Audience (News - Alert) today announced its latest and greatest voice processing chip at Mobile World Congress, the earSmart eS325. The "Advanced Voice processor" is expected to ship in commercially available smartphones by the end of April 2013, delivering HD voice and superwideband support, three microphone processing, Mobile Audio Zoom, Bandwidth Expansion technology, optimized Automatic Speech Recognition (ASR) and Enhanced Stereo Recording.
As mobile carriers move to HD voice, on-platform (i.e., on the phone) voice and sound processing become much more important to avoid calls that are "HD noise" – picking up a wider range of sound, including background noises that distract or mask conversation.
Audience leads the field with its earSmart voice processors, using algorithms based on the process of human hearing to suppress background noise and enhance voice quality.
The company has numerous customer wins, with earSmart processors used in high-end smartphones and mobile devices from companies such as Acer, Dell (News - Alert), Google, Huawei, HTC, LG, Lenovo and Samsung.
The special sauce in the earSmart eS325 starts with using three microphones to process sound, instead of one or two. Using three microphones gives improvements of around 30 to 55 percent signal-to-noise ratio (SNRi) improvements in noisy environments, with up to an 80-percent improvement in "optimal conditions" over two mic, previous-generation products, says Audience.
Lower power consumption also plays a role, with the eS325 a 40nm chip process to deliver 41 percent lower power consumption for narrowband calls, and 52 percent lower power consumption on HD voice calls.
Fraunhofer (News - Alert), promoting its "Full HD Voice" AAC codec, will be happy to know that the Audience earSmart eS325 processes sound at 24 kHz, supporting superwideband codecs above and beyond today's 8 kHz-esque AMR-WB HD voice codec.
But performance is nothing without applications. The eS315 provides the "world's first" de-reverb on a mobile device, removing voice echo from such spaces as a hallway or meeting room. ASR Assist improves the reliability of speech-enabled applications by 40 percent or more in noise conditions, delivering more useful speech-to-text, voice service and command – making hands-free use of Google and Nuance (News - Alert) a lot better in real-world normal noisy conditions.
The two new features I love the best are Audio Zoom and Enhanced Stereo Recording. Audio Zoom enables a device to record a discussion in an interview mode while suppressing noise and capturing both voices with clear accuracy – just the thing for an interview on the tradeshow floor or other area where there's a lot of background buzz.
Enhanced Stereo Recording minimizes background noise while capturing video using two mic 48 KHz noise suppression to capture high definition audio. I suspect there are a couple of camcorder manufacturers that might roll two or three mics and an Audience earSmart processor into its next-generation products.
Smartphones with the eS325 will be available "within 60 days" of its announcement, according to Audience, with Audio Zoom and other features available depending on the manufacturer and carrier turning on the functions. I'm already starting to shop for a new phone, and Audio Zoom just made the short list of must-have features I'm going to look for.
Edited by Braden Becker