In response to my
post number 295,
@hex168 and
@Thomas savage expressed interest in my offer to “describe what I believe to be a stereo setup where something desirable happens to the spatial quality which cannot be evaluated by testing a single speaker.” Perhaps
@richard12511,
@Coach_Kaarlo, and
@tuga ought to be notified as well, since they “liked” post number 295.
And I SHOULD HAVE made it clear that I was referring to LISTENING tests, not the actual MEASUREMENTS of a single loudspeaker. My bad, and I hereby excuse any of the five of you who thought I was talking about MEASUREMENTS from having to read the rest of this post. (I suggest you take advantage of that loophole; this is a dreadfully long post.)
Please consider all of this to be my opinion so that I don't have to remember to insert an annoying number of disclaimers.
Here's the TL;DR version: If you take two speakers with narrow and well-controlled patterns and toe them in such that their axes criss-cross in front of the main listening position, their room interaction will result in better spatial qualities than could have been predicted from mono on-axis listening to a single speaker.
And now the long version:
First, let me re-state my opinion regarding mono vs stereo listening tests so that nobody reads through all of this under the assumption that my claim is farther-reaching than it actually is:
I claim that stereo listening is more revealing of spatial quality than single-speaker mono listening is.
And the intention of this post is to present an example of an unorthodox but valid stereo setup whose spatial qualities cannot be adequately evaluated by listening to a single speaker.
For the purpose of this post, let's define “good spatial quality” as, the recording venue's acoustic space perceptually dominating over the playback room's acoustic signature. In other words, a “you are there” illusion is the goal (which of course is also somewhat recording-dependent, so we can only focus on the speakers and their interaction with the listening room).
In the playback room there is in effect a “competition” between the venue cues on the recording (whether said cues be real or engineered) and the "small room signature" cues of the playback room itself. The ear/brain system make its best guess about the acoustic space you are in by drawing from these two packages of acoustic cues. Ideally we'd like to effectively present the venue cues which are on the recording while minimizing undesirable "small room signature" cues, thus enabling a “you are there” illusion. This is of course easier said than done.
At the risk of oversimplifying, the ear looks at three things to judge the size of an acoustic space: The earliest reflections; the reverberation tails; and the temporal "center of gravity" of the reflections. So we have the early reflections, reverberation tails, and temporal "center of gravity" of the venue cues on the recording; and we have the early reflections, reverberation tails, and temporal "center of gravity" native to the playback room. We want to effectively present the first package of cues while disrupting and/or minimizing the second.
Achieving our goal is complicated by the fact that it is the in-room reflections which are the CARRIERS of the venue cues (at least for two-channel). Yes there are venue cues in the direct sound, but that is actually the WORST possible direction for reflections to come from. Reflections coming from many different directions will do a much better job of presenting the venue's spatial cues.
ALL of the in-room reflections are carriers of the desirable venue cues, but the FIRST in-room reflections are also the STRONGEST source of undesirable "small room signature" cues. So the WEAKER the first reflections are (relative to the later ones), the LESS small-room signature we will have superimposed atop the venue cues.
On the other hand the in-room reverberation tails are very effective carriers of the recording venue's reverberation tails, and therefore can play a significant role in creating the illusion of immersion in a much larger acoustic space than the playback room. So we want to preserve these reverberation tails, and of course we want them to be spectrally correct. (If the in-room reverberation tails are spectrally incorrect then the ear/brain system has a harder time identifying them as reflections, which is detrimental for reasons I can explain if anyone is interested.)
So to recap where we are thus far: We want WEAK early reflections because that weakens the playback room's “small room signature”; and we want plenty of spectrally-correct later reflections because they effectively and benignly deliver the recording's venue cues.
There is one more thing that would help: If we can somehow time-delay the first lateral reflections, this would push back in time the temporal "center of gravity" of the in-room reflections. Doing so disrupts the undesirable "small room signature", because a long time-delay for the first lateral reflections tells the ear/brain system that you're in a much bigger room that you actually are. In effect this undermines the plausibility of the “small room signature” package of reflections, making it easier for the ear/brain system to accept the recording's venue cues as the more plausible package of reflections.
(If I didn't put him to sleep,
@Kal Rubinson is probably shouting at his computer screen that a good multichannel system already does all of this. And he is right! But let's see how much we can do with just stereo.)
Okay with that long-winded background behind us, let's look at what happens when we use the unorthodox approach to stereo that I briefly described earlier, with narrow-pattern speakers and criss-crossing patterns.
The Gedlee Summa is designed to be used like this. The Summa combines a 15” woofer with a similar-sized round, 90-degree, constant-directivity waveguide. The crossover is around 1 kHz, where the woofer's pattern has narrowed to match that of the waveguide. Here is a polar map of the Summa's frequency response:
The designed-for listening axis of the Summa 20 degrees off-axis. You can see this depicted by the red line in the overhead polar response view to the right of the polar map, above. And in the polar map you can see a horizontal black line that is not right smack down the center of the polar map, but is 20 degrees above the center. That is the intended listening axis.
And below is a photo showing how this type of speaker is intended to be set up. The photo is taken from off to one side and you can see that the speaker axes criss-cross in front of the central sweet spot (these speakers aren't Summas):
The combination of narrow, well-behaved radiation pattern AND strong toe-in has two major implications for room interaction. First, the early same-side-wall reflection is very weak over the imaging-critical upper portion of the spectrum because the speaker's pattern is aimed about 45 degrees AWAY from the same-side wall. Second, the first STRONG lateral reflection is actually the long, across-the-room bounce off the OPPOSITE side wall! Earl Geddes on the subject:
“A reflected signal that arrives at the opposite ear from the direct sound is less perceptible as coloration and image shift than if both signals arrive at the same ear. This is because of head shadowing above about 500 Hz and the fact that our ears can process signals between them. When the two signals arrive at the same ear, the signals are physically merged in space even before they enter the ear and no amount of auditory processing can separate them. When these signals arrive at different ears, the auditory processing system can diminish the adverse effects of these early reflections through cognitive processing between the ears."
In a conversation with me Earl used the term "decorrelation" to describe this situation where the first significant lateral reflection arrives at the
opposite ear. Decorrelation increases the sense of being in a large acoustic space and therefore is a disruptor of "small room signature" cues. Decorrelation is used to derive surround channel information from a recording which does not have surround channels.
Earl continues:
“From an acoustics reproduction standpoint then, the loudspeaker system design must help to provide as much delay as possible in the early reflections and allow for speaker placement and orientation such that the earliest reflections occur at opposite ears rather than the same ear. This needs to be done above about 500 Hz. Below 500 Hz other factors, such as room characteristics and our hearing mechanism, may dictate an entirely different approach.
“In order to perform this “trick” for optimizing the early reflections in a small room, two specific source characteristics are required. First, the source directivity must be less than about 90° and second, the listener is not actually on the sources axis! In other words the direct sound, the first arrival sound, is not the axial sound. To achieve a flat response at the listener in this configuration, the loudspeaker must have a flat frequency response off axis. This is virtually never the case with most loudspeakers. Most loudspeakers with a smooth flat axial response will usually not work very well in this configuration.”
So here are the first two advantages of this cross-firing narrow-pattern speaker setup, which would not be revealed by listening to a single speaker on-axis:
1. Reduced image shift (and reduced coloration) due to reduced early same-side-wall interaction, and due to the first strong lateral reflections arriving at the opposite ear; and
2. Reduced “small room signature” due to weakened early reflections, decorrelation, and pushing the temporal “center of gravity” of the reflections back in time.
Another spatial quality attribute of this setup is that the soundstage holds up much better than normal for off-centerline listeners. The ear/brain system localizes sound by two mechanisms: Arrival time and intensity. With a conventional stereo configuration (speakers toed-in little if any), the soundstage tends to collapse towards the nearest speaker for off-centerline listeners because the nearest speaker “wins” both arrival time and intensity. With narrow-pattern speakers toed-in aggressively (like the approximately 45 degrees you see in the photo above), the near speaker “wins” arrival time but the FAR speaker “win” intensity! This is because, as the photo shows, the off-centerline listener is well off-axis of the near speaker but right smack on-axis (or nearly so) of the far speaker. The two competing localization mechanisms tend towards balancing one another out, and you still get a decent spread of the instruments (including the center vocalist remaining near the center with most recordings) even from locations as far off-centerline as where the photo was taken from. The KEY to this working well is that the response of the near speaker must fall off RAPIDLY and SMOOTHLY as we move off-axis, which is not the case with most loudspeakers.
So we can add one more spatial advantage of narrow-pattern speakers in a cross-firing stereo configuration:
3. Soundstaging holds up much better than normal for off-centerline listeners.
As previously mentioned, the correct listening axis of the Summa 20 degrees off-axis. In the following image the black curve across the top is the frequency response at 20 degrees off axis, so (assuming proper set-up) this is the frequency response of the first-arrival sound:
The red line is the power response curve, and the white line is the directivity index. Imo both are very good, with the spectral balance of the off-axis energy matching the spectral balance of the first-arrival sound exceptionally well (and imo this is desirable – I can explain my thinking on the subject if anyone is interested).
One consequence of the first-arrival sound being the 20 degrees off-axis sound AND the extreme toe-in is that the perceived tonality is relatively consistent across a wide listening area. And as we move off-axis to either side, the net spectral balance of the first-arrival sound holds up better than is normally the case because the reduction in energy in the top half of the spectrum from the near speaker is being offset by the increase in energy in the top half of the spectrum from the far speaker. This is of course more of a sound quality advantage than a spatial quality advantage, but without it the spatial quality advantage for off-centerline listeners would imo be of academic interest only. So I'm adding it to my list:
4. Reduced spectral discrepancy between the first-arrival and reverberant sound, for off-centerline as well as on-centerline listeners.
And as is hopefully self-evident, advantages 3 and 4 would not be revealed by single-speaker on-axis listening.
So in conclusion I think there is at least one valid stereo configuration with spatial attributes that cannot be evaluated by single-speaker on-axis listening.