i cant be the only one side eyeing how sexualised the art in uge's demo was for no reason right? especially with how young the character looks 
anyway stan ci chan for clear skin
anyway stan ci chan for clear skin
VOCALOID Vocaloid ß-Studio
- Thread starter Granata
- Start date
-
- Tags
- vocaloid 6
I'm surprised by how much I like Ci, definitely not the voice I expected from her design, but in a good way. Because most VTubers are legally obligated to look 13, it made me underestimate how experienced her singing would sound 
Aaaaa the all sound and look so good, can’t wait for them to get released
Interesting...
Adding support for Piapro NT files is unexpected and intriguing. I wonder if this is just a convenience for users, something driven by competition, or a sign of Crypton's interest in V-beta.VX-β ver.1.1 will be released on 9/26 (Tuesday). The main additional elements are as follows.
・Added 4 voice banks
- Hana Kanade Kanon β
- Ando Yuge β
- Ci-chan β
- Kasukabe Tsumugi β
・Added a corresponding format for external singing voice synthesis sequence files
- .vsqx
- .ppsf (NT version only)
VX-β バージョンアップ情報公開
2023年9月26日 VX-βの新バージョン(Ver1.1)がリリースされます。 新バージョンでは、新たにボイスバンクを追加、読み込み可能なシーケンスを追加、バグ修正などのアップデートが含まれます。 VX-βをお持ちの方はこの機会にアップデートください。まだVX-βをお持ちでない方も、順次抽選して参りますので申し込みの上、お待ちください。vocaloid.beta.yamaha.com
A Miku V6 Beta would probably break the fandom at this point lol. Meanwhile, I would just be glad to hear an official Miku AI. As of now, those AI cover creators on YT keep (unofficially) beating Crypton to the punch.
I would be, too
I'm a little scared of the 'would probably break the fandom' part, though
These rvc covers make me beyond uncomfortable tbh & I avoid them at all costs. Some voice providers made official statements that they don't want to record AI vbs. There are probably a lot more who dislike the thought of having a hyper realistic voicebank fearing it might take their job. Idk it just feels very wrong to me
There's something about these covers that I've noticed: the AI voices used still retain the original song's vocal timbre and other qualities, which is because the original song's vocal stems were used
Take this cover of Rihanna's 'Diamonds' for example:
Plankton's voice still retains Rihanna's vocal timbres and other qualities; you can hear it in the 'bright,' 'tonight,' and 'you and I.'
Rihanna's singing in 'Diamonds' was based on Sia's original demo, and it shows (but not in a bad way):
So the covers are akin to painting on top of an already-painted wall (and in Diamonds' case, painting on top of a wall that was already painted two times), in my opinion
P.S.
If I'm wrong about certain assessments and points, please feel free to correct me
Very unpopular oppinion but stuff like rvc/sovits/diff svc vs engines that require otoing/labeling (utau, vocalsharp, nnsvs etc) is giving ai generations vs real art to me. All those voice converters may sound realistic and all, and have a wow effect upon first time hearing them, but they require very little effort compared to utaus and others, and to me just seem… soulless. Same goes for ai art which may seem impressive at first, but is largely soulless and doesn’t require much effort. So yeah, I’m ok with people making rvc of themselves and stuff, but it just generally seems weird to me that quite a few people are trying to present them as smt superior to atrial singing/talking synths
EDIT: sorry if this a bit off topic
EDIT: sorry if this a bit off topic
On one hand, I do understand this, but on the other... well, for someone like me that has no clue how to label (and is way too afraid to commission someone as they'd have to hear me sing, oh, the horror!) SVCs are very useful for a person who wishes to create an AI voicebank of their own voice, but it's when consent comes into play that things begin to change. There's a big difference in using it because you want to be able to use somebody else's voice that you don't have access to, and wanting to use it in the same way you would by recording and using your own utau... basically, inherently they aren't something bad, but it's all dependant on what they are used to do, of course.
Oh, wow, we're ALL really getting off topic, huh?
To remedy this... I wanna talk about how impressed I am with the vxbeta! I didn't expect us to get MORE voicebanks when there's already a good amount of them, but I also didn't expect the quality of all the banks/the engine itself to be better than, you know, the COMMERCIAL VOCALOID OPTION THAT IS CURRENTLY FOR SALE? On one hand, I'm excited to see how the beta program - and what people end up using it for/how they use it - changes the future of the Vocaloid program... but man, spending all that money on the V6 engine for middling quality when Yamaha are already trying to dabble in their *next* generation is a liiiiittle bit of a kick in the nuts.
If we're lucky, V6 might be the last iteration of the engine, and they might just start to update it much like how SV does, rather than... you know, make us pay for Vocaloid 7 in the next, what, 2 years?
EDIT: realizing seconds after posting that i did not really talk about how impressed i was, and moreso how NOT impressed i am with v6. it was worth a shot.
Voice providers made official statements? Really? I didn't know that. I knew that about some cartoon/video game/anime dub VAs but...
Oh, also talking about AI VBs in general huh? I think Nostraightanswer said something about not wanting a DEX AI for example
the miku rvc covers arent very ethical but i do enjoy seeing them as like a guilty pleasure, since most of the people making them are just teenagers/early adults messing around (i think monetizing this stuff is the issue...since theres alot of grifters and conmen in the ai-sphere too)
to get back on topic though i do hope miku atleast gets the treatment vocaloid did to some other banks (zola project i think?) where they just port her samples over into ai..itll be better than NT lol
(edit) i forgot to refresh the page and my reply is really late omg T_T
to get back on topic though i do hope miku atleast gets the treatment vocaloid did to some other banks (zola project i think?) where they just port her samples over into ai..itll be better than NT lol
(edit) i forgot to refresh the page and my reply is really late omg T_T
i used to mess with the stuff as a guilty pleasure and to mess with ideas + get inspiration. been harder to do since recent provisions (colab banning rvc, kits.ai's new restrictions, rvc not working on pinokio when i tried it, etc.)
I think is a little of both things: Crypton always said they they would keep very good relationship with Yamaha Vocaloid team for future developments on both sides, and I really think this beta program can help.
Also, when you think that the next update of Miku NT will be released around march 2024, same period of this beta program ends, mmmm... maybe some features might be here and there..
Haiyi and Shian's vps made such statements and that's why those two are the only Quadimension vbs that aren't announced for getting an AI bank. Hiroto-P also once said something among the lines of "A CUL AI update would be so cool but then we'll need a new vp". Didn't knew about Lupin before but ig he's also part of that gang
I dunno if Saki herself said anything but I believe Crypton also didn't want Miku AI since it would sound too close to Saki as opposed to Miku's unique sound (they'd probably be more for it if they could manage to make her sound Miku-like enough)
a more rock-orientated usage of ci..interesting!
Okay, Ci-chan's officially won me over despite the terrible impression I got from Mori Calliope's Graveyard Shift
Will probably use her along with +0.400 gender Rikka AI if I ever get the invite from Yamaha
EDIT:
Added '+0.400 gender' and 'officially'
EDIT 2:
(Moved from this thread to a profile post and back again because I can't make up my mind. If this is redundant and unnecessary, I'll take it down)
Just looked up Ci's design again and my God, this woman is petite.
About a few inches shorter than Ariana Grande, presumably
I remember a comment about Lola Brooke that said something along the lines of her having a 'big voice in a small body'
And that fits Ci pretty well
Also, I wasn't sure if it was Fra making Ci look shorter than she really is, but I saw a picture of both of them with Mori and
Ci's head reaches up to Mori's chest, and Mori's 167 centimeters tall
Anyhow, I'm not sure why I went on this random tangent, though
Last edited:
not sure if i should share some of the sorts of demos/closed beta uses in here? havent seen many posts on the forum for any of them. but inabakumori posted an original song using it