For values, oto theory goes like this.
10000/tempo For overlap.
10000/tempo x 3 for preutterance. The rest of the values are adjustable for looping.
say I recorded at 120 BPM, I would do:
10000/120=83.33333 <- This will be my overlap.
I take that value, 83.3333 and multiply that by 3, I get 250 for my preutter.
This is the equations that base UTAUloids use, such as nagone mako and namine ristsu.
If you do not have a tempo, create an average tempo with your samples. Say you recorded [kakakikakukeka] at 136 and [sasashisuseso] at 142 and [ririrurarerire] at 129. You can create an average by adding them together. (equals 407) and divide that by how many samples you got the tempo from (which is 3). So 407/3=135, then use the equation up above, or use the auto generate option with set param and use the 135 tempo average. The more samples you do, the better your average will be, the better the average, the better the result. This alternative method is only if you have recorded in a somewhat consistent tempo and isn't recommended for voicebanks with extreme variations in tempo. (i.e. 120 tempo for one sample and 150 for another.)
[doublepost=1587685269][/doublepost]The best thing I can recommend to you for recording a VCV are:
Record with a guide BGM/tempo.
Practice singing and warm ups everyday before you actually record.
Understand your limits while recording to prevent damage to your voice.
Study study study equipment, mic position, and sound deadening techniques.
Study Japanese pronunciation.
Have a goal for the voice and what genre(s) you want your voice bank to work with.
Plan a scale to record in. (say if you record in C major, you would do C, E, and G.) This helps with consistency in tone and common chord progression in certain types of genres (like C major for Pop songs.)