I found this article misleading. I came in thinking I'd be hearing about the latest and greatest formats for storage and playback of music and sound and find a pretty shameless plug for Intel's new marketing catchword for codec audio. What the author glosses over is that the "standard" Intel proposes is in fact proprietary and only found on Intel chipsets and functionally indifferent from the existing, universally-accepted AC'97 standard.
It's nice that the chipset supports 192kHz, 32b sampling but so do many inexpensive sound cards these days. However, the actual sound quality you hear is determined by the digital-analog and analog-digital converters and the amplifiers attached to these circuits as well as the amplifiers in your speakers and the speakers themselves. Historically speaking, motherboard manufacturers have opted for the cheesiest codecs and wimpiest amplifiers money can buy with few exceptions; I seem to remember a motherboard a couple years back that had
valve amplifiers on it, read: electron tubes. I see no reason for this trend of poor component specification to change. Also, never mind the fact that if you use digital speakers or are connected to your home theater receiver via S/PDIF your computer's sound hardware has absolutely nothing to do with the sound you hear anyway (DSP, decoding, and amplifier tasks are done by hardware in the speakers/receiver).
The author then segues into a shameless Dolby plug. The first two "sound levels" are unacceptably poor for anyone that listens to something other than ground-shaking bass like rap, hip hop, or heavy metal or uses their computer's sound for the various activity noises only. Less than 95dB signal to noise ratio is easily audible noise at normal volume levels. Note that Analog Devices SoundMax codec audio found in many laptops and motherboards is rated for 95dB and in my opinion produces only acceptable levels of noise at reasonable volume levels.
The upmixing the author refers to in the form of Dolby Pro Logic has been available for some time now, but lets face it, every time a signal goes through one of these filters you get distortion. That's the entire point of upmixing in the first place. A fairly smart digital signal processor (DSP) sends phase-shifted copies of the original signal to the different satellite channels (fronts, backs, and centers) and low-pass filtered combinations to the low-frequency channels (subwoofers). These phase shifted signals add together in the air and the combined total excites your eardrums. What you're getting is something that the author found pleasing but is in actuality a highly distorted reproduction of the original signal.
If you don't understand terms like distortion, noise, SNR, THD, etc. I find this to be a good primer:
http://www.ethanwiner.com/audiophoolery.html
-drasnor
