Loudspeaker system design poses several "assumption traps" that causes us to waste time and expense on parameters that have little or no impact in the real world while limiting our choices on the ones that actually do
"You're right - could be a magic driver where the laws of physics don't apply. Or it could be the huge peaks in the upper frequency range that you're not taming helping the off axis. In any case, you can look at off axis measurements of people who use this driver in a full range system and see that its about 12db down at 6000 from its level at 1000"
rob r. sez:
Angle of relative incident (up/down, left side/right side) plays an important role in determining what the useful range of a driver is.
To assume that all drivers must be omnidirectional and that audition occurs 180 deg hor. and 180 deg vert. limits the designer's ability to optimize first arrival flatness (the perceptual leader of your hearing), unity transfer (all the terms in the numerator of the s plane transform of the crossover equalling the terms in the denominator) and sufficiently low IM (6% is considered quite an accomplishment at 90dB@1M).
In fact, a more appropriate assumption would be that the useful angle of audition would be more like +/-20 deg horizontal and +/- 40 deg vertical.
With these parameters established you could then intelligently make decisions on appropriate crossover points to realize the best possible fit between on- axis flatness, minimum IM and unity.
This tends to fly in the face of many first timers as the assumption is made that the higher the xover slopes, the better (wrong), asymetry is better (wrong) or controlling out of band driver behavior is more important than minimizing IM (yes, wrong).
If lobing error is a worry, then one should welcome the MTM or WMTMW array as symmetry about the array center cancels lobing error along the lengthwise axis over a reasonable audition angle.
You may use first order crossovers (unity transfer) with MTM (lobing error cancelling). You may even make the xover points higher thus limiting IM.
"Intellectual games? Ah, here's the scientist trying to intimidate the layman... sorry, I'm just stating basic speaker building ideas. If we go by what the creator of the mtm configurations suggests, the CENTER to CENTER spacing of adjacent drivers should be kept close to one wavelength of the crossover frequency."
rob r. sez:
This incorrectly assumes omnipolar behavior.
"I don't really care that you think your driver is 2.75" in diameter, every other w4-1337 out there is 4 15/16". So you're going to tell me your mtm with a crossover of 5600 does not have any vertical lobing? Guess you're powering that with FM (not frequency modulation)"
rob r. sez:
The directivity limitations of spaced 4" drivers (mechanical mount limited with even a large tweeter in between) would have to demonstrate an amazingly omnidirectional polar pattern to end up with an end-fire lobe of any consequence (even at the crossover).
I believe that jj has made some excellent reality based decisions on his design
rob r. (who has designed more bad speakers than anyone