A Broad Discussion of Speakers with Major Audio Luminaries

[Floyd Toole]

Well, in fact, while you can't 'image' below 90Hz, you can get intense changes in envelopment. This absolutely requires more than one sub.
That's way, way below any transition frequency, and you can't claim "bass quality" until you make that listening room sound as "wide" as the church you started in.

Sorry, but we have to disagree, and there are discussions thereabouts this board detailing others' experiences with it.

Reducing seat to seat variations is only the start.

I don't think we disagree, more like we just have different priorities. In decades of double-blind subjective evaluations it became very clear early on (1985/86) that bass extension, by itself, correlated with sound quality ratings. No real surprise, I suppose. Then, many years later, Sean Olive did his subjective/objective correlation study, and it revealed that bass extension and quality accounted for about 30% of the overall sound quality rating. These evaluations were of loudspeaker sound quality performance, and were therefore done in mono - no spatial component was in the source material.

Small rooms have resonances that modify bass sound quality in ways that are not subtle, and if one wishes to share the audio experience, as in home theater, the seat-to-seat variations experienced in the standing-wave patterns are a major problem. Equalization definitely helps because prominent standing waves behave as minimum-phase phenomena, but it cannot satisfy more than one listener with any certainty. Bass management became part of multichannel sound, which aided our ability to address the resonance/standing wave problem. The realization that multiple subwoofers could be used to manipulate standing waves in predictable ways was a powerful tool in attenuating both the seat-to-seat variations and the resonance problems because EQ then benefitted multiple listeners. This was progress on the sound quality front, but envelopment was not considered.

But, there were voices saying that "stereo bass" was lost, which is true. The next question is "how important is stereo bass?". Within Harman, David Griesinger was an advocate, arguing, correctly, that in concert halls the dimensions allow for "directionality" in long-wavelength, low-frequency sounds and that this could/would/might be a factor in perceived envelopment. The long wavelengths (20 ft/6 m at 50 Hz) compared to the spacing of the ears means that the effect is likely to be subtle compared to binaural effects at higher frequencies, but human hearing is good at detecting subtleties. We mounted a demonstration, set up by David, in which we listened to a variety of stereo vs mono subwoofer comparisons using a wide variety of music and digitally contrived signals that should have been good at revealing differences. This was done at several stereo-to-mono crossover frequencies, using different layouts, with the auditioning being done in a largish living room. It was definitely a serious effort.

Differences were heard, but they were quite subtle. Differences in sound quality were expected, because the excitation of room modes is quite different when two subs are operating in mono or stereo - i.e. receiving the same or different signals. These differences would depend on the stereo separation at bass frequencies in the program material (LPs don't qualify - all low bass is mono). But here we were making an effort to focus on differences in "space/ envelopment", to the extent that such a perceptual differentiation is possible. The differences we could report seemed to fade to insignificance at a crossover frequency of about 80 Hz, a figure supported by other investigations described in Section 8.4 of the 3rd edition, provocatively entitled " Stereo Bass: little ado about even less", with apologies to William Shakespeare, if indeed he wrote the words. The conclusion was that the necessary spatial information exists at frequencies above about 80 Hz, and therefore it is present in bass managed systems. It is a stereo "upper-bass" effect, not a stereo low-bass effect. There is also a subtle problem in A vs B comparisons - hearing a "difference" is not declaring superiority of one option - here we were content to hear a difference.

Is that a definitive statement? Probably not, but it strongly suggests that whatever potential there is for enhancing envelopment by capturing directional bass cues in large spaces and reproducing them in small rooms must be achieved while at the same time reducing seat-to-seat variations and room resonances, both of which are easily audible. Because listening rooms are not standardized any wavefront reconstruction exercise is clearly a custom listening room, multichannel audio system - expensive - solution. Present indications in the audio industry are that it is not likely to be commercially successful. But it is an interesting academic exercise.

Recordings, including classical recordings, are mixed and mastered in recording control and mastering (small) rooms, in an industry that seems proudly to ignore science and lacks even basic standardization. It seems to me that there is a "circle of confusion" problem to be added to the list of challenges facing listener satisfaction.

 

[middlemarch]

Mr. Toole, very much appreciate and enjoy your contributions to these discussions. Having read through my copy of your third edition several times already, I'm very much intrigued by your mentions of the upcoming 4th edition. If I may be so bold, could you give us an approximate availability for this version? I realize there is a portion of the effort out of your hands, but count me as being at the front of the line when it hits the bookstores.

 

[MattHooper]

But here we were making an effort to focus on differences in "space/ envelopment", to the extent that such a perceptual differentiation is possible. The differences we could report seemed to fade to insignificance at a crossover frequency of about 80 Hz, a figure supported by other investigations described in Section 8.4 of the 3rd edition, provocatively entitled " Stereo Bass: little ado about even less", with apologies to William Shakespeare, if indeed he wrote the words. The conclusion was that the necessary spatial information exists at frequencies above about 80 Hz, and therefore it is present in bass managed systems. It is a stereo "upper-bass" effect, not a stereo low-bass effect. There is also a subtle problem in A vs B comparisons - hearing a "difference" is not declaring superiority of one option - here we were content to hear a difference.

Question:

Many audiophiles report that adding subwoofers often aids the perception of a larger acoustic space being reproduced, if that is in the recording. So for instance a recording in a large hall in which the real hall acoustics play part in the recording. The idea is that the size information of the acoustic can be also found in surprisingly low frequencies.
So even if you’re listening to just a cello being played in a big hall recording, if you had subwoofers the sensation of the hall acoustic becomes more convincingly large or enveloping.

Given what you’ve posted above, if this is the case then it wouldn’t necessarily be due to stereo reproduction of those lower frequencies, but perhaps simply restoring them even in mono subwoofer frequencies?

So I guess it’s a general question about what has been studied in terms of low frequency contributions to the apparent scale of an acoustic space.

(I mentioned before I do sound design for movies and TV, so I am actually dealing with and creating “ room tones” of various sizes all the time. If I want to reduce the apparent size of a room tone, EQing out bass rumble is one of the first ways to do it… and essentially the more bass I take out the smaller the room sound generally becomes).

 

[MaxwellsEq]

If I want to reduce the apparent size of a room tone, EQing out bass rumble is one of the first ways to do it… and essentially the more bass I take out the smaller the room sound generally becomes).

If you deliberately increase rumble (e.g. from another large space) into the recording of a small space, does the room apparently get bigger?

 

[j_j]

If you deliberately increase rumble (e.g. from another large space) into the recording of a small space, does the room apparently get bigger?

Depends on the spatial distribution of the "bigger room rumble". If it's very stationary as you move around, it just sounds muddy.

 

[MaxwellsEq]

Depends on the spatial distribution of the "bigger room rumble". If it's very stationary as you move around, it just sounds muddy.

OK that makes sense. So shrinking a perceived room size by subtracting deep rumble is effective, increasing the size by addition of deep rumble is likely ineffective.

 

[Floyd Toole]

Mr. Toole, very much appreciate and enjoy your contributions to these discussions. Having read through my copy of your third edition several times already, I'm very much intrigued by your mentions of the upcoming 4th edition. If I may be so bold, could you give us an approximate availability for this version? I realize there is a portion of the effort out of your hands, but count me as being at the front of the line when it hits the bookstores.

Latest estimate is publication around September. Thank you for being a reader of facts - more of it is needed in the audio scene.

May 2, 2025 UPDATE: I just learned that the publication date is now estimated to be November, 2025. I'm disappointed.

 

[Floyd Toole]

Question:

Many audiophiles report that adding subwoofers often aids the perception of a larger acoustic space being reproduced, if that is in the recording. So for instance a recording in a large hall in which the real hall acoustics play part in the recording. The idea is that the size information of the acoustic can be also found in surprisingly low frequencies.
So even if you’re listening to just a cello being played in a big hall recording, if you had subwoofers the sensation of the hall acoustic becomes more convincingly large or enveloping.

Given what you’ve posted above, if this is the case then it wouldn’t necessarily be due to stereo reproduction of those lower frequencies, but perhaps simply restoring them even in mono subwoofer frequencies?

So I guess it’s a general question about what has been studied in terms of low frequency contributions to the apparent scale of an acoustic space.

(I mentioned before I do sound design for movies and TV, so I am actually dealing with and creating “ room tones” of various sizes all the time. If I want to reduce the apparent size of a room tone, EQing out bass rumble is one of the first ways to do it… and essentially the more bass I take out the smaller the room sound generally becomes).

I share that perception, and I have always attributed it, without scientific proofs, to the time domain information - i.e. reverberation time. Large rooms tend to have much longer reverberation times than small rooms. Reverberation time per se is not a "spatial" perception. Highly regarded concert halls have reverberation times (RT) in the range 2 to 2.5 s at 124 Hz, rising below that frequency. Typical domestic listening rooms tend to be around 0.4 s at 125 Hz, so hearing prolonged bass by itself is a significant cue as to room size. It is no surprise that adding a subwoofer to reproduce those long RT low frequencies creates a "large room" impression. It is not amplitude itself, or space, but amplitude vs. time.

Figure 10.1 in the 3rd edition illustrates several examples. I doubt that it is more complicated than that.

 

[Lion♡]

I share that perception, and I have always attributed it, without scientific proofs, to the time domain information - i.e. reverberation time. Large rooms tend to have much longer reverberation times than small rooms. Reverberation time per se is not a "spatial" perception. Highly regarded concert halls have reverberation times (RT) in the range 2 to 2.5 s at 124 Hz, rising below that frequency. Typical domestic listening rooms tend to be around 0.4 s at 125 Hz, so hearing prolonged bass by itself is a significant cue as to room size. It is no surprise that adding a subwoofer to reproduce those long RT low frequencies creates a "large room" impression. It is not amplitude itself, or space, but amplitude vs. time.

Figure 10.1 in the 3rd edition illustrates several examples. I doubt that it is more complicated than that.

Thank you for the great explanation.

I don't remember the exact source of the paper, but among the various papers I compiled and organized, there was this information as well.

When the T60 was around 0.1s, participants generally perceived the room size to be about 3.0m and described it as small and narrow.
When the T60 was around 0.4s, participants generally perceived the room size to be about 4.5m and described it as the size of a realistic living room.
When the T60 was around 0.8s, participants generally perceived the room size to be about 6m and described it as a large space or resembling a concert hall.

Also, when I manipulated and listened to binaural room impulses that I recorded myself, which were larger than 30m in size, I noticed that adjusting the decay time changed the subjective perception. Moreover, when listening to the full spectrum, it felt like the space I was in, but when I cut the low frequencies, it felt somewhat off. It didn't feel as big as I expected. This was my subjective perception, beyond just the tone balance differences.
So, conversely, when I applied a high-cut filter, the vast and spacious feeling disappeared completely, leaving only fluctuating low frequencies. From my small personal test experience, it seemed that not only the decay time but also the response that harmoniously matched the room size made that feeling more pronounced.

 

[Floyd Toole]

Thank you for the great explanation.
View attachment 444786
I don't remember the exact source of the paper, but among the various papers I compiled and organized, there was this information as well.

When the T60 was around 0.1s, participants generally perceived the room size to be about 3.0m and described it as small and narrow.
When the T60 was around 0.4s, participants generally perceived the room size to be about 4.5m and described it as the size of a realistic living room.
When the T60 was around 0.8s, participants generally perceived the room size to be about 6m and described it as a large space or resembling a concert hall.

Also, when I manipulated and listened to binaural room impulses that I recorded myself, which were larger than 30m in size, I noticed that adjusting the decay time changed the subjective perception. Moreover, when listening to the full spectrum, it felt like the space I was in, but when I cut the low frequencies, it felt somewhat off. It didn't feel as big as I expected. This was my subjective perception, beyond just the tone balance differences.
So, conversely, when I applied a high-cut filter, the vast and spacious feeling disappeared completely, leaving only fluctuating low frequencies. From my small personal test experience, it seemed that not only the decay time but also the response that harmoniously matched the room size made that feeling more pronounced.

All of thiis fits a well established pattern. Most discussions of reverberation time (RT) in concert halls focus on mid frequencies, but bass obviously matters as well. The music being performed decides what the optimum value is, with vocal oriented operas needing lower RT than instrumental music - closer to 1s as opposed to around 2s at mid frequencies. Conserving bass energy is important - bass viols cost money - but not at the expense of articulation. A rock performance in a concert hall is dreadfully "boomy", lacking the necessary punch and articulation -an acoustical mismatch.

Around the time I studied in London, in the early 1960s, the Royal Festival Hall underwent renovation because too much bass absorption had been built into the structure. Later they added artificial reverberation using many microphones, amps and loudspeakers - I believe it was the first time it had been done. Anticipating that audiences and concert critics would rebel at the thought, its was done in secret, and revealed publicly only after reviewers commented on the "mysterious" improvement in sound. Now, it is often done and one contributor to that technology was Lexicon's LARES reverberation enhancement system. Build an acoustically "dead" hall, good for speech, and add reverberation as appropriate for other kinds of performances. These days it makes economic sense.

 

[Lion♡]

Conserving bass energy is important - bass viols cost money - but not at the expense of articulation. A rock performance in a concert hall is dreadfully "boomy", lacking the necessary punch and articulation -an acoustical mismatch.

Thank you for the great explanation. I also subjectively agree with that part and strongly support it.

The interaction between reflections in the 50-150ms range and the subsequent RT was fascinating. When adjusting it, it felt like swinging between muddiness, clarity, and space. I was once again amazed by the appropriate settings and theory for large spaces.
When I first became interested in sound and started as a hobby, I used to think in simple terms, such as more reverb or fewer reflections, in a very flat and straightforward manner. However, after studying the documents from David that I attached, along with your works and insights from j_j, I've found it incredibly enjoyable to study various keywords related to notes, spatial variations, auditory masking, perception, and cognition on my days off, and then test and listen based on that. (To be honest, I'm not really into music. I seem to prefer testing the theories of contributors like you and feeling the differences through that process rather than just listening to songs.)

Around the time I studied in London, in the early 1960s, the Royal Festival Hall underwent renovation because too much bass absorption had been built into the structure. Later they added artificial reverberation using many microphones, amps and loudspeakers - I believe it was the first time it had been done. Anticipating that audiences and concert critics would rebel at the thought, its was done in secret, and revealed publicly only after reviewers commented on the "mysterious" improvement in sound. Now, it is often done and one contributor to that technology was Lexicon's LARES reverberation enhancement system. Build an acoustically "dead" hall, good for speech, and add reverberation as appropriate for other kinds of performances. These days it makes economic sense.

I haven't heard the LARES reverb enhancement system from Lexicon, but I think that approach is highly innovative and valid.
I will definitely experience such a magnificent concert hall at least once.

 

[Floyd Toole]

Thank you for the great explanation. I also subjectively agree with that part and strongly support it.

View attachment 444792
View attachment 444793
The interaction between reflections in the 50-150ms range and the subsequent RT was fascinating. When adjusting it, it felt like swinging between muddiness, clarity, and space. I was once again amazed by the appropriate settings and theory for large spaces.
When I first became interested in sound and started as a hobby, I used to think in simple terms, such as more reverb or fewer reflections, in a very flat and straightforward manner. However, after studying the documents from David that I attached, along with your works and insights from j_j, I've found it incredibly enjoyable to study various keywords related to notes, spatial variations, auditory masking, perception, and cognition on my days off, and then test and listen based on that. (To be honest, I'm not really into music. I seem to prefer testing the theories of contributors like you and feeling the differences through that process rather than just listening to songs.)



I haven't heard the LARES reverb enhancement system from Lexicon, but I think that approach is highly innovative and valid.
I will definitely experience such a magnificent concert hall at least once.

Lexicon (David Griesinger) created what was called LARES "light" for home audio systems. We had a demo at Harman for a while. Four mics and small loudspeakers connected to a digital controller created an amazingly credible "large room" impression in a home theater. Talking, clapping hands, etc. all indicated that one was in a large space. Playing stereo became much more a "concert hall" experience, all the more convincing because the hall illusion did not disappear when "pause" was pressed. I consider this to be one of the most compelling stereo "upmix" methods I have ever experienced. It was not commercially viable at the time, but could very well be now that digital horsepower is inexpensive and systems like Atmos already incorporate some of the necessary hardware. The acoustical properties of the "hall" were adjustable.

 

[MattHooper]

If you deliberately increase rumble (e.g. from another large space) into the recording of a small space, does the room apparently get bigger?

FWIW:

It’s a lot more artistic than scientific.

In general, if I want the recording of a large space to sound like a smaller space, taking away the low frequencies tends to make it sound smaller.

But if I have a smaller space recording that I want to sound bigger, than I don’t normally EQ in more bass (often it’s not there to bring up in the first place) but rather I will process it through other reverbs in order to expand the size of space. So it’s easier to EQ a large space into a smaller space, but it’s not so easy to EQ a recording of a small space into a larger space - I tend to have to augment that small space recording either with a wider sounding reverb, or I also can lay in some separate low rumble tones that have been designed (usually by me manipulating stereo width effects, and reverbs ) to evoke a large space.

The thing about this job is that In the end I’m not dealing so much with “ the thing that was recorded in front of the microphone” as much as I’m dealing simply with “ what the recording itself happens to sound like. What that recording evokes.”

So keeping it to room tones, I may have scene taking place in a large, empty factory.
And I may search out some recordings in the library that were recorded in a large empty factory. But that doesn’t mean they actually end up sounding like a large empty factory.
Depending on the choice of microphone, pick up pattern, the way it was recorded, and actual large interior factory recording can end up sounding like you are in a smaller basement with some furnace rumble - it actually sounds more claustrophobic than it sounds large.

This is the issue that Dr Toole (and others here, including me) point out: the way microphones don’t pick up sound the way our brains interpret sound. So you’re always dealing with the microphone’s interpretation, which doesn’t always convey how it sounded in the real space.

Whereas I may have a recording of an average size room, and yet due to the stereo spread of the recording there is an open airy space quality, and I might use that as a basis for the large space I want to create. I might augment it with some reverb and/or run some larger rumble beneath it. So in the end, it doesn’t matter so much what was in front of the microphone during the recording, but rather what the recording ended up actually sounding like and what impression I can create using that recording.

 

[Heinrich]

Many audiophiles report that adding subwoofers often aids the perception of a larger acoustic space being reproduced, ...

Impulsive sounds are the core of human language. We, as the humans we all are, are quite trained to analyse everything about it, auditorily. (We might not be quite as successful with the sabre-toothed cat anymore, as some stereo addicts think; they got extinct.)

Switch from steady state to sound packets flying around, interfering with themselves.

Many times I proposed extra speakers for reverberation simulation (popular in the late 80s). And avoiding the early cut-off in bass of especially small ported speakers.

All ignored in favor of Dirac style in-room equalisation down to the very last decibel (Harman curve!) at some preferred listening position alone. If one cannot provide a spinorama granting excellence, gone ... ;-)

And recordings get more and more demanding in regards to stereo-exactness, they are mostly unlistenable to me. It's becoming a self-referential attempt to have a good stereo ;-(

 

[j_j]

It is no surprise that adding a subwoofer to reproduce those long RT low frequencies creates a "large room" impression. It is not amplitude itself, or space, but amplitude vs. time.

Figure 10.1 in the 3rd edition illustrates several examples. I doubt that it is more complicated than that.

Unfortunately, that doesn't work. A single sub does not create the spatial sensation of width (or depth for that matter when the listener moves their head).

You need the ability to radiate sound, that's for sure, but there is more to it. Merely adding ONE subwoofer is only a start. This is actually true well above traditional subwoofer frequencies, up to the 100Hz or so range, as well. I think you underestimate how sensation works at low frequencies.

 

[Lion♡]

Lexicon (David Griesinger) created what was called LARES "light" for home audio systems. We had a demo at Harman for a while. Four mics and small loudspeakers connected to a digital controller created an amazingly credible "large room" impression in a home theater. Talking, clapping hands, etc. all indicated that one was in a large space. Playing stereo became much more a "concert hall" experience, all the more convincing because the hall illusion did not disappear when "pause" was pressed. I consider this to be one of the most compelling stereo "upmix" methods I have ever experienced. It was not commercially viable at the time, but could very well be now that digital horsepower is inexpensive and systems like Atmos already incorporate some of the necessary hardware. The acoustical properties of the "hall" were adjustable.

Such aspects seem similar to what some users in the Korean community I participate in refer to as "반사음파이=Reflection-Fi, Reverblation-Fi," where multiple small (or identical) speakers are used to create additional spatial impressions (tailored to personal preferences and goals), and I think this is a very good approach.
It was part of an attempt to overcome the lack of spatiality that arises with the limited room space conditions and only two speakers. (Of course, the playback shouldn't overpower the recording, but during personal tests, I tried playing speakers directed toward a high ceiling, around 15m to 20m, and artificially emphasized the strength of the ceiling reflections by adjusting the speaker's gain and timing. That, too, was quite an interesting experience. Taking a brief break from the recording/reproduction loop, subjectively, the direct sound started from the front, and the background image began from the ceiling, creating a sound that felt like stardust falling from the sky.)

 

[Lion♡]

Unfortunately, that doesn't work. A single sub does not create the spatial sensation of width (or depth for that matter when the listener moves their head).

You need the ability to radiate sound, that's for sure, but there is more to it. Merely adding ONE subwoofer is only a start. This is actually true well above traditional subwoofer frequencies, up to the 100Hz or so range, as well. I think you underestimate how sensation works at low frequencies.

The insights from j_j and Floyd Toole on that topic will be valuable lessons for many users. What do you think is the minimum threshold for this? If I recall correctly, according to Thomas Lund's research, it was around 40Hz. I'm curious about what you consider to be the average meaningful range.

 

[Heinrich]

Unfortunately, that doesn't work. A single sub does not create the spatial sensation of width (or depth for that matter when the listener moves their head).

You need the ability to radiate sound, that's for sure, but there is more to it. Merely adding ONE subwoofer is only a start. This is actually true well above traditional subwoofer frequencies, up to the 100Hz or so range, as well. I think you underestimate how sensation works at low frequencies.

As you say, it won't work. In order to replicate any spatial cues in the bass in small rooms the small room's own acoustics has to be neutralized. To sum up many subs would, in connection to the resonances, set up a diffuse field. That won't be exactly helpfull in that it loses all phase / time correlations.

I wonder if the asymmetric nature of impulses could be a factor. Of course the DC part of a drum kick cannot be replicated. But the impulse is defined by a more or less continuous spectrum. Which f3, e/g 20Hz, is worth to be realized?

Starting from 40Hz, then 30Hz, finally 20Hz in-room changed things for me personally.

 

[j_j]

As you say, it won't work. In order to replicate any spatial cues in the bass in small rooms the small room's own acoustics has to be neutralized. To sum up many subs would, in connection to the resonances, set up a diffuse field. That won't be exactly helpfull in that it loses all phase / time correlations.

Done right, I'd say that is exactly helpful for the diffuse field. Of course, for direct sound, you don't want to do that.

And that's possible. Just because you have, say, 7 LF radiators does not mean you have to use all 7 for every auditory cue. Using all 7 to clean up problems in the room (done under control) remains possible, of course. But "flatness" is not the major criterion once you've cleaned out any major peaks and cancellations.

But, again I'll say the best way to get rid of standing waves is absorption. Adding more energy to a room that already stores too much is not generally a great way to start your day. So you can't get completely away from the room you've got without some way to get energy out of it as well as into it.

 

[Heinrich]

Done right, I'd say that is exactly helpful for the diffuse field. Of course, for direct sound, you don't want to do that.

And that's possible. Just because you have, say, 7 LF radiators does not mean you have to use all 7 for every auditory cue. Using all 7 to clean up problems in the room (done under control) remains possible, of course. But "flatness" is not the major criterion once you've cleaned out any major peaks and cancellations.

But, again I'll say the best way to get rid of standing waves is absorption. Adding more energy to a room that already stores too much is not generally a great way to start your day. So you can't get completely away from the room you've got without some way to get energy out of it as well as into it.

Don‘t agree on loading the room with even more. That‘s not how it works.

But to first find a constellation that would neutralize the room, and then, using the same facilities, inject spatial cues on top is possible; the two would just multiply on a „linear“ basis.

That brings me back to ask, if the self-interference of those sound packets flying around is a low frequency phenomenon?