By the way, I don't know if anyone has shared highlights from the Maghni AI stream yet? I caught the beginning when it came out but it was 2AM and the demons told me to go to sleep. I decided to watch it now, and I thought I'd do it over multiple sessions because livestreams can be a little stressful for me but honestly everyone in the team seemed really friendly and I didn't mind watching the entire thing (at 1.75x speed of course,) but either way, here are my notes:
Natalia's cats are named Chip and Pancake.
Third parties will reveal their voicebanks at their own discretion. First party voicebanks will be revealed/demo-ed, if I understood well, whenever they're ready; especially those that should be available on or near launch.
No set project timeline yet (for features, voicebank reveals, and such); maybe in a few months when they're more certain. It should be revealed around March or April.
They are planning to share weekly updates for the crowdfund and to host monthly livestreams.
Re: "when are we getting Aurum and Audine demos?" Demos will come out as time goes on; it seems the VPs aren't done re-recording yet. (Oliver will have his own demo over the time of the CF too, most likely with more recordings than what was shared so far.)
Oliver's old recordings are owned by YAMAHA. Vocatone hired a new voicer; Natalia mentions he is an adult. They chose to work with an adult for ethical reasons. He, as well as other voice providers, made the decision to remain anonymous.
No plans for free voicebanks à la SynthV Lite. They are looking for ways to have people test voicebanks before they make their purchase, but they're worried about having a version that isn't an accurate reflection of the finished project.
Languages such as Spanish or Portuguese, which can vary largely depending on region, should have separate support. I'm definitely not wording this well AT ALL but basically, Maghni AI is ready to deal with European Spanish sounding different from South American Spanish. Canadian French and correct French will also be separate. (<- I'm French and a little jokester)
There are no plans for Standard voicebanks. However, Caleb Long has been working on a """robot mode""" meant for people who preferred the sound of their previous line of products.
They are not ready to show off the program's crosslang abilities, because they don't feel like it's up to their standard of quality yet. Hopefully in the next few months.
Some items are going to be CF exclusive, but others not, such as Library Physicals, which will be available to purchase later.
Base program is going to be free; libraries should be about USD$85 each, give or take, with potential differences for third parties. Vocatone's voicebanks should be $85.
Crossy (art designer) will post artworks on Bluesky, not Twitter.
No TTS plans for now; it would be in the very distant future.
The ability to create your own plugins would also be in a fairly distant future.
tl;dr: while singers are meant to keep their "native" accent, if a voicebank was recorded in multiple languages, it should have a dataset for each accent. For example, Oliver, whose languages are English, French, Japanese and Mandarin, will be able to sing in French with a French accent. When it comes to other supported languages, he would be able to sing in, let's say, Norwegian, but not with a Norwegian accent (or, if he can, it will be generic and not native to his voicebank). You could choose to give him a British, French, Japanese or Mandarin accent in Norwegian. Makes sense? No? Watch the livestream I'm just a girl
Oliver wears shoes in some languages. Do with that as you will.
Merch will be shown very soon.
You will be able to import MIDI, VSQ, UST, SVP files and such into Maghni AI.
Voicebanks will be linked to an account, not a number of activations. You will be able to log onto two devices at a time.
Maghni mobile is in the plans for the future, but it'll require more resources so not for now.
Oliver is slightly older! He's 13-ish.
Ref sheets soon.
The editor will have many different themes, some character based, some based on different accessibility needs, and will be customisable. (The character based themes were inspired by VOCALOID1.)
Customisable phonemes
They really said "vocal synth fans are largely minors" (or basement dwellers?) (would they be wrong?). Either way, in order to avoid piracy, with the knowledge that their fans may not have their own source of income, they're looking to make their product as affordable as possible and are considering Free Limited Trial voicebanks, although I mentioned their hesitations above.
You should be able to input breaths as their own phoneme (assumedly the way "br" works in SynthV).
Vocatone's characters don't have official genders. But in related properties, Oliver uses he/him, Aurum uses they/them and he/him, and Audine uses she/her and they/them. For example, in narrative-led projects, Aurum is non-binary, but as a voice library, he is genderless.
No VST/i support planned for launch, but in the future.
There will be tutorials.
They're planning to have less expensive vocals too; read "lower quality so the price can be decreased, but of course still up to their standard."
I think that's about it. If anything is unclear I'd be happy to try to clarify - that is, if anyone cares about this post. 😎 How did I get famous in my previous fandoms again . Was it the youth? It was probably the youth
They've also answered some questions in the comments of the crowdfund, so feel free to double-check that!
28 notes
·
View notes