Engineers at the University of Washington have developed a way for some deaf people to enjoy music. The findings could help people with cochlear implants, a bionic inner ear that allows deaf or hearing-impaired people to hear speech, albeit in kind of a robot voice.
Cochlear implants can be a lifesaver for people without hearing, but when it comes to music, this very practical device can’t carry a tune to save its life.
The implants simply aren’t sensitive to pitch and what’s called timbre—the qualities of a sound that make, say, a guitar sound different from a harp.
This, for example, is a simulation of how a familiar melody might sound to someone with a conventional cochlear implant:
And here’s the original:
But UW engineering professor Les Atlas has come up with a different way for the implant to process sound, turning those monotones into this:
It may not sound pretty to a hearing person’s ears, but it conveys basic changes in pitch. And that, says Atlas, is much better than nothing.
“If you're able to provide some of that, even a degraded versions, it’s something they don’t have now,” Atlas says.
Atlas and his colleagues, including UW's Dr. Jay Rubinstein, achieved that effect by changing the algorithm used to encode the sound. Implants capture sound in a series of rapid micropulses. Atlas’s algorithm lets those pulse rates vary in a way that conveys pitch.
The advance may do more than bring melody to implant uses. Atlas says it could help then hear speech better in a noisy environment: the so-called “cocktail party problem.”
“You walk into the room, several people are talking. They’re talking at different pitches,” says Atlas.
But if you can’t hear pitch, all those voices squish together and become just noise.
