AI transforms VocaTone offerings

Since winning the People’s Choice Award in the Elevator Pitch Olympics at the Early Stage Symposium in 2022, the VocaTone team has been hard at work developing the latest version of its vocal synthesizer technology.

VocaTone Studio developed a novel piece of technology that allows content creators to access a digitized human voice that can do an astonishing number of things- in 40 different languages! First, the user selects the type of vocalist desired, everything from an operatic voice to a heavy metal vocal style are available. Then the user enters the words the electronic vocalists will sing or speak. With just a few keystrokes, human vocals pour from the computer speakers exactly as the creator imagined it.

Launched in 2021, founders Gregory Grissom, Natalia Shmueli, Diana Outlaw, and KK Jenkins met each other online in a vocal synthesis chat room while still in high school and have been creating innovative technology together ever since.

The use of AI technology has allowed the VocaTone team to recreate the human voice faster and more accurately than it ever has since the project began in 2011.

“It used to take about two hours of recording to just record one pitch of English,” Shmueli said. “We used to record pitch per pitch, and it would be four or five pitches sometimes. Then to develop it would take a month per pitch, but now with AI, we really need just two hours of singing data, period.”

“Now, we also have the ability to control how powerful the voice is, how soft the voice is. We have the ability to control different aspects of the voice. Developing a library is now 30 times faster. We have other new features, like the ability for one library to sing in one language and then sing in another,” she said.

“I wish I could convey how massive of a switch it is for vocal synthesis, “Grissom added. “Before, we would need to have someone come in and record every single combination of every single sound in whatever language we wanted. But now, with our generative AI model, we just need some singing, and we can infer the sound, which lets us generate samples instead of playing back the combination of every note that was recorded.”

“The giant shifts in technology have been life-changing,” he continued.

The company participated in both the gBETA Kenosha accelerator program, as well as the WiSys accelerator, VentureHome. The company found both accelerators useful and was particularly excited about the industry-specific connections made through the WiSys program.

“Wilson Fay has been so great. He connected us with an AI ethics company and set up a framework for us to work with the University of Wisconsin-Parkside Music Department. They’ve been so helpful,” Grissom said.

To expand, VocaTone tapped into its loyal fanbase of current customers for its first round of funding. The company launched a crowdfunding campaign in the final days of 2023 on the Indigogo platform. The campaign exceeded its funding goal of $20,000 dollars in less than one week.

“We’ve been planning this for the longest time and a lot of what we’ve been working on are vocals and languages,” Grissom said.  “With this crowdfunding, we were finally able to show some vocal plans and concepts and what languages we’re going to be launching and prioritizing.”

“One of the most fun things about this fundraiser is finally being able to show what we’ve been working on this whole time behind the scenes,” he said. “These are people (investors) that have been using that same (Vocaloid Oliver) software for years, making stuff with it and consuming the content made with it. With our new software, we are bringing Oliver over, and users are all in. Most of them are people that have been following us the whole time or following the people that have been following us the whole time, so it’s been pretty cool.”

As of this writing, the campaign has more than 200 backers who have invested more than $25,000 to support the development of an enhanced version of the technology. The funding round is still open and expected to attract additional investors before it closes on January 26.

The company anticipates releasing the new version of the software, named Maghni AI, before the close of 2024.

To learn more about this certified Qualified New Business Venture business, and see the evolution of VocaTone software, connect with the company here.