(For a personal account of the voice-banking experience, see companion article, “In His Own Words.”)
Every human voice is unique—a treasured part of a person’s identity. Yet for decades, people with verbal disabilities had to make do with generic text-to-speech voices. The earliest versions were painfully mechanical-sounding; even after they became more humanlike, options were limited. No one really wants a second voice that “doesn’t sound like me.”
Voice Banking Matters
The solution was voice banking, which uses recordings of original voices to create personalized text-to-speech voices. Voice banking has become more widespread and affordable in the past few years, thanks largely to advances in artificial intelligence.
In fact, you may already have a voice-banking option in your pocket. In September 2023, Apple released Personal Voice—the first, and so far the only, voice-banking tool included in device accessibility settings at no extra charge.
Personal Voice developers at Apple relied heavily on input from the ALS community. People who are losing their natural voices due to ALS, muscular dystrophy, or other muscle-degeneration disorders are particularly interested in continuing to talk like themselves after physical speech is gone. But voice banking is an option even for those who are already nonverbal: recordings from pre-speech-loss days, or from close relatives, can work almost as well as recent recordings from the actual user.
Personal Voice and Other Banking Tools
The one drawback to Personal Voice is that it doesn’t work with devices made before iPhone 12, or with operating systems older than iOS 17. Otherwise, it’s found in the Settings app under Accessibility–Speech. The original version required reading 150 sentences of text prompts for about 15 minutes, then connecting the device to Wi-Fi for several hours of AI processing. Nine iOS updates later, the same job took 10 phrases and 60 seconds of processing.
To date, Android has nothing comparable. There are, however, downloadable apps with voice-banking features. Here are a few options (note that most require a purchase or subscription, so check what your health insurance covers).
- The Voice Keeper: Can build a voice from five sentences of input. The brand focuses largely on children’s voices and emphasizes “privacy by design” (no company-maintained voice archives, no direct company connections to personal devices). Users with PRC-Saltillo devices can get free access: others pay $4.99/month.
- Voice Back: Created by an independent developer for a relative who disliked the “robotic” sound of other voice apps, this app is free to download and trial. Ongoing use requires either a $23.99 one-time payment or a $2.99/month subscription.
- TD Talk and TD Snap: Not actual voice-banking software, but popular AAC apps that work well with banked voices. Both are products of Tobii Dynavox, a major provider of assistive communication technology (including many devices found in our BridgingApps labs). TD Talk is free; TD Snap requires a $9.99/month subscription.
Working with the Pros
There are other text-to-speech options that, like the TD apps, offer customizable speech but require users to provide their own banked voices. These options include Proloquo4Text ($119.99 to purchase) and Grid for iPad ($299.99 to purchase, or $10.99/month on subscription).
Voices for such apps are often created by professional voice-banking services. My-Own-Voice, first launched in 2015, can create a computer-generated voice from 50 sentences or 10 minutes of recordings: it “speaks” over two dozen languages and dialects. An account (which includes basic customer support, a choice of software packages, and a free 3-month trial period) costs $99/year—or $999 as a one-time “licensee” purchase.
Other voice-banking providers include ModelTalker and SpeakUnique. Ask your doctor, speech therapist, or insurance agent for help finding the best option.
In closing, remember: Talking is talking, whether it’s done verbally, via text-to-speech, entirely in writing, or in sign language. What matters is that everyone has every opportunity to contribute (and be valued for) their input.