Feel free to contact me if you need someone to teach you.
At UfB we can coach you, but I will put some general information here for free.
Recording is as simple as record as "ka-ka" from there you take the first "- ka" the middle "a k" and the second "ka".
Unfortunately the resources that Shirahime has linked are. Very-very rough. Lacking lots of necessary explanation.
You really can't learn how to oto a voicebank by simply looking at the samples- it won't make sense to a lot of people.
It would of been better if she explained the process in her own words, rather than linking tutorials from deviant art. Especially because it doesn't explain recording.
I'll explain it briefly here. Remeber, otoing methods work for all voices, you just need to make minor adjustments.
the CV part of a CVVC voicebank would be otoed much like a normal CV.
For the sake of simplicity this will be veey general, you can be more exact than this.
With the example of "ka ka" you have two CV samples in Japanese CVVC, the starting "- ka" for the first note and the second "ka" for usage between notes.
For the starting note:
hard consonants (like k p b t r d and ch) will get an overlap of 30 and that will be spaced evenly before consonant so there is a moment of silence before the sample. Preutterence goes between consonant and vowel
(reference 1)
for fricative consonants (like s sh z v etc) you set the overlap to 35 and cut off amd silence before the consonant.
Preutterence goes between the vowel and consonant.
For fluid consonants (like n m y w etc) you can oto them much the same as the fricative consonants. You may increase the overlap to 40
For second CV:
cut of any of the previous vowel-
hard consonants:
lower the overlap if needed to keep it just before the consonant (no negatives)
Fricative consonants are otoed the same. preutterence should be no more that 110ms, cut of some of the consonant if it is more.
(reference 2)
Fluid consonants are otoed the same. Max preutterence of 100ms.
(reference 3)
The VC samples:
Overlap 85, preutterence 115.
Preutterence goes right at the end of the consonant.
You can use a higher overlap if you want/need- just make sure the preutterence is at least 20ms higher than the overlap.
------------
If you happen to record stop-consonant (where you record as "kak" instead) then oto the hard consonants like "reference 1"
Oto the fricatives like at "referece 2"
Oto fluids like at "reference 3"
-------------
If you have any questions, feel free to contact us at UfB! We will make sure recording goes smoothly.