What is Solaris?
Solaris is a project being developed by Eclipsed Sounds and Dreamtonics. It is a new AI vocalist that recreates the voice of singer-songwriter Emma Rowley for the Synthesizer V software programme.
It’s been in development for two and a half years and now Eclipsed Sounds is launching a crowdfunding campaign which you can check out here. So far just under £12,000 has been raised since its launch a few weeks ago.
The idea is to have Solaris be a production tool as part of a larger sound kit. It could remove the need for human vocalists entirely which, although a little concerning for musicians, may open doors for indie producers and songwriters.
It’s hoped the crowdfunding campaign will allow Eclipsed Sounds to fully realise its vision – with the eventual aim of making the AI near flawless and indistinguishable from authentic singing.
Why is it a big deal for the industry?
Solaris is another example of tech industry teams creating uncanny and realistic vocal AI programmes – further indicating that the industry will utilise a mix of human and robot vocals fairly soon.
In recent weeks, for example, we’ve seen more automated vocal software become publicly accessible.
One such programme called Uberduck AI allows anyone with a Discord account to create voice clips for various famous figures, including Eminem, Jay Z, and even Ratchet and Clank. I’ve been having particular fun with it the past few days I must admit.
TikTok has also been forced to update its text-to-speech system after the original actor filed a lawsuit – though some users also discovered other options for celebrities voices that work in a similar fashion to Uberduck AI.
It seems we’re very close to voice AI becoming widely accessible and understood – and all for free. It doesn’t sound too realistic just yet, but it’ll no doubt become more convincing with time.
We could be in for a seismic shift and it could be happening very soon.