VB testing: Kira Hitorine (New VB)

well I'm done with his new voice...yet there is some problems. His voice gets errors time to time and I have to fix alot of it.
here is a new Voice Sample of him singing BRS



http://www.4shared.com/audio/EPvSKPVt/BRS_Test_new_kira.html


Well I take Harsh words...so say it!!!

(Don't be too kind to me...okay?)

Edit:changed the music(no instrumental and shortened)

Hmm, I can't download that 17 MB file right now, uh, sorry. Maybe I'll comment another day when I'm in the other computer.

I guess the advice I could give you right now is to upload the test as mp3 because the quality is almost as perfect and it saves a lot of time and dowload bandwidth for slow conections (or torrent-full)... sorry.
Here you can read a how to do it with Audacity. (Note: in the new Audacity, it's "Export selection" and you choose mp3 as file of type.)

And, well, if you think it's just a test you could also cut the instrumental start and go directly for a sample that has voice at full test. It's the voice what we want to hear in a test, not the cool karaoke, isn't it?

Sorry for not being useful right now.

Meh it's okay..

*gone to upload the MP3 Version*

Changed the Voice sample (no instrumental)

Some of his samples sound really... Demonic. Seriously.

i agree... i like the voice (w/o the demonic sound) cant wait for you to be finished =3

Much better. Less than 2 MB is a cool size to me right now. Thank you for the change of size.

Now the voicebank... I'm confused, sorry. I'm not sure if I like it.
Some notes sound very human and nice while the other ones sound very robotic and extreme.

I'm ok with voicebanks sounding human with some roboticness or VBs being totally extreme (I like TREE even if it's joke example it can end being interesting to hear).
But this one is like a heterogeneous mix in between and it gets a feeling of changing too fast and extreme to make the song being normal. It's like there's a song performed by two singers with very different voices and, even if they're in tune, they alternate after each word.

But if that's what you wanted then ok. I think what you want is more important than the opinions of other random people like me. You're not hurting anybody and you're doing this for fun.
The oto seems quite correct (not sure about "n" and "m" and maybe "sh" and "s" needs shorter consonant but he's not having that problem with the silences of the start of the samples) and he's actually very clear so you can understand very well what he's saying.

I'm not very good giving opinions around and you should wait for more input... but I would aim for one of the types only. Well, maybe make both in a form of separate voicebanks ("human" and "demonic" tone).

smeen wrote:Some of his samples sound really... Demonic. Seriously.

O.O hmmm....now that you mention it...His name "hitorine" would really fit right?
but anyway My original mic(the one I used with the first VB) Broke down.

zinfadel wrote:i agree... i like the voice (w/o the demonic sound) cant wait for you to be finished =3

Thanks but I don't know how to take off the Demonic sound...

irei1as wrote:I'm not very good giving opinions around and you should wait for more input... but I would aim for one of the types only. Well, maybe make both in a form of separate voicebanks ("human" and "demonic" tone).

I don't think I could do that cause it's just an accident when I made this...(plus the working on the oto.ini was very hard cause the waves are very,very messy).


thanks anyway.Hmm...maybe I should save more money for a new headset with a microphone.

The clearer sounds really somewhat remind me of KAITO. o_o;
I have no idea why.
But I like KAITO, so, by extension, I like this UTAU's voice.
The other sounds MIGHT need some fixing, but that's cool, right?

R.K. wrote:The clearer sounds really somewhat remind me of KAITO. o_o;
I have no idea why.
But I like KAITO, so, by extension, I like this UTAU's voice.
The other sounds MIGHT need some fixing, but that's cool, right?

Kaito? =_=

Anyway, yeah this might need some fixes though.

try deleting your .FRQs and remaking them...

Remake? hmm...sounds like a plan.Okay I'll do it!

Yeah, that's what happens with UTAUloids with the wrong frqs. The sounds become demonized. I've had that happen to me with Saki Awane, while she was singing Only my Railgun.
|D Just re-make the frq files, it should be good as new.

I really really like how he sounds in the clear samples! It's so clear and human-like~

LupinAKAFlashTH2 wrote:Yeah, that's what happens with UTAUloids with the wrong frqs. The sounds become demonized. I've had that happen to me with Saki Awane, while she was singing Only my Railgun.
|D Just re-make the frq files, it should be good as new.

I really really like how he sounds in the clear samples! It's so clear and human-like~

I'm done now all I have to do is to make him sing again..


oh yeah there are times that when the FRQs are being made, the Utauloid engine suddenly stops and then an error occurred....maybe thats the cause of bad FRQs.