One big subwoofer or two weaker subwoofers?

Size matters. Of course so does quality, but 20 Hz sound at any reasonable volume is an uphill battle for an 8-10 inch driver.

It is possible to make a small cone vibrate at subwoofer frequencies, but what happens when the sound goes out into the room? It spreads out and becomes less loud. And with a small driver, the SPL falls off rapidly. If you have a small room, and listen close up, a small subwoofer is OK. In the extreme case, tiny earphone diaphrams can reproduce LF very well, but require immediate proximity to the ear.

Take a look at what the cone of a small subwoofer is doing...long excursion is necessary. For any dynamic loudspeaker, (good ones and bad ones alike) performance deteriorates as cone excursion increases. A large cone will need much less excursion to move the same volume of air.

In keeping with rules of full disclosure, I state that my SW system consists of three 15" plus three 12" drivers mounted in large sealed enclosures embedded (for cosmetic reasons) in the wall of my room. These systems "drive" the entire room, so that the sound is very uniform throughout, and there is an effortless quality that I miss with even excellent smaller subwoofers. Ideally, I would have the entire wall consist of drivers, and it would not matter if it was 20 15" cones, or 120 8" cones. My subwoofer design philosopy is consistent with the nature of my main speakers...three planar MG1.6.

The SW systems are in the wall because I like the characteristics of a sealed (acoustic suspension) system. There are many drivers suitable for sealed box alignment, but few designed for infinite baffle operation. Also, it is an outside wall.

It is three separate SW systems, each having 15" and 12" drivers. Each driver has it's own amp. This is necessary because the 4 ohm driver impedance discourages parallel operation. Each 15" driver is powered by half of a CarverPro ZR1600 digital amp, 600 watts. The 12" drivers are powered by 250 watt SW "Plate amps" and their signals are equalized for phase and frequency response. I intend to use the 12" drivers to tweek the overall system performance, but in practice little or no equalization is needed. The enclosures are about 7.5 cubic feet each. The front baffle of each enclosure (which resemble coffins standing on end) is covered with acoustic foam to absorb some of the back radiation from the Magneplanars. The whole SW system is concealed behind fabric wall covering. In spite of its massive size, it is nearly invisible. (WAF you know). At present I am using 80 Hz 24 dB/octive for crossover. This makes life easy for the Maggies.

Because I designed and constructed the whole thing it didn't cost too much. I reckon it was about $3200, including 2550 watts of power amps, and sheetrock, plaster and paint to remodel the living room. Besides, the project was interesting. Better than dropping $3200 in a one hour visit to an audio shop.

"Faster". What does this mean? Here is what I think.

Frequency response is part of the story. A subwoofer that is flat to 400 Hz will handle 80 Hz better than one that is flat to 200 Hz.

However, a loudspeaker that reproduces the LF waveform perfectly can still be "slow" in the sense that the amplitude of the sound builds up slowly, and perhaps persists after the electrical signal is cut off. This characteristic does depend partly on the size of the driver, with large ones tending to be slower. However, apart from the enclosure, the full story involves not only the size of the driver (actually the moving mass of the cone) but also the strength of the motor that drives the cone: the voice coil, magnet, and amplifier. My 15inch drivers, for example, have 4 inch voice coils, massive magnets, and are driven by powerful amps. They are not slow. The characteristic of a driver that would correlate with slowness is a ratio of motor force/volt to the moving mass.

Also, I must mention my view that most very low frequency sound sources are, by nature, slow. A large organ pipe, for example, does not sound forth at full volume when activated, and it continues to emit sound after it is turned off, not to mention the hall resonance that can be strong for several seconds. I think that what many people take to be "slowness" in subwoofers is really the nature of the sounds they reproduce.

Finally, aren't subwoofers fun! How often do you see similar debates about tweeters?

Sean...Those small drivers with big voice coils wouldn't be Dynaudio by any chance. I have a set of Dynaudio MTM boxes with 5inch "woofers" that have 3inch voice coils. They bark furiously, but no bite. Very nice close up in a small room.

Size matters.

Sean..What no 30 inch EV30W? Back in the old days one would install one of these monsters in the ceiling, vented to the attic. Hartley 18 inchers were second best.

If you are looking for 15" subwoofers, check out the JBL SUB 1500 drivers that Parts Express is selling at a good price. I just got three to replace the JBL LE15 that I have been using (which are good, but not really designed for the SW application). I am impressed.

Sean...Those JBL 1500 seem to work well in my sealed enclosures, but then they are 7.5 cubic feet. How would I know that these drivers are designed for vented use, other than the FS = 24 which is higher than some but still well below the typical 15" Prosound driver?

The only thing I can cite to match your EV 64" thing is the vibration machine that we use to test our missile GS. This machine really is a huge loudspeaker, and has a "voice coil" four feet (yes feet) in diameter, and shakes two hundred pound objects at awesome levels. Interestingly, its frequency response goes a lot higher than most subwoofers. We use it up to 1000 Hz or so (but not with flat response). I don't know what the power amp is...it's old enough to be some kind of tubes. Maybe amplidyne. I will find out. The drive is a closed-loop servo using accelerometers mounted on the test fixture (a la Velodyne).

Thanks Sean....I have copied the info so I can read it at leisure. This appears to be the best definition of TS parameters that I have ever seen. What I am looking for is some kind of single dimensionless number (like Bernoulli's number that relates to fluid dynamics) that would describe where a driver falls on the optimization spectrum between vented and sealed. Maybe I will invent such a parameter.

Sean...I guess that QTS is the parameter I had in mind, although as the writup suggests, it does not reliably describe vented/sealed suitability . The JBL 1500 QTS is 0.32, not far from the 0.4 suggested as the boundry between vented and sealed. I am sure that enclosure size, shape, and stuffing has a lot to do with it.

I could easily put a vent in my enclosures, but mice would probably take up residence. (Deaf mice).