The human hearing system is exquisitely sensitive to directional cues that let us instantly determine where a sound is coming from. This was critical for our survival in prehistory when the snap of a twig or a low growl might be the only harbinger of impending doom in the jaws of a hungry predator.

These days, we don't have to worry about becoming something's dinner, but the ability to discern the direction from which sounds reach us remains. This is one reason why 2-channel audio reproduction is ultimately unsatisfying—even with a system that exhibits excellent imaging, all sounds appear to originate in a relatively narrow window directly in front of us or, in the case of headphones, inside our heads.

Many have tried to simulate a 3-dimensional soundfield using 2-channel and even multichannel audio systems. But unless there are many speakers arrayed all around and above the listener—a prohibitively expensive approach—there's always something missing.

This problem intrigued Jerry Mahabub as a teenage prodigy working on brain-imaging technology while attending Rensselaer Polytechnic Institute in the late 1980s and early '90s. He recognized that previous attempts to simulate a true 3D sound space were based on head-related transfer functions (HRTFs), mathematical descriptions of how sound diffracts around human heads and enters the ears. However, Mahabub realized that this is only part of the solution—to achieve a more convincing simulation of a 3D sound space, a system must also model how the brain processes audio information, a field of study called psychoacoustics.

After 17 years of research and development, Mahabub's system, dubbed AstoundSound, is now being introduced. In 2003, he formed a company called GenAudio to commercialize various AstoundSound-based products for professional and consumer applications.

In the pro realm, AstoundSound can be applied to monaural, two-channel, and surround recordings during production, and the result can be heard on any playback system with no decoding required. Engineers can place up to 100 individual sound sources anywhere in 3D space and automate their movement over time, and the system is compatible with any audio format, including PCM, DSD, MP3, etc. (In the screen shot above, different sound sources are represented by colored dots placed within a virtual environment. The arrow indicates the direction in which the listener is facing, and the dot within the arrow is the listener's position.)

For pre-existing recordings, AstoundSound can be applied after the fact. A consumer-oriented Windows Media Player 11 plug-in called AstoundStereo Expander processes audio files using more than 90 DSP parameters to create a much broader, deeper soundstage. This software is also available for Macintosh computers running OS X 10.5.1 and can process audio from any application, such as iTunes, QuickTime, DVD Player, and others.

Unlike most current soundstage-expansion and surround-simulation technologies, which are primarily based on phase manipulation, AstoundSound does not introduce phase anomalies, as indicated by a phase monitor during a demo I recently attended. The system manipulates frequency response, interaural time delay, and interaural level differences among other physical properties, and very sophisticated EQ techniques are used to maintain tonal balance.

The demo I heard was, well, astounding. First, I listened to a clip on headphones and found myself turning around to see if someone had closed a door behind me while helicopters flew convincingly overhead. Next, I listened to Beyonce's "If I Were a Boy" on two speakers, switching between conventional and processed 2-channel versions. The difference was striking—with AstoundSound, the soundstage widened significantly and even increased in height, enhancing the clarity of each instrument while the vocal remained rooted in the center.

Among the first commercial releases to use AstoundSound are Hellboy II: The Golden Army on DVD and Blu-ray, Robin Thicke's Something Else CD, and Deprived, GenAudio's video game. You can check out some examples and get AstoundStereo Expander here. I'm very impressed with this technology, and I look forward to hearing it grow.