We're currently having issues with our e-mail system. Anything requiring e-mail validation (2FA, forgotten passwords, etc.) requires to be changed manually at the moment. Please reach out via the Contact Us form if you require any assistance.
Btw, about the flexible funding since no one's mentioned why they're doing it yet:
The crowdfunding page says that if she doesn't reach her goal, they'll try to come up with the remainder. If they can't do it by six months after the campaign ends, every backer will be refunded.
I think generally the people behind this are pretty trustworthy... though that still is a lot of trust to put in them, so I don't blame people for being a little skeptical about it (haven't seen that many, though!). It helps that we know for sure they're working with Dreamtonics though since Dreamtonic's official twitter has been talking about Solaris. Either way, she's doing pretty well so far - hopefully she can keep up this momentum.
i was worried about a flexible goal being the way this would go, as i would of had a hard time beleiving theyd be able to produce the bank if they didnt have all the money. but that is good to hear, that they will refund if they cannot get the rest themselves (and i imagine thatll depend greatly on how much they need to get if they dont reach the goal). so i hope that stays true if they cant reach their goal, because everything else about their campaign seems trustworthy. the project seems good, and im glad to open up the page and see how much progress its made already!
yeah it does make sense the standard test voice they made wouldnt be fully finished (which points to why they have standard as a stretch goal- AI is their goal so why work on it? the demo standard isnt even close to presentable so in terms of being sold). i gotta imagine they prioritized recording the sounds specifically used in the demo song too and nothing else! it is neat to realize its so limited and yet, it sounds pretty alright even so. but it also make me think the final AI voice will sounds very very different (in a good way).
I don‘t know how exactly SynthV works internally, but I think it means they only recorded the phonemes and transitions that they needed for the deo song instead of everything. Or maybe the engine normally uses several recordings of the same transitions, and they only recorded it once
They very likely recorded a much shorter list which only contains the absolute necessary transitions, whereas normally you would record to cover as much as possible. Which makes sense, because if you recorded a full pitch that's essentially a Lite Standard voice. They aren't planning to produce a Standard voice anyway (even as a stretch goal) so there's no reason to do more than the bare minimum for it.
I focused on information that hasn't been mentioned much in previous documentation.
As mentioned, if the team comes close to their goal, they will try to fund the rest out of pocket or via a small business loan.
Standard voicebanks are not cheaper than AI voicebanks, as they usually require the voice provider to fly to Dreamtonics' studio in Japan. They will consider a standard voicebank only if they have the funds and there's community demand. This is an interesting insight into the pros and cons of Standard Synth Vs vs AI - neither is a simple addition.
Their company goal is to expand English vocal synthesis. They all speak English only, so in the interest of quality they are not pursuing voicebanks in other languages (unless there is significant support in another region). They would love to create more Synth Vs, and if they created another character it would match Solaris thematically. However, all future voicebanks (including appends or talk) are entirely dependant on Solaris' success.
Solaris will have exvoice: you can request a custom word or phrase if you donate to the Solaris Superfan tier.
Emma Rowley is extremely enthusiastic about vocal synthesis and its possibilities. She is comfortable with the concept and excited to have her voice on the engine. She rerecorded the standard voicebank 6 times for fun! (Interview coming soon)
They aim to give Solaris a powerful high range, like her voice provider.
The physical box, and all merch except for the acrylic stand, are exclusive to the crowdfund.
Once you buy Solaris, you have an unrestricted licence to use her voice for whatever you wish. Her official illustration can be used up to a profit of $1000 before permission must be sought.
If funded, Solaris will come with some special Synth V plug-ins created by Eclipsed Sounds.
Eclipsed Sounds does not want to profit from any free work, so extra demos, official 3D models etc., will be dependant on the funds they have available.
Everything about Solaris is up to fan interpretation. She was conceptualised as an adult sun goddess, but you are free to change this. She has no canon height or weight (Author's note: why do vocal synths even have those?!)
Solaris will have at least five breath samples.
You currently can't add falsetto to an AI voicebank.
They aim to match their voice provider's range as best as possible.
The item Solaris is standing on in her official art is a fancy sundial. Example
They are still accepting suggestions for crowdfund tiers and social media content you would like to see! (Finances withstanding, no bishoujo Solaris figure yet guys.)
Solaris' voice will likely be very flexible in terms of genre, if they had to pick a specialism for her it would be pop (to match the main genre Emma Rowley works in).
They want to give the community what they want, they are monitoring the community's feedback and will factor it into future decisions.
'What is Solaris' favourite food?' Hydrogen- jk (I laughed)
The name 'Solaris' is inspired by the company name, Eclipsed Sounds, which preceded her.
They are prioritising power over softness, as in Synth V it is easier to make a power voice soft than vice versa.
The story for how she came to be is in the crowdfund rewards.
A great way to help is by making fanart, they will do a fanart spotlight next stream.
They thank everyone for their support so far!
It's been heartwarming to see the support for Solaris. It seems like a lot of people in the fandom love her and really want to see her succeed. I think, right now, there's one big hurdle for crowdfunded English voicebanks:
Based on activity in the Western fandom, from social media to survey response numbers, there's probably about 100~200 or so people in this part of the community who're actively buying new English vocal synths. This means that even though she's been very well received, funding Solaris with the Western fandom alone will be very difficult.
Something I've noticed is that Japanese companies will often market their English banks to Japanese fans rather than overseas ones. It's telling that there are more non-native speakers buying English banks than native speakers. We are basically a very, very small niche.
Yamaha has the money to do extensive market research and analysis. For V5, they chose to pivot completely to professionals. Simply put, they probably realised that even a voicebank that's a smash hit with the fandom will struggle with sales. Appealing to say, only 0.1% of indie musicians outside the community would still be a bigger demographic.
Luckily, Solaris' voice and design have potential to appeal beyond the current community. Her voicetype is popular outside the fandom, her visuals are reminiscent of a lot of popular games and comics. The base concept of a realistic AI singer is interesting and provides a clear function for musicians. Likewise other parts of the fandom, like the Japanese side, could be interested in a professional American singer, as access to them IRL is difficult.
If I were to offer unsolicited advice - and all this assuming it isn't something Eclipsed Sounds is already doing - it would be to go crazy and post about the campaign on absolutely any music community that permits it. It may be met with more scepticism, but you potentially gain a much bigger audience. There's nothing to lose.
It was cool to hear how Emma + her audio engineer friends' minds were blown by the concept of vocal synthethis. Listening to newcomers realise how fun and creative the software could be reminded me of what it's like to discover vocal synths for the first time. It's great to have a voice provider who's so excited and proud to be part of the project.
I really think if eclipse sounds can market Solaris right we can have a influx of western producers using her. I think the technology will finally be there with ai that more people than just vocal synth fans would use her. They just got make an easy way to buy her and the editor and have it in USD and teach people how to use synth v and people will come.
Half of the crowdfunding goal has been reached (20 grand out of 40), and half a month remains until the end of crowdfunding. Honestly I'm a little bit worried.
The box
They also held a illustration contest, with a prize of digital copy of Studio Pro.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
