Can you please elaborate on that? I thought that high frequencies have a higher absorption at increasing distances so at these where sound power/off-axis start to play a bigger role such a behaviour would be counterproductive?
Directivity / size (area) of HF radiator compensates absorption to long distances. Larger domes, rings, planars, HF line arrays etc. "cut the air" to longer distance and reduce reflections acoustic keeping resolution up. Too steep tilt in power is also bad due to limited listening area so we should know or find suitable compromise especially if listening distance is not fixed.
Maybe you already have searched the message I referred, but here is a bit longer replay:
Frequency response of human ear depends on direction. Relative level of high frequencies varies quite much when sound source moves from 0 to 80 deg. 0 deg is the center speaker or listening test in mono in front of single speaker. Angle of default/fixed listening triangle in stereo is usually ca. 50 deg. Some speakers and listeners tolerate more, and some less because physical tolerance and preference is personal.
Close to ideal constant directivity i.e. horizontal flat directivity index is not the most usable in stereo if listening distance is not constant but speakers are at fixed locations and aimed to the longest common listening point. Some day you may want to listen at the far point to relax and avoid "too high" acoustic resolution causing brain-tiring, or long distance suits for some genres or recordings. Some other day you move the chair much much closer to enjoy full acoustic resolution and envelopment. In that case HF would be clearly emphasized with constant directivity speakers, without new aiming or EQ. Too high relative subjective level at HF increases the risk of sound stage collapsing. Especially with conventional speakers with sharp edges having diffraction problems. Also timing error (long excess group delay at LF...MF) increases the risk for sound stage collapsing with multi-ways. Larger heavily rounded and actual half space concept can be quite close to zero diffraction system. It can be fun experience to listen close to perfect balance and sound stage, head almost between the speakers. For example constant directivity at LF...MF + conventional 1" dome without wave guide and minimal diffraction concept can do that trick.
We have also seen many kind of room correction target curves. Harman's and
dirac etc. Most of those have some level drop at HF. 11 listeners (below) should not override anyone's personal preference but room curves commonly dropping at HF indicate that too constant directivity up to 20k may not be the most adaptable option especially without tilting on-axis down. Tilt in factory setting could be difficult to explain to potential customers, and questionable/subjective judgements and conclusions in reviews don't help.
