Voice is the new user interface, and Audience is rapidly capitalizing on its technology and experience to get into new areas. This week the company released its first voice processing software product for x86 PCs and the new e800e Series processors with far-field and direction features that should have enterprise phone manufacturers taking notice.
Audience S1.0 is the software product targeted at higher-end Intel -based PCs, tablets and all-in-one devices. It delivers what the company describes as "improved VoIP capability" along with intelligent key click removal, improved speech recognition performance, and 360 degree processing using only two microphones.
Super wideband and "Full" HD voice are all the rage among mobile devices. S1.0 can provide super wideband noise suppression for clear VoIP calls in nearly any environment, in addition to acoustic echo cancellation and de-reverberation to reduce double-talk and echo effects in larger spaces. Intelligent key click removal works both during quiet pauses and speech, so you won't have to hear the click-click-click if you are talking to someone – say, like myself – during an interview or otherwise doing work at a keyboard.
The 360 degree voice feature is intriguing both for first generation applications by OEMs and potential in future IP handset and conference phones. Users can participate in conference calls sitting around a device and all can be heard clearly, regardless of their respective positions in relationship to the microphone. OEMs can be much more flexible in putting microphones and openings on devices, not having to drill holes through the front into metal or glass.
Audience's eS800 series is the company's fifth generation architecture in silicon and is loaded with interesting features. The line supports far field capabilities for distances up to five meters. There's a "barge-in" capability allowing voice recognition to work during media playback without interruption, such as if you were listening to music or a video and you wanted to tell the computer what to do.
Voice Compass may be the hottest feature in the new silicon. It provides the ability to recognition the position of the speaker relative to the device to improve speakerphone usage, enabling a device to pick out multiple speakers sitting around it.
The eS854 chip in the family includes high quality playback and capture up to 192 kHz, the sort of range audiophiles love, and a slick feature for advanced accessory detection and multi-button capability for universal headset support. For example, it can tell the difference between a Beats and a Bose headset and support multiple control buttons if they are present.
While Audience is focused on consumer devices, both products are likely to gain attraction in the enterprise OEM market. Dolby currently holds the high-ground in 3D sound processing with its proprietary system, but Audience has the silicon for IP phone manufacturers to build similar devices using existing HD voice codecs, with third-parties providing backend processing of higher-quality audio. The IP handset market could certainly use a shot in the arm with more processing power at the device level.
