+  RHDN Forum Archive
|-+  Romhacking
| |-+  ROM Hacking Discussion
| | |-+  TokiMemo- First Patch Released
Pages: [1] 2 3
Author Topic: TokiMemo- First Patch Released  (Read 2 times)
mijae
Guest
« on: April 30, 2008, 05:18:33 am »

Hi guys, I came to this site a while back hoping to find hackers in need of translating. But now I'd like to learn some hacking myself.  Then I can go ahead and translate whenever I want!

I've started learning on one of my Japanese roms, and I think I'm doing pretty well for a total newbie. I figured out how to find the text, made up a table for the kana and romaji, and dumped a test excerpt of the script (with crude placeholders for its groups of umpteen million kanji, but I can recognize a lot of stuff from the kana alone and can tell it's working)

Now I need to find the font and I'm having a real hard time here.  I'm not sure if it's because the font is compressed or because I'm not using the right tools. (or because I have no clue what I'm doing...) There's bugger all tools for hacking on my Mac, so I'm using my hubby's computer which runs ubuntu. Some of the tile editors I've tried are in Wine rather than their native Windows and I'm not sure if they're really working right.  The one that seems to work best for me so far is Djinn Tile Mapper.  I could have sworn that near the end of the ROM I could star t to make out the "symbols" part of the font where it has punctuation and numerals from 0-9, but they're always a bit stretched and garbled and despite madly flailing at the tile settings, it never comes into focus, so to speak.

I should be able to try more Windows stuff in the near future once my hubby gets a virtual windows machine up and running again in VMWare, in case anyone has some good stuff to recommend.  Are there tricks I should try before assuming the font's compressed? And if it is compressed, what do I do now?

Or of course, if anyone happens to know anything about this game's font in particular, that would be even better, it's SNES Tokimeki Memorial. (I know, shut up  Wink)
« Last Edit: July 06, 2008, 10:33:41 am by mijae »
RedComet
Guest
« Reply #1 on: April 30, 2008, 06:47:35 am »

Tokimeki Memorial uses a 12x12 font. It isn't compressed, but you'll need to dump it with FEIDAN most likely. It's been literally years since I looked at the game, so I don't remember much else about it. Hope that helps. Smiley
mijae
Guest
« Reply #2 on: April 30, 2008, 01:19:26 pm »

allright, that FEIDIAN program's really helping, I can definitely see I've found something now.  But for some reason, 12x12 isn't working for me.  I've played around with lots of different settings, and oddly enough the one that gets things looking lined up the best is... 16x28?  Shocked   Even that's still off a bit, but it's the only one I've had come out like it's actually in a grid instead of smeared all over the place, and the kana look like they're almost there.

Do games sometimes use 2 different sizes, one for kana and one for kanji?  (although, so far this is the only one I've got looking decent for EITHER group...)

using this command:
./feidian.php -r 16,28,32,32,0x3E8200 tokimeki.smc tokifont2

I got this bitmap:
RedComet
Guest
« Reply #3 on: April 30, 2008, 06:30:30 pm »

Quote from: mijae on April 30, 2008, 01:19:26 pm
Do games sometimes use 2 different sizes, one for kana and one for kanji?  (although, so far this is the only one I've got looking decent for EITHER group...)

That's typically how game do it. Although the kana are usually stored as half the width of the kanji, e.g. 12x12 kanji and 6x12 kana. So, you might try dumping only the kanji. Additionally, you'll want to make sure that you're dumping from the right start point; it's quite possible the font doesn't begin on an even address like $3E8200. Smiley
BRPXQZME
Guest
« Reply #4 on: April 30, 2008, 09:10:15 pm »

It’s been years since I looked at tokimemo for SNES, but IIRC there’s more than one group of kanji.

(I am working on the PSX version, myself. In secret.

I do have access to a complete dump of the SNES version, but I still have no freaking clue if it’s okay for me to give it to anyone – if anyone from j2e can give me a go-ahead, I’d appreciate it!)

And yes, the font is stored all weirdo-warp like that. This isn’t normal, but Konami is never “normal”.
mijae
Guest
« Reply #5 on: May 01, 2008, 05:37:07 am »

Nice to hear that the "weird-warp" is normal, I just tried nudging it over a byte at a time till I saw the next letter scroll around, and the best it ever looks is with that little off-set.  After staring at these tiles so long, I actually started to wonder if they'd always been a little italicized or something and I'd somehow never noticed, so I played the ROM for a bit just to make sure I wasn't imagining things...   :huh:

But anyway!  I found out the reason things looked so funky in the last one is the tiles aren't 16x28, they're 16x14!  I'd been silly to assume the height would have to be equal or more for some reason.



Amazinly, it appears to be the same for the kanji too, at least for a while.  I can recognize the first kanji from the "a" group and some others moving along through the "ka" group.
BRPXQZME
Guest
« Reply #6 on: May 01, 2008, 06:52:34 am »

The characters are not supposed to be tilted like that. The offset is actually half a character-width to the right, and then the halves are swapped (this is why the left half is lower than the right half in what you have here).

IIRC, Tokimemo SNES stores kanji in SJIS order (which lists kanji in phonetic order at the beginning, pretty much). But I’m not checking here, so I may be wrong.
mijae
Guest
« Reply #7 on: May 01, 2008, 07:02:44 am »

Hm, any way to fix that last offset?

As for the kanji, you're right about the phonetic order.  When you name your character at the beginning, you can pick from all the kanji sorted into kana groups that you can page through in phonetic order.  The font here seems to be following that naming table exactly once it gets to the kanji.  But now my tilemap's starting to get a bit messed up again partway through the big list, so I might need to break it into 2 (or more) chunks after all.

And then I need to figure out how to get the edits to work, I just jumbled everything up when I tried to change a couple characters and re-insert with FEIDIAN.

(edited to add: at least things are going more smoothly with the script! I just did a small replacement to test it out. The "stress" stat is now: "GAH"  Tongue)
« Last Edit: May 01, 2008, 09:24:45 am by mijae »
mijae
Guest
« Reply #8 on: May 15, 2008, 05:22:54 pm »

Well I've made some progress!  I've figured out how to re-insert new font tiles without stuffing it up, and tested it out by hacking all the stats in the stat window.

(...would've liked to insert a screenshot here, but photobucket is just not working for me this morning, will have to try again later)

Might need to play with it more, it looks a bit ugly, but I can fix up the cosmetic stuff later. I'm just glad it finally worked.  Thanks for pointing me to FEIDIAN, it's a really great program.

So now I'm back to looking through the script. I was able to find the stat window and some other random stuff like the girl's profiles at the start of the script.  Then farther down, I've found uncompressed dialogue from the endings.  But in between are big chunks of unreadable stuff except for the odd kana and kanji floating about.  Looks like the script is mostly compressed (no big surprise there).

Now, having the endings is actually pretty sweet, if I could make a patch that translates them that would be a nice accomplishment for me in itself.  But while I'm working on that, I'm hoping I can try and wrap my head around this compression stuff too.  I've looked all over trying to find info on previous Tokimeki Memorial translation attempts, but can't seem to find anything.  Any links or advice will be appreciated!
Guadozoku
Guest
« Reply #9 on: May 15, 2008, 05:49:00 pm »

If you suck at hacking and want to get a script dump, try the PC Engine version, the script is in SJIS.
dshadoff
Guest
« Reply #10 on: May 15, 2008, 09:30:43 pm »

I understand that the PC Engine version is actually more than one version (identifiable by a glyph on the rear of the package, in Konami's address: star/dot/dot-in-circle/etc., 5 in all).  These apparently have different minigames and so on.  I ripped a couple of different ones once, and noticed different binaries, but I never played it enough to know what was different.

Anyway, PC-E version's script might be nice for a running start, but maybe not so reliable for a 'definitive' script rip.  Maybe the versions on other platforms also have the same variability ?
Soluzar
Guest
« Reply #11 on: May 16, 2008, 02:18:59 am »

I wish you the best of luck with all this, but remember that the technical problem is only a part of the overall difficulty of taking on Tokimeki Memorial as a project. There's a lot of text, so it's a big project. Don't be disheartened, because this game really is loved by many. It's worth putting the time in, but you should know from the start how long this will take.

I feel I should add that the reason I say this is because I am anxious to see you succeed. It might not have come off the right way, but I'd rather see you know what you're getting into and succeed than become disheartened and quit.
« Last Edit: May 16, 2008, 05:23:30 am by Soluzar »
mijae
Guest
« Reply #12 on: May 16, 2008, 08:41:40 am »

Yeah, I've heard a lot about this being practically a cursed project.  It's been fun for me to learn the basics of hacking though, by trying to get started on it!

I have plenty to chew on now with the uncompressed text alone, but I'm hoping that trying to figure out the techy stuff too in the meantime might help give me a distraction when translating gets too tiring.  I'm also planning to make lots of SMALL patches.  (once I figure out how to make patches...)  If I make it through the uncompressed stuff, my first goals in the compressed text will be translating the "bonus" characters of Miharu and then Rei, who never get in on the standard dating procedure.

oh yeah, Photobucket's finally working again:
("Lit" is slightly dodgy as a translation, but makes a nice short stat for your "book-learning". "Trivia" is much more dodgy, but mangling "General Knowledge" down to some crazy abbreviation doesn't seem right. I think "trivia" works okay when the main way you get it is just chatting with your buddies)

BRPXQZME
Guest
« Reply #13 on: May 16, 2008, 02:10:55 pm »

文系 is more like “liberal arts” or “humanities”.

雑学 may as well be “trivia”.
Djungelurban
Guest
« Reply #14 on: May 18, 2008, 01:45:45 am »

Just wanna point out, if you finish this translation you'll be my hero forever. Cheesy This was even the game that brought translation hacking to my attention in the first place back in... I don't know... Maybe 99? Well, this or Seiken Densetsu 3... And been longing to play it even longer, since 96 atleast. And thus you=hero. Cheesy
Pages: [1] 2 3  


Powered by SMF 1.1.4 | SMF © 2006-2007, Simple Machines LLC