Early last week, I discussed my experiences attempting to record my wife reading some of her poetry and, in the process, pointed out the value proposition of a digital-output microphone technology such as the MEMS-based approach Akustica is now touting. I met Akustica for lunch yesterday, and their demos (all of which compared analog and digital mics built into a notebook PC's display bezel) uncovered additional justification for their transducer approach.

First, using Adobe Audition (in both its spectral- and time-based display modes) they showed me the additional RF noise generated by a laptop PC's display when it was running in a power-managed battery state versus 'full-on' via AC. They didn't know if this noise was being generated by the display drivers, the backlight, or both, but it noticeably coupled to the analog signal-carrying wire running between an ECM (electret condenser microphone) and the system board. The Akustica-based system's digital signal trace, in contrast, produced pristine results in both AC and DC scenarios.

Their next demo showed me the noise impressed onto an ECM configuration (and, in contrast, not impressed onto a digital mic configuration) in the presence of active Wi-Fi traffic. This was initially surprising to me, since 802.11b/g and 802.11a operate, respectively, within the 2.4 GHz and 5 GHz FCC Part 15 unlicensed frequency bands, neither of which directly falls within the 20-20,000 Hz range that's perceived by humans' auditory systems. However, harmonics and carrier signals do, apparently, interact with mics in audible ways; at least that's what Audition and my ears both suggested during Akustica's presentation.

Since Bluetooth's spectral 'fingerprint' is different than Wi-Fi, they (and I) don't yet know if it'll similarly degrade an analog mic configuration. But it's important to note that Wi-Fi and Bluetooth antennas both commonly locate in the display (along the side or top edges). Therefore, those antennas' likelihood of interfering with a similarly-located analog mic and its cabling is high.

Akustica is one of those rare companies who emerges in the midst of a timely 'perfect storm' situation. PC-based audio communication is hot and getting hotter; look at the profusion of soft VoIP configurations, or at Microsoft's demos of Windows Live Messenger during Monday's WinHEC keynotes, for examples. As my hands-on testing last Monday suggested, a motherboard-proximal mic isn't a palatable approach; users' voices are obscured by CPU and system fan whine, various HDD noises, key-pressing clicks, obscuring fingers and hands, and other distractions. And, if Akustica's demos are to be believed, display-mounted ECM mics are prone to RF noise-created degradation.

Clearly, it's possible to surmount these ECM challenges given enough time, money and effort; cell phones (which nowadays also comprehend GSM data and voice communications, Bluetooth, Wi-Fi, and power-managed displays) make this point. Akustica is candid about the fact that its MEMS digital mic price is roughly 2x the ECM alternative, at equivalent volumes. And, whereas Akustica admits that there are roughly 40 ECM suppliers in the world, at the moment you'll be going sole-sourced (and reduced-source on the corresponding digital-mic-input audio codec) if you pursue an Akustica-based design. So where will you spend your noise-suppressing money; on additional ECM-based R&D, or on the more straightforward Akustica-based approach?