Search results

Oh the neutrino banks are 2 banks they talked about how hard it was to make 2 banks that sound like their voiceroids while being distinct.

Out of nowhere Kotonoha Akane and Aoi are getting free neutrino banks for owners of their ' ai voice ' tts. Important it's only for owners of their Japanese 'ai voice' library not English ai voice or voiceroid 2 or synthV A little sad it's not synth v just because their synth v standard...

Sorry for the triple post but more info came out Haiyi's AI is in development and Cangqiong is in beta and kanru is having a well deserve vacation

Another Cangqiong ai english sample I don't know how to upgrade her bank or if it's even out but still

Here's what Cangqiong ai english sounds like https://www.bilibili.com/video/BV1134y167fc

Do we know if she's gen 3 or 4 in these demos

Excited for a lite version since her demos are kinda mixed quality. I wonder how her automatic tuning will sound like.

I don't know what to pick. For me it's all about the voice. I guess it's the tone it's hard to express in words but it's the naturalness of the vocal tone. Like I'm okay if it sounds like a singer drenched in auto tune if the voice still sounds like a human. Like take Elenore standard lite I...

Wow, this is shaping up to be a decent software chinese kiritan sound interesting I wonder how she sounds to chinese natives

I wonder if they use this tech https://google.github.io/tacotron/publications/multilingual/index.html I wonder who they got for the English pronunciation data set to train them if that's the case. They sound nice but a little muffled and hard to understand at first.

Would recommend renaming the xml to sample1 since that worked and retrying

Maki's a Japanese collage student going to school for music. She plays guitar in her band with her friends because she had a throat injury. She recently got vocal surgery and her voice came back but it has a slightly different tone. She's also studying English at college for her foreign language...

Also turn on pitch snap. But if you want the real way to do it Is render out the eh sample and in your daw put it in a midi sampler

Hmmm looks like og maki has a ai bank wonder what's going to happen to it...

If it's a volume thing you can just use a vocal rider on them. If it's a length thing use the velocity parameter. I never really noticed a problem with avanna she sounds great out of the box a lot better than most English vocaloids. On the subject of fukase, his English bank is really high...