Subwoofer

A normal home listening room will impose huge peaks and dips on the output of a subwoofer or pair of speakers.

As we increase the number of intelligently distributed bass sources in a room, we significantly reduce the magnitude of those peaks and dips. The only way multiple subs could fail to make the net response smoother would be if they all had identical in-room response, and the distribution ensures that nothing even close to that will happen.

So while adding a single subwoofer will offer the benefit of improved bass extension, adding multiple smaller subs and spreading them around intelligently will additionally result in smoother bass throughout the room. By way of anecdotal evidence, it seems like most Maggie and Quad owners who add a single sub go back to using their speakers without subs within two or three months. But it seems like most Maggie and Quad owners who add a pair of subs keep them in their system. Those few who have tried more than two subs never go back, as far as I know.

I have multiple customers who report +/- 3 dB in-room from 20 Hz up through the bass region without any EQ from using four small subs. And smooth bass is "fast" bass, subjectively speaking, because where there are peaks in the bass region, there is ringing. Smooth the peaks and the ringing has been addressed.

While the output of a single subwoofer can be equalized to be very smooth at a single location, or almost as smooth within a small area, such equalization will inevitably make the response worse elsewhere in the room. Because a distributed multisub system results in much less spatial variation within the room, any need for EQ is probably correcting a global (room-wide) problem rather than a local one. So EQ is arguably rendered even more effective by a distributed multisub system, assuming it’s even needed.

A distributed multisub system results in a greater number of smaller peaks and dips in the in-room response. The subjective improvement is often greater than the raw numbers would lead you to believe, and here’s why: The ear/brain system tends to "average out" peaks and dips that are within about 1/4 octave of one another, so when we have a lot of smaller peaks and dips bunched up together, they start to behave as a continuum (as far as the ear/brain system is concerned). This is what happens higher up in frequency (those hashy-looking room-reflection peaks and dips are there, but they’re so close together that we don’t hear them discretely), and this is also why large rooms have subjectively better bass than small rooms. A distributed multisub system can make a small room behave like a much larger room at low frequencies.

Here is what UCLA mathematics professor, concert violinist, and respected audio reviewer Robert E. Greene had to say about a distributed multisub system: "Makes even the most magnificent of one-point subwoofers into dinosaurs, something grandly impressive in their time, but their time is over."

So imo there is arguably significant room for improvement beyond the addition of a single good sub.

Duke

dealer/manufacturer

"Duke, does your opinion change if I am only interested in good bass at a single location in the room?"

If you're going to be equalizing for that single location, I can only think of one theoretical disadvantage of using just one sub:  If you have a significant dip at that location, you could conceivably overtax your sub trying to fill it in with EQ.  That being said, dips are subjectively much more benign than peaks. 

I have been told by someone who tried it both ways that the distributed multisub system sounded more natural in the sweet spot than a single equalized ubersub, but that could have been due to other factors.  And my sample size (in this case, one) is too small to draw any conclusions from. 

Duke

Just for the record, I am skeptical of the notion that cone material makes an audibly significant difference in the low bass region, as long as the cone is stiff enough to behave as a piston. I don’t doubt that a person could hear a difference between different subs and that difference could seem to correlate with cone material, but cone material would not be the actual cause of the difference.

As for "speed", the ear has incredibly poor time-domain resolution in the bass region. We cannot even detect the presence of bass energy from less than one wavelength, and we must hear several wavelengths before we can hear pitch. So whatever it is we are perceiving as "speed", it is actually something other than the leading edge of a low frequency signal. We will come to exactly what it is in a minute.

What the ear is extremely good at, in the bass region, is hearing differences in SPL. Look at a set of equal-loudness curves:

http://hyperphysics.phy-astr.gsu.edu/hbase/sound/imgsou/eqlou.gif

See how the curves are close together below 100 Hz? What that means is, a small change in SPL is perceived as a much larger change in loudness. So a 3 dB change at 30 Hz can SOUND LIKE as much of a difference as a 10 dB change at 2 kHz!

(Incidentally the ear’s exaggerated sensitivity to small changes in SPL at low frequencies is one of the reasons why it takes a long time to get the gain dialed in properly on a subwoofer system.)

Low frequency speakers in rooms have a minimum-phase characteristic, which means that the time-domain response tracks the frequency response. So when we have a peak in the frequency response, we also have ringing at that frequency (conversely if we fix the peak, we fix the ringing). At low frequencies it is this MODAL RINGING ASSOCIATED WITH FREQUENCY RESPONSE PEAKS that sounds "slow"! This happens on the TRAILING EDGE of a low-frequency waveform, where our perception is well-equipped to detect what we interpret as "speed" (because we are good at hearing loudness), NOT on the LEADING EDGE!!

Recall how it takes us several cycles to hear the pitch of bass energy. By the time we hear bass, the energy has reflected off of room surfaces numerous times. So PERCEPTUALLY, we cannot hear the leading edge of a bass waveform, nor can we separate the subwoofer from the room. They are a SYSTEM, as far as human hearing perception goes. So imo it makes sense to design subwoofers that work with the room instead of against it. If the in-room frequency response is smooth, then we don’t have some frequencies decaying significantly more slowly than others, which contributes to the perception of "speed".

Now it is common knowledge that sealed boxes sound "faster" than vented boxes. This is explained by their in-room frequency response trends. A vented box tends to be flat down to a much lower frequency, and then roll off rapidly below the tuning frequency. A sealed box starts rolling off higher up, but more gently. Once we factor in "room gain" from boundary reinforcement, the vented box typically has a net rising in-room response as we go down in frequency. Couple this with room-induced peaks plus the ear’s significantly increased sensitivity to small SPL changes in the bass region, and vented boxes do indeed tend to sound boomy. The more gentle rolloff of a low-Q sealed box has a much better synergy with room gain from boundary reinforcement in most cases.

"Typical" room gain from boundary reinforcement is ballpark +3 dB per octave below 100 Hz. A low-Q sealed box has a rolloff rate that approximates -6 dB per octave, so the synergy there is good. Imo it still leaves room for improvement, but that would be another topic for another day.

Controlled blind listening tests have shown that group delay on the order of what we might expect from a well-designed high-quality vented box is inaudible on program material, and just barely detectable on test tones. But the difference in frequency response is highly audible. In the real world the time and frequency domains are linked in the bass region, but it is our perception of loudness in the frequency domain that dominates our perception of "speed". When a kickdrum goes from full-on to very quiet very quickly because there is no modal ringing (because the in-room response is smooth), that sounds very FAST and TIGHT. (The frequencies that give the kickdrum impact are largely above the subwoofer range, so we use the gain, frequency, and phase controls to optimize that blend.)

I’m not saying there would be no audible difference between a sealed box and a vented box equalized to have the same in-room frequency response, but I am saying that any difference would have nothing to do with the leading edge of the waveform.

Anyway my point is, the differences in the perceived "speed" of subwoofers are due to factors other than cone material or the leading edge of a bass waveform. Perceived "speed" in the bass region is more closely related to the in-room frequency response than to anything else (which is in turn linked to modal ringing or lack thereof), so that is the thing we should focus on first. I’m not saying that’s the only thing that matters, but I am saying it matters far more than the subwoofer cone material or the leading edge of a low-frequency waveform, neither of which human hearing is well-equipped to hear.

Duke

bo1972, there is nothing wrong with making experimental observations as you do, but in my opinion one critical element seems to be lacking: A strong background in acoustics and psychoacoustics.

The human mind loves to have explanations for what it observes in the world around it. If we do not have enough correct information in our mental database, we will manufacture explanations using what limited information we do have. So we may conclude that this subwoofer sounds faster than that one because of cone material or woofer diameter, and once this has become one of our core beliefs, our mind selectively interprets our experiences to preserve and reinforce that belief. We confuse correlation with causation and come up with erroneous "laws".  And we humans do not like to have our beliefs challenged because we identify with them. We love to "be right" and we hate to "be wrong" (see, even the wording equates our self with our belief, which is a mistake). Our highest allegiance should always be to the truth, not to a favorite theory.

In my opinion it is the responsibility of those of us who claim to be professionals in this field to acquire a professional level of expertise. That doesn’t necessarily mean we need a college degree in the subject, but it does mean that we should at least read and study for our niche as much as a serious college student would.

I suggest starting out with Dr. Floyd Toole’s "Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms", which has chapters devoted to the stuff this thread is about. Don’t stop there, but at least start there, and then look in the bibliography for further reading in areas of interest. I also recommend Geddes, Everest, and Blauert.

The advantage of going to so much effort is this: Now when we make an observation, our minds will be able to draw from valid scientific theories and explanations stored in our mental database, instead of having to manufacture its own explanation from very limited information.

Imo, ime, ymmv, etc.

Duke