Unprecedented voice processor mimics human audio path

Feb 14, 2008 3:18 PM

Mobile start-up Audience has developed the industry’s first voice processor based on human hearing system to provide unparalleled voice quality in handsets operating in noisiest environments. Unveiled at the Barcelona’s Mobile World Congress, the developer claims that the processor enables instantaneous noise suppression of 25 dB (in < 500 ms) in both stationary and non-stationary noise.

Article Tools

According to Audience, its voice processor is the first custom IC that is modeled after the most efficient and accurate auditory system, the human hearing system. By thoroughly understanding the entire audio pathway - from the cochlea to the brainstem to the thalamus and cortex - Audience is the first company to deliver a commercial product based on the science of Auditory Scene Analysis (ASA), or the grouping and processing of complex mixtures of sound, said the developer. Because it handles signals the way people actually perceive specific sounds, it is able to identify and suppress noise sources in an extremely efficient and accurate manner, said the supplier.

"The Audience voice processor defines a new standard for robust, noise suppression, enabling unprecedented voice quality and more usability of rich voice and data applications," said Peter Santos, CEO of Audience. "People talking on mobile handsets will be able to hear and be heard everywhere from their cell phones, even in the noisiest places."

The voice processor receives a complex mixture of sound at overlapping frequencies, and organizes it into individual sources, in the same way people actually hear sounds. Regardless of whether the noise is local to the caller or remote over the mobile network, the processor uses several grouping cues to group all sound sources instantaneously, suppressing the noise and delivering the voice of interest clearly.

The voice processor A1010 is a compact single chip that integrates audio acceleration logic, a powerful custom digital signal processor, audio peripherals and control interfaces. Its low power and small package design is optimized for mobile phones and portable wireless peripherals, according to Audience. The A1010 is packaged in a 2.7 x 3.5mm WLCSP that meets RoHS requirements.

Designed and priced to fit into mid- and high-tier phones, the A1010 is sampling now. Sample pricing ranges from $5 - $7.