How to oto CVVC voicebanks

0.Introduction

This tutorial will teach you how to oto JP-CVVC voicebanks.

Otoing CVVC has 3 steps.
First oto CV.
Next oto VC.
Finally oto only consonants.

Hiragana and alphabet correspond according to this table

1.Oto CV

This section explains how to oto CV and only vowels.

1.1. Beginning VOWEL [necessary]

For example:[- a][- i][- n]...


[Figure 1] [- a]

Offset is at the beginning of the sound.
Preutterance is 0.
Overlap is 0.
Consonant is at the stable point of the sound (stable pitch and stable volume).
Cuttoff is at the stable point of the sound from the right.

1.2. Crossfade VOWEL [not necessary]

For example:[a][i][n]...


[Figure 2] [a]

Offset is at the stable point of the sound.
Overlap is 50-100.
Preutterance has to be the half of the overlap
(For example: If the overlap is 80, preutterance has to be 40).
Consonant is between the overlap and preutterance.
Cuttoff is at the stable point of the sound from the right.

1.3. Begining CV (except plosive)[unnecessary]

Plosive consonant in Japanese:k,t,p,g,d,b,ts,ch,j and r.
For example:[- sa][- ni][- myu]


[Figure 3] [- sa]

Offset is at the begining of the sound.
Preutterance is at the beginning of the vowel.
Overlap is a third of preutterance.
Consonant is at the stable point of the sound .
Cuttoff is at the stable point of the sound from the right.

1.4. Begining CV (plosive)[unnecessary]

For example:[- ka][- tsu][- pyu]


[Figure 4] [- ka]

Offset is soundless point.
Overlap is soundless point.
Preutterance is at the beginning of the vowel.
Consonant is at the stable point of the sound.
Cuttoff is at the stable point of the sound from the right.

1.5. CV (except plosive)[necessary]

For example:[sa][ni][myu]


[Figure 5] [sa]

Offset is at the end of the previous vowel.
Preutterance is at the beginning of the vowel.
Overlap is a third of preutterance.
Consonant is at the stable point of the sound .
Cuttoff is at the stable point of the sound from the right.

1.6. CV (plosive)[necessary]

For example:[ka][tsu][pyu]


[Figure 6] voiceless consonant(k,t,p,ts,ch) [ka]

Offset is soundless point.
Overlap is soundless point.
Preutterance is at the beginning of the vowel.
Consonant is at the stable point of the sound.
Cuttoff is at the stable point of the sound from the right.

[Figure 7] voiced consonant(g,d,b,j,r) [de]

Offset is in the region of before consonant.
Overlap is in the region of before consonant.
Preutterance is at the beginning of the vowel.
Consonant is at the stable point of the sound.
Cuttoff is at the stable point of the sound from the right.

2.Oto VC

This section explains how to oto VC.

Otoing VC has 3 steps.
First,Preutterance and Overlap is determined by the recording BPM in the same of way as VCV.
Preutterance is 30,000/BPM.
Overlap is a third of preutterance.

For example)
recording BPM:100
preutterance:300
overlap:100

recording BPM:120
preutterance:250
overlap:83.3


Next,You should set the preutterance at the end of the previous vowel in the same way that you oto VCV samples.


Finaly,set consonant and cutoff.

2.1.VC(except plosive)[necessary]

For example:[a s][i n][e my]


[Figure 8] [a s]

Consonant is at the stable point of the sound.
Cuttoff is at the stable point of the sound from the right.

2.2.VC(plosive)[necessary]

For example:[a k][i t][e py]


[Figure 9] voiceless consonant(k,t,p,ts,ch) [a k]


[Figure 10] voiced consonant(g,d,b,j,r) [a g]

Consonant is in the region of before consonant.
Cuttoff is in the region of before consonant.

If plosive consonant's VC has sound of the consonant,the consonant sounds double at the crossfade point.

2.3.ending-VC(plosive)[unnecessary]

For example:[a kk][i tt][e ppy]


[Figure 11] [a kk]

Consonant is at the stable point of the consonant.
Cuttoff is before the beginning of the vowel.

If you don't want to crossfade like the end of phrase,you use this.

3.Oto only consonant[unnecessary]

This section explains how to oto only consonant.

3.1.plosive

For example:[k][t][p]
If you don't oto plosice consonant's ending-VC,you can substitute a combination of VC and only consonant like this.


[Figure 12]


[Figure 13] [k]

Offset is soundless point.
Overlap is soundless point.
Preutterance is at the beginning of the consonant.
Consonant is at the end of the consonant.
Cuttoff is before the beginning of the vowel.