Opalchip, interesting points you bring up. I can't help but have a question come to mind: When you listen to a grand piano in a recital hall, the vast majority of the sound power that reaches your ears from the piano is reverberant energy, not direct energy. Would you consider this reverberant energy (which cannot possibly be time and phase coherent with the original signal) to also be "distortion"?
You see, the ear treats sounds arriving at different times in different ways. Different cues are extracted from reflections than are extracted from the first-arrival sounds. I believe the correct approach is to see them as two separate events, and to try to get them both right.
Twilo, I didn't hear the Hsu bipolar setup, but believe your description. However, note that the back-to-back speaker pair will probably have a deep notch in the response centered on the frequency where the path length difference from the two sets of drivers to the listening position is equal to 1/2 wavelength. Assuming the back-to-back speakers were each 8" wide and 6" deep the on-axis path length difference is 6 + 6 + (8/2) = 16 inches, so at approximately 420 Hz you'd have severe cancellation, along with partial cancellation at nearby frequencies. So back-to-back speakers may not be the ideal solution.
If I recall correctly, Mirage used a single bass driver on the front of a wide cabinet, and a rear-firing midrange and tweeter on the back of the cabinet crossed over higher than that 1/2 wavelength notch frequency. Definitive Technology patented a technique for using side-firing woofers along with forward and rearward facing mid/tweet arrays, once again to avoid that 1/2 wavelength cancellation notch.
You see, the ear treats sounds arriving at different times in different ways. Different cues are extracted from reflections than are extracted from the first-arrival sounds. I believe the correct approach is to see them as two separate events, and to try to get them both right.
Twilo, I didn't hear the Hsu bipolar setup, but believe your description. However, note that the back-to-back speaker pair will probably have a deep notch in the response centered on the frequency where the path length difference from the two sets of drivers to the listening position is equal to 1/2 wavelength. Assuming the back-to-back speakers were each 8" wide and 6" deep the on-axis path length difference is 6 + 6 + (8/2) = 16 inches, so at approximately 420 Hz you'd have severe cancellation, along with partial cancellation at nearby frequencies. So back-to-back speakers may not be the ideal solution.
If I recall correctly, Mirage used a single bass driver on the front of a wide cabinet, and a rear-firing midrange and tweeter on the back of the cabinet crossed over higher than that 1/2 wavelength notch frequency. Definitive Technology patented a technique for using side-firing woofers along with forward and rearward facing mid/tweet arrays, once again to avoid that 1/2 wavelength cancellation notch.