Audience demonstrates industry-first voice processor based on human hearing system at CTIA Wireless 2008
By Will Strauss
At the Mobile World Congress, I had 28 scheduled interviews and skimmed by several times that many booths. Although I observed many improved products and even a few new wireless chipsets, the product that most impressed me was not wireless per se. Actually, it was an audio chip that I can actually call a "breakthrough" product.
The A1010 voice processor chip by Mountain View-based Audience is the first one that I've encountered that effectively demonstrated noise suppression from both stationary sources (such as repetitive noise sources like automobile highway sounds) and non-stationary noise sources (like rock music in the background or a gaggle of people talking in your airport waiting area). Based on the company's proprietary DSP engine and logic, the tiny A1010 chip provides an interface to all basebands. The basic speech processing algorithm is the secret sauce, based on intelligence of the human hearing system and can be ported to other chips.