Translation The kenshuui Tendo dokuta translation is officially cancelled

[Deleted member 776740]

I really dont want to sound condescending or rude, but I strongly believe OP is a kid. I was young as well when I tried to check up the decompressed files of this game, but had no idea how to work on it/reading and understanding the binary. Wokann seemed to give insights and thats great! This game deserves a proper translation (As its sort of funny how the second game of all things is completely translated to english)

I am a kid and I did try really hard to do this. I did try and use crystaline. I probably would have worked harder on this project if I didnt have to deal with GCSE's

 

[cearp]

if I didnt have to deal with GCSE's

Sorry, I shouldn't have gone by the age in your profile (28 lol), good luck with your exams!

 

[MegalithAce]

I am a kid and I did try really hard to do this. I did try and use crystaline. I probably would have worked harder on this project if I didnt have to deal with GCSE's

Oi buddy, dont feel demotivated at all. You can always try to seek ways to improve and obtain new knowledge, that's the magical aspect of learning something new!

 

[MegalithAce]

View attachment 540876
I didn't examine the other sub-files extensively.
As for the /word/wd_e01_001.zzz file I looked at, it appears to be a simple LZ compression that can be directly extracted using CT2's built-in decompression tool.
It seems that each plot segment has its own independent font and text. As shown in the image, the decompressed file begins with image blocks of all the characters used in that plot segment, with the charmap rearranged accordingly.
Following the font images is the corresponding text content (arranged according to the charmap encoding of the fonts mentioned above).
Based on the information available so far, localization should be feasible; at least the text and charmap of some files can be found.
So don't give up!

Okay, there's something I don't follow.
I remember making a post in the past, about the decompression of the zzz files.
How did you adapted that font into Crystaltile? I've reached the same address and the one you provided in the image, but I can't find anything of use.

Please explain your procedure to follow.

Post automatically merged: Jan 19, 2026

One extra thing, I've seemed to find some resemblance of characters, but I can't find the correct values to display them correctly. Any tip?

 

[Wokann]

Okay, there's something I don't follow.
I remember making a post in the past, about the decompression of the zzz files.
How did you adapted that font into Crystaltile? I've reached the same address and the one you provided in the image, but I can't find anything of use.

View attachment 551833

Please explain your procedure to follow.

Post automatically merged: Jan 19, 2026

One extra thing, I've seemed to find some resemblance of characters, but I can't find the correct values to display them correctly. Any tip?

View attachment 551845

Sorry I haven't been able to access my personal computer while traveling for work these past few days. Let me explain what I did now.
First, after lz77 uncompressing the /word/wd_e01_000.zzz file, we can get a size of 0x6B48 file.

From header we can get these information:
1. number of subfiles is 2.
2. the address of subfile1 is 0x10.
3. the address of subfile2 is 0x5A14.
3. the size of this file is 0x6B48.

Subfile 1 is the tile picture of charmap. you can set it in tile screen of ct2:
start offset: 0x14 (not from 0x10, i don't know what the value 0x00001680 in 0x10 means, maybe some magic value announcing charmap.)
width: 16
height:16
tile form: GBA 4bpp
image pattern: ObjH-1234

Now let's give charmap for each tile from 0x0000.
0000=こ
0100=が
0200=今
0300=日
0400=か
0500=ら
0600=お
0700=世
0800=話
0900=に
0A00=な
0B00=る
0C00=聖
0D00=命
0E00=医
0F00=大
1000=付
1100=属
1200=病
1300=院
1400=す
1500=ご
1600=く
1700=キ
1800=ン
1900=チ
1A00=ョ
……

Then go to subfile2 0x5A14.
There appear to be many script instructions here, and their specific meanings still need to be reverse-engineered and analyzed. Let's skip this part for now and look directly at the text we can understand so far. Load the charmap file we just made into CT2, and you'll see the actual text. If the charmap loading function is not available, it's recommended to find a newer version of CT2.

We can find from 0x5ADA is the first text we need.
……00 00 00 00 01 00
02 00 03 00 04 00 05 00 06 00 07 00 08 00 09 00
0a 00 0b 00 00 80 07 80 01 00 0c 00 0d 00 0e 00
0f 00 10 00 11 00 12 00 13 00 07 80 00 00 04 00
01 80 00 80
-> ここが今日からお世話になる\n
{textcolor:red}聖命医大付属病院{textcolor:white}か\p\n

The character name is an image call, invoked by a script preceding this text whose usage is not yet clear.
By comparing hex values and in-game performance, in addition to the fonts' charmap, we can deduce that 00 80 is used for line breaks (\n), 07 80 xx xx is used to switch colors (xx xx is red when 01 00 and white when 00 00), and 01 80 is used to switch text pages (\p), usually used in conjunction with 00 80.
0080=\n
0180=\p
07800100={textcolor:red}
07800000={textcolor:white}

 

[cearp]

@Wokann thanks for the write up! From that, dumping all of the text shouldn't be too difficult.
it's just annoying that a separate charmap will need to be made for each game episode.
I'm not sure how many there are.

But when adding the English text, things will be simplified a bit because we can just the same "static" charmap since there will be far fewer characters
I think that will be beneficial when adding VWF - each "A" having the same value will make things easier when looking up the width.