, but anyone who's tried to sort out a ni-he encoded font encoded in arbitrary order would likely say it's a pain in the butt.  Unfortunately, it's rarely the translator that gets the job of muddling out character tables.

[Blank_User]

I don't know if the game is compressed. but i think its not. most of its scripts and game menus are found on the .dat files. I really don't have an idea what the difference between the compressed and uncompressed ps1 iso. so far I don't have much problems in extracting the files in the .bin file. only the .str files that cannot be extracted because my computers says they have invalid MS-DOS function.

I don't know how to start making a table file. My problems with the relative search is you still need a table file to view the japanese characters. I don't know how to use the a tile editor. I tried watching videos about it but most of them teaches how to rip sprites and replace the spites of the characters in a game. can anyone pls teach me how to use one?

Thanks for all the answers. As a beginner in romhacking I learn a lot from all the people. pls continue to teach me all the things I need to learn especially with ps1 iso hacking. I am really willing to learn how to translate. Thanks

[VicVergil]

No, listen...

You initially have only the iso file with you, and nothing else.
You don't know how the text is stored, and what values does it use for letters.

To know those values, you must do a relative search (you write a word that you saw in the game).
The relative search generates the TBL file.
If you don't find anything with relative search then ... logically the text you see while playing the game is in the iso, but you don't find it. That means it's not uncompressed, but compressed.

Sometimes the game uses the values for the characters used worldwide in computers to print text (Shift-Jis, Ascii, etc) and thus when you open it with a text editor it displays Japanese text like in the game.
That means you have your work cut for you, and it's Shift-JIS. You can use a  Shift-JIS TBL.

Oh, and windows doesn't display the whole iso, since the developers use often an uncommon non-standard structure for PS1 isos. You may most likely need to open the whole iso with a hex editor and search...

..............
Listen, it's better if you start with something simple.
At least learn the basics, before doing this game.
Try to modify the text in an English nes game, the font, etc...
Build your experience, and THEN try something like a big PS1 iso. You CAN'T do this if you don't even know how to tackle simpler games.

and read this: http://www.romhacking.net/start/

[Blank_User]

Thanks for the answer but most of the hex editors that display japanese characters does not have relative search and you need a table file to view japanese characters with hex editor that has relative search.

I will try your suggestion to edit simple games like on nes so I could learn the basics of translation.
Thanks

[Zoinkity]

To be honest, most people don't read the japanese directly from a hex editor.  They identify where it's at and extract it.  you don't even have to get into fancy programming or anything; you can copy strings from the file into a new file and save it with the extension .txt.

Hex editors aren't really designed to be an end-all tool.  It's like screwdrivers.  Sure you can use one as a lever and half-heartedly bash nails in with the butt, but you're better off using a dedicated tool for that.

It's probably easier to learn hacking simpler software (and there's certainly more docs and tools), but not everything crosses over to later stuff.  If you have an interest learning NES then by all means go ahead.  If you're only learning it to do PS1, you'd be better off learning PS1.

[BRPXQZME]

anyone who's tried to sort out a ni-he encoded font encoded in arbitrary order would likely say it's a pain in the butt.  Unfortunately, it's rarely the translator that gets the job of muddling out character tables.

Maybe someone is gonna step up and say it’s just because I’m the resident expert, but I’mma go ahead and say it’s not any more of a pain in the butt than it was for whoever set it up in the first place. If you’ve ferreted out the font, you just need someone with kanji literacy and the time to copy the tables by hand.