Suppose that what the data actually shows is that your definition of spatial quality is flawed. Instead of being defined by the stereo field, where channel differences create a masking illusion of space, what if spatial qualities are really dictated by a speaker’s ability to create a timbrally accurate field of sound, by which I mean all the sound from the speaker, including reflections that arrive variously delayed, are timbrally consistent with how the brain identifies “space”?
After all, when I listen to a cellist, I hear the reflections of sound from all around the room in my two ears. The cellist is monophonic, but I still hear the cellist in space.
Thus, perhaps the data are showing us that wide and smooth directivity creates a sound field that gives the impression of space?
Your reduction of that concept to “interact more” is missing the point. The point is that speakers with wide and smooth directivity create timbral accuracy and spatial qualities, while speakers that are noticeably different in timbre off-axis may still create the illusion of stage width when you listen to two of them, but don’t create a spatial effect beyond that. And they also make instruments not sound like themselves when listening away from the sweet spot.
It took me a while to grasp this point. Stereo presentation is a masking effect—an overpowering distraction—the same way loud music masks distortion.
(For a person who claims to be a questioning amateur, you sure are displaying some fixated impressions as the basis for accusing experts of Missing The Point, however you temper them by use of passive voice.)
Rick “respectfully submitted” Denney
(Edit: Dear Apple: timbrally, timbre, and timbral are words, dammit!)
I am not sure that I fully understand your questions but I'll give it a go.
The study used three different speakers: one monopole with wide-ish directivity, one monopole with narrowing directivity and one dipole with narrow constant directivity.
None of the speakers was particularly flat on-axis nor apparently had very smooth directivity.
In other words there were several and not just a single variable at play:
• on-axis FR was different,
• off-axis FR was different,
• directivity characteristics were different,
• diffraction-generated artifacts were different,
• low frequency extension was different,
• the floor bounce cancelation effect was different,
• room interaction was different.
All these variables will contribute to sound quality and thus affect the perception of spatial quality, so one cannot say which one or ones are responsible for the perceived better spatial quality.
To make matters more complicated when listened to in stereo all three speakers rated similarly in terms of spatial quality.
We could remove the effects of room interaction and floor bounce cancelation from the equation by listening in anechoic conditions, of frequency response by EQ'ing flat on-axis and listening on-axis, of low end extension by high-passing all speakers. We would end up with directivity as the only variable, tainted somewhat by diffraction-generated artifacts and potential box resonances.
Because in order to determine the effects of directivity no other variable can interfere with the listening assessment perhaps the BeoLab 50 or 90 would be good candidates for the test because, if I'm not mistaken, they allow the user to change their directivity characteristics (at least in part of the spectrum); one single variable at play.
In my experience it is possible to perceive a sense of space with mono recordings over a pair of speakers. I have never tried listening to a mono recording with a single speaker (my Tivoli One isn't a fair comparison) but I would expect less "spaciousness", with the sound more concentrated on the location of the source/speaker. And without reflections (anechoic) I suspect that different directivity width will not affect spatial quality because it is not interacting with the room.
I am not questioning the importance of smooth directivity in reverberant, untreated rooms. Directivity is on the one hand a matter of taste and on the other a way to control how your speakers interact with the room. From my understanding and experience, and even taking into account that it is ultimately a matter of taste, wide directivity is, like omni, not suitable for narrow rooms (unless you like it).
You probably agree that the most significant difference between mono and stereo is the latter's ability to produce phantom source localisation and multi-dimensional spatial presentation (soundstage). These stereo effects result from using two speakers (and idealy a 2-mic 2-channel recording) but there seems to be some conflict between obtaining image "sharpness" and increasing "spaciousness". D'Antonio and others seem to agree that early reflections are damaging to both image "sharpness" and the "reconstruction" of the original venue acoustics (when there was one) and that is my experience too. And even though ultimately it's a matter of preference, if (you find that) early reflections are damaging then wide directivity does not make sense unless the room is very wide.
In regard to your cellist analogy there is a very concrete difference in goals between a musical instrument and a sound reproduction device or system, the former creates sound and the later reproduces recorded sound. Unless the cello was close mic'ed or recorded outdoors the mic will have picked up ambience cues. If your room's acoustic footprint is too strong (a reflective room, or wide-directivity speakers which interact more with the listening environment) it will overlay itself to or mask the recorded ambiance and mess it up. Whilst the result of this interaction may sound pleasing with pop/rock studio mixes it is in my view and/or for my taste damaging with classical music.