Sub-woofer fast enough for Horns?

It is the shape of the in-room bass frequency response curve, rather than the "speed" of the woofer, that is primarily responsible for how "fast" the bass sounds.

Lumpy bass = "slow" bass, as over-emphasis somewhere in the bass region gives the subjective impression that we tend to describe as "slow" or "fat" or "boomy".

And,

Smooth bass = "fast" bass, it's actually not any faster, but it sounds that way when the fundamentals and their overtones are in proper proportion.

Group delay is of far less subjective significance than we'd tend to think, because the ear's time-domain resolution is quite poor in the bass region. Controlled listening tests that used digital signal processing to isolate group delay from frequency response have shown that group delay on the order of what we'd get from a vented speaker system is barely audible on test tones and statistically inaudible on program material. The reason sealed subs generally sound faster is the shape of their frequency response curve (which is generally more room-gain-friendly), rather than their actual "speed".

Anyway, imo the key to smooth, and therefore "fast", in-room bass is how the subwoofer(s) and room interact. What the room does to the sub's output dominates. Yes we can hear the difference between one sub and another in a given room, but still the room's generally detrimental signature is imposed on whatever the sub is doing.

The solution I advocate is, multiple small subs distributed around the room. Each will inevitably have a different room-induced peak-and-dip pattern, and the sum of these multiple dissimilar peak-and-dip patterns is much smoother than any single sub's output would be. And to address the issue of room gain (more precisely boundary reinforcement) boosting the low end a bit, I prefer subs whose native frequency response starts out as approximately the inverse of typical room gain.

Let me go off on a psychoacoustic tangent here. The ear/brain system tends to average out peaks and dips that are fairly close together, but if they are too far apart, the ear/brain system can't average them and they stick out like sore thumbs (especially the peaks - dips are harder to hear). The room-induced peak-and-dip pattern of a typical home listening room with a single sub has the peaks and dips spread way too far apart for the ear/brain system to average them. But with a distributed multisub system, not only are the peak-to-dip ratios significantly reduced, but we also have more peaks and dips bunched up closer together, so that the ear/brain system's averaging-out characteristic can work to our advantage.

Imo this approach has advantages over a single equalized sub, in that the bass response is much more uniform throughout the room. And if you do want to equalize a multisub system, you get better results because you don't have as much response variation from one location to another.

For deepest-loudest-bang-for-buckest bass, go with a single mighty ubersub. For smoothest (and therefore subjectively fastest) bass, I think the acoustics and psychoacoustics both favor a good distributed multisub system.

Duke
dealer/manufacturer (yes, of multisub systems... grains of salt all around)

Something that isn't obvious at first glance is that we literally cannot hear bass in our home listening rooms without hearing the room. The ear cannot detect the presence of bass energy without hearing at least one full cycle. And it cannot detect the pitch of a bass tone without hearing several cycles. Consider how long the wavelengths are in the bass region, and the size of our listening rooms, and you'll see that by the time you begin to hear a bass note, the room's signature is already all over it. So if high quality is the goal, we have to consider sub(s) + room = a system, because working on the sub(s) alone is ignoring the elephant in the room (which is the room itself).

A subwoofer may start out with a perfectly flat (and perfectly "fast") response, but by the time you hear it, the room has modified that flat response into a roller-coaster of major peaks and dips. Some subs will synergize better with a particular room acoustic situation than others, but no single sub can give you reasonably flat (+/- 3 dB) response anywhere in the room without equalization, and no equalized single sub can give you reasonably flat response across a broad listening area (because our room interaction peaks and dips change places as we move around the room, EQ that improves the response in one location is almost certainly making it worse somewhere else). So if you have a high quality speaker system, imo you're simply not going to be able to extend that quality down into the low bass region with a single sub, no matter how good it is. And having low bass that doesn't blend into the rest of the spectrum is often worse than having no low bass at all.

Nothing against bass traps, they can certainly help, but cannot correct the basic room acoustic situation of what the room is going to do to a single bass source. With multiple bass sources distributed around the room, they average out much smoother (and subjectively "faster") than any one alone. Intuition may tell you that the differing arrival times would cause smearing and blurring, but that intuition would be wrong, because the ear doesn't begin to have that kind of time-domain resolution at low frequencies. Remember our inability to even detect the presence of bass energy from less than one full cycle, and our need to hear multiple cycles to detect pitch? That's an indication of the ear's poor time-domain resolution at low frequencies.

I originally developed my multisub system (based on the ideas of Earl Geddes, which I'm using with his permission) to work well with planars. In the planar world, most who try a single sub go back to using no sub at all, because the discrepancy between what the sub is doing and what the panels are doing is just too great. On the other hand, most who try two subs end up keeping them, as they offer a worthwhile net improvement. Two subs are roughly twice as smooth in-room as one sub. Each time we double the number of subs, we essentially double the smoothness (or, halve the lumpiness), as long as our subs are spread around somewhat. This approach absolutely does not require special subs - you don't need to buy my system, you just need multiple decent subs. And they don't need to be large, high-output ubersubs because you're basically still getting ubersub air-moving capability by using multiple smaller subs.

Imo, ime, ymmv, etc.

Duke

The two primary researchers in the field of multiple subs are Todd Welti of Harmon International, and Earl Geddes of GedLee, LLC. Briefly, Welti investigated symmetrical configurations while Geddes advocates asymmetrical configurations.

Todd Welti on multiple subs, in layman's terms:

http://www.harman.com/EN-US/OurCompany/Innovation/Documents/White%20Papers/multsubs.pdf

Earl Geddes on multiple subs (you may have to cut and paste this to a new browser window):

https://docs.google.com/file/d/0BxgUOGOB5HbfR0JTRF9XZjkyUms/edit?usp=drive_web&pli=1

And a well-written blog article on the subject (author uses Geddes' approach):

http://mehlau.net/audio/multisub_geddes/

Enjoy!

Duke