Re: Another translation completed

From Dave Shadoff <xxxxxx@interlog.com>
Date
Hi Scot, sorry about the delay in responding.

Each game has its own method of storing the text.  But there are really 
only a few ways, when it comes down to it.  Unicode is never used in these 
games, since it has only gained any sort of popularity since Win 95 days.

(1) SJIS - there are 1-byte and 2-byte storage formats for this.  2-byte is 
seldom used in HuCards, because of the complexity of interpreting the code 
- but it is nearly universal in CDROM games because of the built-in support 
in the CDROM system card.

So 1-byte SJIS is often used in HuCards instead.  However, 1-byte SJIS only 
defines code entries for katakana, in the 8-bit range of ASCII (above 
regular A-Z, a-z characters, instead of the wide variety of European 
characters added by ISO-8859-1).  In these cases, the game might also 
choose to use the regular range for katakana, and replace the 'normal' 
english characters with hiragana instead.  But it can also choose to have a 
special 'switch' character to swap between katakana and hiragana sets (like 
in Lode Runner Lost Labyrinth).

In this case, you would try to search the ROM by using a SJIS-based viewer, 
or a SJIS viewing aid such as NJWIN.

(2) OR, they might simply have their full font laid out in a completely 
non-ASCII format, and completely ignore standard formats for saving (like 
Valkyrie no Densetsu, where 0 = 'A', 1='B', etc.).  In this case, you would 
need to use TMOD2 to find the font definition, and copy it to paper.  Then 
you would play the game until you found some sort of text appearing, and 
copy it down.  You would use the font definition as a 'magic decoder ring' 
for numeric representation of the letters, and you would use some program 
like Relative Search to find this first bit of text (because these letters 
might start with "0", or they might start with "1", or even "20" as far as 
you know).  From there, it is a mathematical transform to the ROM in order 
to view the text and find the rest (and a mathematical transform back when 
the new text has been substituted).

I believe there are utilities already written to help with this, but I 
don't know enough about the translation world yet to know which work 
best.  It seems that the NES and SNES worlds are most developed for this 
type of thing, but for text translations the PCE should be able to 
piggyback on the work already done in both the NES and SNES worlds.

There are a variety of good documents out there on sites such as 
http://www.romhacking.org and http://donut.parodius.com and 
http://rpgd.emulationworld.com .

Please let me know if you find any good utilities for displaying and 
modifying non-ASCII text.

Find TMOD2 and some other tools and stuff at my page:
http://www.interlog.com/~daves .

At 09:45 AM 3/5/2002 -0500, you wrote:
> > > I think I will try something a little more challenging for my next
> > > translation.  ;-)
> >
> > Maison Ikkoku? :-)
>
>Hey guys,
>
>I'll take a stab at Maison Ikkoku... how do I extract and read the hiragana
>from the program?  Is it in a form of unicode?
>
>Lemme know