shenmue

[Invité Invité]

Ben non, la table des caractères (standards) est dans le bios de la DC, pas dans les binaires du jeu. Et si elle était dans les binaires du jeu, il est plus que probable que ABCDEFGHIJKLMNOPQRSTUVWXYZ y soient...

[Invité Invité]

ben c'est ce que je sais pas justement avec les consoles eur ou jap si la table des caractères est universelle??? Faudrait faire le test

[Invité Invité]

Si si, les lettres de l'alphabet y sont forcément... Les productions japonaises sont rarement uniquement en kanjis, n'est-ce pas.

 

La table des caractères n'est pas universelle (les ß allemands, les ç français, etc...) mais il est ici question d'alphabet standard (par satandard, on entend "anglais", langue qui n'utilise ni accents ni zigwigwis...). Ce n'est pas bien grave si au lieu de "entrepôt numéro 8", on écrit "entrepot numero 8".

 

Une autre preuve est que quand tu lances un jeux d'origine japonaise, tu n'as pas plein de points d'interrogation ou du n'importe quoi mais bien les kanjis...

[Invité Invité]

Oki alors ça veut dire aussi qu'on peut avoir les caractères japonais ?

[Invité Invité]

oui .

[Invité Invité]

J'ai trouvé ça...

 

.:How-To Translate DC Games:.

Step 1:

Determine if the text used in the game is stored somewhere and that it is not encoded. The

easiest way is to create an image (for educational purpose of course ) or to copy the files to

you harddisk of your PC. Use a editor like Ultraedit. It has a hard time reading a full image

because it always copies the file to a temporary file first. But after it's finished copying, it is fairly

fast. Use the 'find' option to locate a unique string out of the game. Preferably a normal

alphabet-string. Be aware that JIS also has 2-bytes codes to refer to alphabet characters but

these are in most cases as wide as the Japanese characters (that is half as wide as normal

alphabet characters). Also be aware that some text strings can be actually pictures and thus will

not show up in your binary. In some cases the text is encoded and in that case is is really hard to

find. In most games, the text used in menus and in the actual games is seperatly stored in

various files.

If you use Ultraedit to search a string in a (hex) binary, make sure to select the option 'Search

ASCII', else it will not search properly. If you found your unique string, you can like if you can

locate more alphabet text fragments. Most strings will be seperated by one ore more 0x00 bytes.

The easiest part is to just select a portion of the data before and after the location your unique

string showed up and paste it into another file. Change this file into a HTML file by adding: .html

Include you copied section heren and do not forget at the end of the file. Use a standard browser

with Japanese language support to load the file we just created. Make sure the character coding

is set to Japanese SJIS. IE will start whining immediately when you don't have Japanese support

and make you download it. You can now see is some more Japanese text is around the location

where you found the unique string. If it is there, then we are in the right location. If not, then

you'll have to continue searching. In case only the second byte of a JIS character is copied, a

strange characte is show in most cases.

Step 2:

Now we have found some Japanese text (maybe not all of it), we can test if we can alter this

text. replace all the text (maybe leave the zero's unchanged) by the complete alphabet and all

other characters on your keyboard. Replace a much strings as possible. Save your image and

burn it again to CD. (There is no DC emulator unfortunately so we can't try the image on our

harddisk directly). Go/play to the position where you're unique string (you used to search

previously) should be show. If the text you just replaced the original strings with, shows up, then

we can be sure we can change the file without any problem. Some programs use a CRC or other

checking mechanism to prevent people tinkering with their work. In that case the binary neesds

to be dissasembled in order to disbale that check (that's quite difficult).

Step 3:

Get a 'fresh' image with the original strings in it. Located your unique string again and start

working from there. Locate and note every start and end position of an original string. Remember

that they are seperated by zeros in many cases. The most easy way is to copy the original string

to an HTML file (like we just made) and note the positions of is in the same HTML file. You get

entries like: 1c12e4-1c1303 Japanese text you will get a long list of these entries, and you can

also see if the parts you copy will also make sense. If you also load this HTML file in a

webbrowser, then you will only have to save the file when you add something and then you can

hit the reload button of the browser too look.

Step 4:

After locating the strings, we can translate them.

Step 5:

Since we can only replace, we have to make sure that our translation fits in the original location.

Because of the 2 byte JIS coding, you will in most cases have twice as much space for the

normal alphabet. Always leave the zeros untouched since these are used as seperators. You can

enter the translations by hand. But if you have al lot to enter, , make (or find on the internet?) a

program to inject your translation in the specified location. A 'space' is 0x20 and you can use this

to remove unwanted characters.

Step 6:

Burn the patched image to CD and check if you replaced and translated all strings. Repeat the

last steps until you are satisfied or can't find an other text to translated. Be prepared to ruin quite

a few CD's. For a simple translation like the CVS2 fandisk, I spend over 5 CD's to fully translate

the CVS2 fandisk. If you translate for example a fighting-game, try to get hold of a good FAQ

where the original japanese terms are used. For a good online dictionary, try

http://www.alc.co.jp/ You can just cut and paste text into it (do not forget to set it to Japanese

English). If some text is impossible to translate, just try to translate it as close as possible. But if

you just translate 90% of the text, it will be more than enough for non-japanese speakers to

understand. It's not an easy task, and takes quite some time.

 

La, c'est pour la jap et c'est bien sûr beaucoup plus simple de l'anglais au francais... (pas besoin de faire des tests dans un navigateur web)

 

Bon courage. De nombreux projets n'ont pas abouti. Comme le remplacement des voix japonaise de shemnue 2 DC par les voix anglaises rippées de la version XBOX. Un superbe travail de Jeff.Ma mais qui n'a été effectué que sur le 1er des 4 disques.

[Invité Invité]

Par contre la traduc sur les 2 premiers cd de shenmue1 en espagnol sont inpressionant!!!et le 3éme cd est 50% traduit!!!

viva espagne

Quel bonheur de mettre fait le premier shenmue en comprenant tout

[Invité Invité]

Qui s'y lance?

 

En tout cas ça serait sympa surtout pour les anglophobes

[Invité Invité]

Il faudrait tous démarrer de la même image et se répartir les tâches. Parcontre existe t il un moyen d'avoir l'aperçu des caractères modifiés tout de suite ou bien il faut vérifier en testant le jeu à chaque fois? Si c'est ça c'est fastidieux quand même mais on arrive à travailler chacun sur un chapitre ça peut aller plus vite...

[Invité Invité]

c'est comme tout faut tester et surtout etre rigoureux...

 

après c'est selon sa propre confiance en soit... mais si tu testes qu'à la fin et tu te rends compte que le jeu ne passe plus (genre tu as ecrasé des octets qui n'étaient pas du texte) et bien amuse toi pour trouver l'endroit (bon y'a moyen...).

 

quand à tester sans lancer le jeu euh... tu vois bien ce que tu as tapé en lieu et place du texte initial, non? il est pas codé le texte que tu tappes... comprend pas ta question.

 

bref, y'a au moins 2 mois à quasiment plein temps par cd pour faire ça (avec tests); alors faut être réaliste, hein, personne ne le fera

mieux vaut prendre ce temps pour prendre des cours d'anglais si lacune il y a...