presamp error

[xyompltrine]

when I was making a presamp.ini for my Arabic VCCV voicebank, when I try to play an error message keep popping up saying,
"otoini
traceback (most recent call last):
File "presamp.py" line 129, in main
File "IOpreset.pyo" line 498, in readiniv
File "IOpreset.pyo" line 75, in __init__
IndexError: list index out of range

<type'unicode'>"

then:
"note change

Traceback (Most recent call last):
File "presamp.py", in line 201, in main
File "Anaalias.pyx",in line 552, in Anaalias.alias_chek_loop
(Anaalias.c:11857)
IndexError: list index out of range


how do I fix these errors?

 

[Kiyoteru]

Hello, we were able to fix this issue after a discussion in St. Defoko's School of UTAU! For anyone else who's searching for a solution, here's what we did.

In the VOWEL section of the presamp.ini, it was mistakenly written like this:

a=a,ba,ta,da,t'a,d'a,ka,ga,'a,qa,pa,fa,va,tha,dha,dh'a,sa,za,s'a,sha,xa,Ga,ha,Ha,ma,na,ra,Ra,la,wa,ya=100

This only contained the ending vowel, list of CVs, and loudness. However, vowel entries need 4 items: ending vowel, main vowel, list of CVs, and loudness. The error was caused because presamp was searching for 4 items but only found 3. This is how we fixed the line.

a=a=a=100

For the list of CVs, we only needed to put "a" by itself, because presamp matches characters at the end of the lyric. It doesn't need to have an exact match for the whole lyric.

Another unrelated issue we discovered was that there was no REPLACE section. Presamp tries to merge voicebank settings with its default settings, and in the default settings, there is a REPLACE section that converts romaji to hiragana. However, that was unnecessary for this voicebank because it's meant for Arabic, which most definitely does not use Japanese hiragana. In order to delete that, we added a REPLACE section to the voicebank's presamp.ini.

[REPLACE]
a=a