Other Maghni AI by VocaTone Studio and Misbah Studios

[morrysillusion]

im glad there is an actual prototype demo-- their first tweet didnt even imply theyd show that at all which brought my initial concerns... not sure why they didnt straight up say they had a demo prior?? bc they would bring hype certainly

anyways-- i def dont mind oliver's sound however, his original provider was a kid wasnt he? and i think what sounds off to me is that it sounds like someone voice acting as oliver rather than actually being a kid? there is always a certain sound to adult voicing children imo. so i wonder if thats the case, idk if its mentioned who is his new provider.

in terms of crowdfund, that is a very small amount to produce not just an engine but also presummedly their voices. i dont recall what solaria's original goal was but she earned about $40k, and thats for one voice. im not sure if vocatone is expecting to get more than that and is lowballing in order to make that happen (its not an uncommon practice to set the goal lower, as it will sometimes make people donate more).

i also agree w pico on the fandom-ness of vocatone's presentation-- audiologie definitely leans on being fandom-y but they do present themselves properly like a company. its always been a thing i noticed w vocatone and i think one issue w it is espcially not reaching a wider audience, an audience that could also support them for this campaign monetarily. because yes, there is a lot of fandom love for characters like oliver but... lets also take into account how many fans dont use the softwares and just like characters. are those people going to back the campaign? when they dont use any of that stuff? they are certainly hyping a lot of fans but i cant not think about that too. it may sound like theyll get a lot of support bc of all the fandom talk but i worry that issue would cause it to fail. and i would like to see it succeed.

 

[pico]

At the rate this is going, I think they’ll reach their goal but I agree with the concern about the total sum. To be fair, we don’t know what licensing looks like for Synthesizer V, but professional studio time is expensive. Even if pandering only to the community is successful crowdfunding, if their goal really is to support this many languages not expanding their reach is a huge missed opportunity.

I hope they know what they’re doing with regard to language support. I seem to recall that early on, one of the goals of collaboration with Misbah was preserving cultures and dying languages. If the results aren’t accurate, more harm can be done than good in these cases.

 

[Wario94]

It passed the 50% mark and it's currently at the 57% mark!

 

[Leon]

Their announcement said they were gonna show off Maghni "in action". AKA a demo, as they clarified when someone asked if there would be one

I guess the idea with the CF is to just get the remaining funds. IIRC the original plan before was to CF the total development and people were unimpressed with the idea of funding a large sum with absolutely nothing to show, so they've been funding things themselves to get a working partially completed editor before launching the CF to get the rest of what they need. I'm okay with that. And at this rate. They'll definitely overshoot :)

 

[PearlStarLight5]

I read the crowdfund thing and I can't help but be off-put by that log-in screen. Is Maghni AI gonna require internet like Ace?

 

[Leon]

Oh, I hope not. Someone should double check if it's gonna be cloud based or not, they said they'll answer questions in a QnA on Thursday

 

[PearlStarLight5]

Here's the screen in question

 

[Leon]

Okay, I asked. Does not require a connection aside from activation! Interesting that it wants to regularly reauthorize, though

 

[Wario94]

Update on the Maghni AI campaign; it's currently on the 79% mark and have 101 backers!

 

[Vector]

Curious as to where they're located. $20K is like...a single engineer's time for maybe three months, and I'm not even in California, where it would be 50-100% more due to the higher cost of living. I could see maybe somewhere in continental Europe, where salaries are much lower, but that still seems like a stretch for a project of that scope.

Miku Expo was an order of magnitude more expensive, for scale. (And since it had a credit roll you can kind of see how many people worked on it.)

Eyeballing project sizes in terms of developer-hours is an occupational hazard.

 

[morrysillusion]

i was just thinking about this as i looked at the campaign but the two other studios they say will be producing voices as well, are entirely new companies. not to say they cannot be trusted but.... is there any further info on them? who opened them, their history? understandable theyre new so they dont have anything to show but. i wonder where theyre coming from. (and as much as keeping the characters a surprise makes sense, i dont think the studios should be a 'surprise' in terms of details lol)

 

[pico]

I hate to phrase it like this but I’m not sure how to put it in a polite way. I think these other “companies” are just friends or acquaintances of the developers.

 

[cocoaCacao]

Hi! I'm the CEO of Altvox, LLC, as shown on the fundraiser page. We nor Soundcake Studios have had too big of a public appearance yet, but please rest assured we're not random friends -- both of the teams behind each company contacted VocaTone Studio in 2020, but were later enlisted to help develop the software to its current point. There isn't too much to share about each individual company, as the members behind them prefer to have fewer eyes on them, but they'll be available for further discussions and demonstrations further down the line. We revealed the names of the companies a bit early as to help drum up hype and reveal the silhouettes for their lineups -- I completely understand the concern, but please rest assured that we're not random friends of the developers, but rather qualified individuals who have been assisting with Maghni's development for many years. Thanks!


ETA: The funding allotted to the programmer is to assist in payment -- we're not planning on utilizing *only* the funds from the fundraiser to help make Maghni, they're just meant to help speed it up significantly :)

 

[nene]

@cocoacacao Sorry if this isn't the right place to ask, but have members of Altvox or Soundcake Studios worked in the industry professionally before? Whether it be developing music program plugins/sound packs/ect. commerically or working for a company that has developed synths specifically?

Likewise, if anyone has worked for an software engineering firm that have produced projects of a similar scope to Maghni? :o

I couldn't find if either company has a formal website (other than GitHub) or a white paper? But I also can't locate any company page for Soundcake at all- so I apologize if i'm wrong. I also understand wanting to keep private!

Thank you so much! Good luck developing!

 

[pico]

Good to know, thanks for reaching out @cocoacacao . Sorry for throwing you guys under the bus but there was rather little context and what I did see kind of reminded me of UTAU fan development projects rather than traditional audio software companies. Echoing nene's comment, also curious if you guys have past work to share that could inform expectations for Maghni AI.

With regard to the fundraising funds "only existing to speed up development" instead of wholly funding it, what does that actually mean? Are Maghni AI's developers funding it out of pocket? Is there a grant?

 

[cocoaCacao]

@cocoacacao Sorry if this isn't the right place to ask, but have members of Altvox or Soundcake Studios worked in the industry professionally before? Whether it be developing music program plugins/sound packs/ect. commerically or working for a company that has developed synths specifically?

Likewise, if anyone has worked for an software engineering firm that have produced projects of a similar scope to Maghni? :o

I couldn't find if either company has a formal website (other than GitHub) or a white paper? But I also can't locate any company page for Soundcake at all- so I apologize if i'm wrong. I also understand wanting to keep private!

Thank you so much! Good luck developing!

Thank you for the well wishes! All members of both teams have formal education in their respective fields, be it computer science, marketing, or art. They're both highly qualified and are deemed so by Maghni's internal development. Barham, the engine developer, is well-trained in both musical development and AI. I can't say much else without disrespecting the privacy wishes of my colleagues. We are, by all accounts, a passion project formed by dedicated and talented members of the community -- we don't have any formal large projects we can share due to being an upstart, but I hope that our work will speak for itself!

Good to know, thanks for reaching out @cocoacacao . Sorry for throwing you guys under the bus but there was rather little context and what I did see kind of reminded me of UTAU fan development projects rather than traditional audio software companies. Echoing nene's comment, also curious if you guys have past work to share that could inform expectations for Maghni AI.

With regard to the fundraising funds "only existing to speed up development" instead of wholly funding it, what does that actually mean? Are Maghni AI's developers funding it out of pocket? Is there a grant?

We've been funding it entirely out of pocket! We're investigating other avenues as well, but a fundraiser is by far the best way to provide the community with the best product as soon as possible (with some neat tier rewards too :D )

 

[Vector]

We are, by all accounts, a passion project formed by dedicated and talented members of the community

Ah, I wondered about that. That does clarify the financial situation, if it's a side thing of sorts. I'm definitely no stranger to pouring hours of time into hobby projects...

If that's the case, it's definitely impressive.

 

[pico]

The confidence on the part of the Maghni AI team is definitely something I can appreciate. It's clear you guys really love this technology and the creative possibilities.

As long as you're here, if I could offer general advice-- it seems like you guys have the groundwork for building rapport with the english-speaking community, and you say everyone is qualified as a professional, which is fantastic. But to put it bluntly, we wouldn't be asking these kinds of questions if the IndieGoGo pitch communicated what you're saying clearly and directly. I appreciate that you guys want to address the community in a direct capacity via Twitter, forum posts, and livestreams, but when you go to the public for money, the underpinnings of what you're describing really should be more clear from the start. I respect the want for personal privacy and trust from the community, but that trust should go both ways, and a fantastic way to do it would be demonstrating the hard work you guys have already done with the engine and being transparent about the plans for your finances.

More specifically, these are the kinds of things I think you guys should include in your IndieGoGo/future escapades of this nature:

• What is the full financial outlook/profile for the project? How much will Maghni AI actually cost? How much are we paying for, and how much are you guys paying for?
• What is the actual nature of the engine technology you are developing? I understand not wanting to reveal specifics if you are developing proprietary technology, but the studies and existing AI frameworks engines are being built off of can help us understand how your product will actually differ from products like CeVIO AI and Synthesizer V in practice.
• How will the purported language support actually work?
• Seeing as the companies involved have no published background in digital signal processing/AI development or have no desire to present their background, are other more loose 3rd parties going to be able to contribute to your platform in the future as well? What will the outlook be for future voicebanks beyond the assumption that deals are happening between new companies behind closed doors?
• I really want to see you guys take this confidence and bring it into your marketing instead of seemingly relying on OLIVER's vocaloid bank and the wordmarks/characters of other companies in your promotional material.
• Finally, it truly does seem silly that your prototype is only featured for a handful of seconds and buried in a video that otherwise means nothing to people outside of the community. That is a huge accomplishment-- why not utilize it more?

I truly am rooting for you guys and I am truly excited to see what comes next but it's just hard as an individual to be outspoken and excited about this when it seems like there are still so many things up in the air. But I really want to be able to get there.

 

[cocoaCacao]

I hope to be able to answer more of your questions in both the livestream tomorrow and as the crowdfund progresses/we release more updates. I'm unfortunately unable to elaborate on anything else you asked. Thank you for all of your advice!

 

[nene]

@cocoacacao

Thank you for responding to what you could! I'm sure I speak for everyone that we all look forward to hearing more from you guys.