Originally Posted by Nuance
Thank you for the information. I have studied some of Linkwitz's methods, and he seems to know his stuff. I don't, however, know how I feel about omni-directional speakers.
Having drivers on the front of a speaker doesn't mean it's not omni-directional at some frequencies. An 11' long 100Hz wave wraps around a domestically acceptable 8" wide speaker like it wasn't there radiating uniformly in all directions, and a 1KHz wave at over a foot has substantial energy to the sides and rear. Your vocal fundamentals are being sprayed all over the place from a conventional speaker.
The some frequencies part is where you get into trouble. While your brain does some separation between the direct sound and reverberant field, you need the frequency spectrum from the two to be sufficiently similar (trending slightly downward with increasing frequency in the reverberant field) for a natural sound.
Ideally reflections would be delayed at least 6ms or attenuated perhaps 10dB; although it's more important for natural sound to get the uniform spectrum than to be only missing pieces of the spectrum in your early reflections. Imaging still shifts when you have a strong side-wall reflection at some frequencies from a conventional speaker, so the increased directivity at high frequencies doesn't even help you in that department.
Not coincidentally, Vandersteens have baffles that are smaller at high frequencies for more uniform but wider dispersion.
You can (and should) still treat the early reflections from an omnipolar speaker, preferably with diffusion that will cut the reflection magnitude and preserve the power response.
Wave guides and acoustically small dipoles (While they have output in both directions, planar speakers are a poorer approximation of a true dipole's cosine alpha radiation pattern) get you the uniform output with reduced sidewall reflections thus making the treatment less important.
Off my first Orion setup in a 13x19x8' room, speakers on a short wall 4' off the front and 8' apart, listening position 11' off the front wall, 30 degree toe-in I calculated first reflections at
front wall 8ms delay, -6dB attenuation from distance, 19 degrees
229 degrees off-axis total
dipole difference: -3.7dB
total attenuation: -9.7dB
ceiling 4ms delay, -3.5dB attenuation from distance, 45 degrees
45 degrees off axis total
dipole difference: -3dB
total attenuation: -6.5dB
side wall 3.4ms delay, -3dB attenuation from distance, 37 degrees
67 degrees off axis total
dipole difference: -8.2dB
total attenuation: -11.2dB
While that front wall reflection may not be there at some frequencies with conventional speakers it is 8ms out and nearly 10dB down (insignificant) while a conventional speaker would have a -3dB side wall reflection just 3.4ms out which the dipole reduces to -11.2dB. That's a good trade-off.
Isn't the point to treat the reflections of the room so the reflection waves don't reach your ears before the direct sound waves from the speakers?
Since a straight line is shorted than any two connected lines sharing its endpoints the direct sound will reach you first. Imaging will be better if you limit the reflections via absorption, diffusion, or loud speaker directivity. Increased but uniform directivity will get you more clarity at larger distances as from a wave guide or acoustically small dipole.
I have listened to Def Tech's and Mirage, and I didn't like the sound.
Other than the characteristic of tonal balance not changing much when you move around the room there isn't any similarity with the Deftechs (I've owned BP8s in the same room). The Deftech problem is in the implementation (they're not accurate compared to other speakers at their price tag and the imaging is vague, perhaps due to unit-to-unit variations in the low cost drivers, perhaps for other reasons) and not the radiation pattern (which may still narrow too much in the midrange due to the high XO required by the 3/4" tweeters used). I haven't heard Mirage bipoles.
Of course, I have more faith in Linkwitz's design and engineering skills than I do in those who work for the later mentioned companies.
You might look for an audition at http://orion.quicksytes.com
We LInkwitz builders+owners are a bit fanatical and often willing to demonstrate.
I am in total agreement about minimizing reflections and cabinet resonance. This is why I was eventually drawn to the Vandersteen line of speakers (due to their design and limited baffle).
While the Vandersteens aren't bad for a box, a damped transmission line still does a better job but doesn't get used much (B&W's shell-shaped nautilus would be one commercial example).