Ever wonder what it would be like if you could “talk” to your great-great-great-grandchildren? Or listen to Frank Sinatra sing a song that has yet to be written?
That’s the promise of voice cloning — the next generation of text-to-speech technology that could replace the robotic, emotionless computer voices that dominate today. The digital process aims to capture and computerize a person’s distinctive vocal qualities in order to create entirely new speech.
Israel-based startup VivoText is at the forefront of voice-cloning technology. “We are focusing on that conversion from written text to spoken text to be as humanlike as possible,” says founder and CEO Gershon Silbert.
A former concert pianist, Silbert doesn’t believe the robotic voices we are familiar with today (Hello, Siri?) are the future of synthesized speech. Rather, he thinks that computer-generated speech expressing emotion will raise the bar. “The human voice to me is like the ultimate instrument ... to convey anything you would want to convey,” he says.
So how good does a cloned voice sound? According to Silbert, “good and getting better.”
Neal Conan — anytime, anywhere
Just ask recently retired NPR journalist Neal Conan. He might be the first broadcaster in history to have his voice cloned, thus allowing him to deliver a newscast anywhere, anytime. Basically forever.
[Related: NPR’s 'Talk of the Nation' is signing off]
“One truly intriguing part of this project, if this technology works as well as we expect it to, is that essentially my voice could live forever,” says Conan.
Conan, who is known for his soothing melodic voice, spent over nine hours in a sound booth recently. He recorded 2,000 sentences, with various inflections, that VivoText uses to record every nuance his distinctive voice can make.
One of the first uses of his cloned voice is “Project Kronkite.” The ultimate goal is to have Conan’s voice deliver essentially anything the user chooses, from personalized newscasts to wake-up calls or even weather checks.
“I was talking to my daughter and describing the technology, saying you can have my voice synthesized, read articles that you’ve chosen and downloaded to any length you want played on your commute or on your jog. That’s the first time my daughter has been impressed by anything I’ve done in quite some years.”
Sinatra back in the recording studio?
The future of voice-cloning technology holds infinite possibilities, even allowing deceased celebrities such as Elvis or Sinatra to, in effect, speak again.
“To clone Frank Sinatra’s voice, we would actually go through his recorded legacy. Now we have the voice library of Frank Sinatra ... and then we can use it,” says Silbert.
The technology could also usher in a new era of celebrity endorsement as stars with distinct voices find new creative opportunities.
“If you go, for example, to Morgan Freeman and suggest he clone his voice, he has to agree to it. I’m sure there are tons of things that come through his agent, and he only does a tiny fraction of it. And of course if there is the ability to have it done through technology, he can do much more,” says Silbert.
As voice-cloning technology continues to improve, no one can be quite sure of what these voices of the future could be saying. Perhaps even a bedtime story for your great-great-great-grandchild.