Welcome Guest [Log In] [Register]




Welcome to the UTAforum!

UTAU is a Japanese voice-synthesis program used to create singing vocals for music. It is much like the more popular, commercially sold Vocaloid products, but UTAU is freeware!

To learn more about UTAU and various UTAU voices/characters, you can click here.

You're currently viewing our forum as a guest. This means you are limited to certain areas of the board and there are some features you can't use. If you join our community, you'll be able to access member-only sections, and use many member-only features such as customizing your profile, sending personal messages, and voting in polls. Registration is simple, fast, and completely free.


Join our community!


If you're already a member please log in to your account to access all of our features:

Username:   Password:
Add Reply
Kana VB into romaji VB manually; Used if the locale fails
Topic Started: Mar 24 2011, 09:08 AM (1,211 Views)
irei1as
Member Avatar

First let me state that having an incorrect locale (not Japanese) cripples the use of UTAU. So it's always recommended to switch locale when possible.
There are a lot of tutorials around of how to change it, just google them or use youtube for video tutorials.

With a correct Japanese locale if you want to use a kana encoded voicebank (a voicebank with the names of the samples being hiragana/katakana/kanji, such as "あ.wav") with a romaji ust then you only need to do an alias. And even that is made automatically thanks to the awesome tool provided By Aster Selene:
https://sites.google.com/site/fuzzlesanutau/important-links/utau-toolkit



Now, sometimes to do a change of locale is just impossible.
Maybe you lack the administrator rights or you need always another locale to run more important programs.
In these cases it's impossible to use kana encoded banks as the programs can't work well with Japanese filenames.
Uncompressing the voicebank will make appear mojibake and UTAU will show the "Bad file name or number" error.

Typical mojibake after uncropessing a kana encoded voicebank if your locale fails:
Posted Image


The normal solution is ask somebody with a correct locale to do the change to romaji encoding for you...
But not always you have a friend in hand, so let's try a (complicated, lengthy and very silly) method to do it yourself even in our wrong locale.

First a warning: normally in the rules of the voicebank is there a line that say that you can't redistribute the voicebank.
So, don't redistribute the converted voicebank! Use this romaji version only for personal use.

The programs you'll need are: an internet browser with net access (I'll use firefox but IE works too, for example), an uncompressor (I'll use 7-zip but others like Winrar or Winzip may work) and notepad.
Also if you can't read hiragana you can use a table chart like
http://www.tokyowithkids.com/fyi/hiragana_chart.html
but it's faster if you can read it directly.

As example I'll use Tei Sukone's CV as a file being .zip and Pumpkin as a .rar example.
Both are encoded in hiragana but .zip and .rar makes the approaching a bit different.

Be it .zip or .rar create a new folder and call it with a romaji name for the voicebank. For example "Scotteiver1.1" for Tei Sukone as it's real name is "すこっていver1.1".
Then extract the oto.ini into it.
Now open the oto.ini with firefox (File->Open file->Browse where you have that oto.ini or just drag the oto.ini inside firefox instead).
It will appear as something like:
Posted Image

Now, go to, View->Character Encoding->More Encodings->East Asian->Japanese(Shift_JIS). And it will go back the original normal hiragana aspect:
Posted Image

Now, open oto.ini with notepad and start the traslation based on what you can read in firefox.
The line order is the same one so it's only a matter of do each line orderly.
You can save and refresh the saved file in firefox from time to time if following the lines end a bit confusing.

Also there are some slightly troublesome exceptions.

You can see in Scottei there are some special aliases like:
う.wav=u,19,53,34,0,0
So normally it would end as
u.wav=u,19,53,34,0,0
but that would gave a conflict with the previous declaration of u.wav. So it's better to give a different sign to the alias like (notice the added asterisk):
u.wav=*u,19,53,34,0,0

Or you may find kanji filenamas. Those are usually the breath symbol -息1- that you may translate as b1 or even bre1.
Other kanji may be gimmicks of that particular voicebank. Just translate the kanji to it's translation (you can use http://translate.google.com/?hl=es&tab=wT#ja|en| for example).

After repairing the oto.ini into romaji you must transform the wav files that go with it.
Now, it's different if the file is .rar or .zip.

.rar is quite easier so I'll start with that one.
After opening the file in 7-zip you see the filenames are a bunch of squares:
Posted Image
Just copy that name with squares and paste it in any place in firefox that lets paste text.
The search bar, google, the fast reply box of this forum or even the search bar of youtube:
Posted Image
And that's it. You have it understandable.
In the image you see it's "あ.wav" so just unrar that file to the folder we created and name it "a.wav".
And then just go for the next file.

As it's easy to get lost it's better if you do it in one sitting and don't change the order with pushing some of the buttons that regroup the files.

Now the problem comes with the .zip voicebanks as their codification breaks the squares and make it mojibake:
Posted Image
(The mojibake may change depending of locale. This image is for Spanish locale.)
Those signs don't fix the same way as oto.ini with firefox or the .rar file so we need a special feature given by this page:
http://2cyr.com/decode/?lang=en

That page can re-code the mojibake text into the Japanese filename we wanted.
To do so we need to know the code page that runs our locale.
If you know how to use dos commands you just need to execute chcp to get its number.
If you have no idea of DOS commands just run this file I made:
http://www.4shared.com/file/yfYZhwzt/chcp_PAUSE.html
(it just run "chcp" and "PAUSE" so you can read it).
For my computer it's:
Posted Image
That is code 850.

So, in http://2cyr.com/decode/?lang=en
(Click spoiler to see the image that contains the numbers:)
Spoiler: click to toggle


We input our mojibake filename in (1): éá.wav
In (2) we input the Japanese codepage, that is Shift-JIS.
In (3) we input our code page. My code is 850 so it's for ibm850 as shows in the list. ibm850 is the one used in Spanish locale. This code may be different for your computer.
Be sure to press the OK of (4). That is the one of the line of (2) and (3).
And in (5) the real filename will show. In this case it's あ.wav (a.wav).
Then just rename éá.wav to a.wav in the folder we created.
If you have set (2) and (3) already, you only need to do (1) and (4) as it keeps saved.

If using chcp doesn't give you an easy to find number you can use the second menu ("select the successful sample") and search until you find the correct filename for the sound (play the file to be sure it is correct). After pressing OK of that line your code page will show correctly and you can keep using the option like before.

To do the frq files too is up to you. You can always ignore those and make UTAU render them later.

After translating all the wav files you only have to paste the remaining files like character.txt or prefix.map and translate its contents if needed like we did with oto.ini.
Edited by irei1as, Nov 24 2011, 04:07 AM.
Offline Profile Quote Post Goto Top
 
irei1as
Member Avatar

Updated method.
Old method:
Spoiler: click to toggle
Offline Profile Quote Post Goto Top
 
1 user reading this topic (1 Guest and 0 Anonymous)
ZetaBoards - Free Forum Hosting
ZetaBoards gives you all the tools to create a successful discussion community.
Learn More · Sign-up for Free
« Previous Topic · Tutorials and Resources Directory · Next Topic »
Add Reply

- Affiliates

- Chatbox

Welcome to the Cbox! Please be sure to follow the rules of the forum when posting. To use an avatar, simply upload a 45px by 45px image to an online host, and paste the direct link to it in the URL/Email blank next to your name!

• No flooding/spamming.
• No roleplaying in the Cbox.
• Your Cbox username must match your forum username.
(not exactly, but close enough that other people can tell who's who).