Imagine that every word you uttered was spoken in a voice that was not your own. This voice does not match your personality, age or ethnic background and is not unique to you, but shared by many. Around the world, there are more than 10 million people who cannot access their own voice for a multitude of reasons and instead rely on a generic automated voice which is used by everyone with a voiceless condition from Stephan Hawking to young girls, as well as those in between. Although taken for granted by many, a person's voice is an integral part of their identity that marks them as different from others. Dr. Rupal Patel and Dr. Tim Bunnell, a speech synthesis expert, understood this and set out to engineer synthetic voices that were created with the individual in mind.
With minimal funding, Patel and Bunnell set out to create voices that reflected an individual's personality and age. They approached this project with the understanding that at different stages in a person's life, they are bound to sound different. Additionally, in her TED talk, Patel gave a quick overview of speech science to help people understand the basic of what produces our unique voices. She discusses how your voice is a combination of source and filter with source being, "the vibrations generated by your voice box" and these vibrations are filtered by "chambers in your head and neck that vibrate." Patel explains that these chambers filter sounds in order to allow one to produce vowels and consonants.
During her studies of speech, Patel discovered that while many people who had voice impairments could not actually speak, they could control the volume and pitch of sounds they uttered. With this discovery Patel had the idea of recording the source of the person who was in need of a voice and then finding a person who had a similar source in order to utilize their filter. Her idea was to mix the source of the "target talker" (the individual in need of a voice) with that of the "surrogate speaker" the person who donates their voice to be a filter. Patel's idea was to find people with a similar source, age, and gender, to those who needed a voice. She reasoned that this would then help to create a voice for the "target talker" that would most closely resemble the voice that they would have if they could speak on their own.
Although it may seen difficult to find a vocal donor, Patel assured that it was not, as all an individual needs to do to be one is to read a few thousands utterances that VocaliD provides for them. Regarding these utterances Patel explained that, "the idea is to cover all the different combinationsof the sounds that occur in the language." This way the recipient of their voice can say anything, and it will be vocalized in the voice they receive. It must be a wonderful thing for a parent to really hear their child for the first time or for an individual to finally unlock a part of themselves that they never had access to, and you can help these scenarios becomes a reality for someone, with little sacrifice on your part.
You can become a voice donor by visiting VocaliD's site and creating an online profile. At your own leisure, through your tablet, phone, or computer you can begin to record your voice and as you do that discover potential matches for your voice. So, if you ever find you have some free time (students just coming home from college, I'm talking to you) consider helping someone reclaim a part of themselves, by donating a part of you.
