There are millions of people who are unable to speak, and as a result they rely on computerized speech generating devices to communicate. One of the most easily recognized people using such a device is Professor Stephen Hawking, a world renown British physicist and mathematician.

Sadly, as unique as each of these computerized communicators is, they often speak using the same generic voices; ones that don’t fit their gender, their age, their accent, or their personality due to the lack of personalized synthesized voices. This mismatch came to the attention of Rupal Patel, founder and director of the Communication Analysis and Design Laboratory (CadLab) at Northeastern University in Boston, and she had an idea about how to eliminate the mismatch, which became VocaliD!

Her passion to create voices unique to each device user, paired with scientific and technological advances, resulted in techniques that can now create customized text-to-speech voices. This is done by blending several hours of donor speech with as little as a vowel sound from the device using voice recipient. This is where YOU come in!

VocaliD is asking each of us to share our gift of voice. Their goal is to gather the largest and most diverse collection of speech recordings from ‘everyday talkers’ worldwide. From this large collection of banked voices, they will create personalized voices for device users at a price they can afford.

Becoming a voice donor is simple. All you need is an Internet connection, a microphone, a quiet place, and a computer or smartphone running the Chrome browser. You will be asked to read or repeat short sentences, and after about 15 minutes you get a visual prompt to select next steps. You can choose to continue recording or return to record at another time. VocaliD needs about three to four hours of speech from each donor. I recorded my first segment, and plan to return regularly to complete my voice file – it really was easy.

To be that voice for another, begin here. To more fully understand the impact of your generosity, view the video below.


Skip to content