Tight bass sub recommendations

My short answer is, two smaller subs instead of one bigger sub.

The following is my long answer.

Off and on over the course of a decade or so I tried building a sub that was "fast" enough to mate well with Maggies and Quads, on the theory that there might be a market for such. I built sealed boxes, low-tuned vented boxes, transmission lines (many different geometries), equalized dipoles, aperiodics, isobarics, and pretty much anything that seemed promising except for a full-sized horn. Some were better than others, but none passed the test.

The one day a really smart guy, Dr. Earl Geddes, taught me that the problem is room interaction, and regardless of how "tight" and "fast" a sub is, the room will impose large peaks and dips that will dominate its response. It is the peaks that are especially detrimental, in that they decay slower than the rest of the spectrum. His suggestion was to use four small subs asymmetrically distributed, such that each produces a different room-interaction peak-and-dip pattern, and the sum of the four dissimilar peak-and-dip patterns would be much smoother (and therefore much "faster") than any one alone.

This made sense to me. I was aware of an AES paper that showed a dipole has significantly smoother in-room bass than a monopole, and a dipole is two monopoles back-to-back with the polarity of one reversed, plus a path-length-induced time delay between them.

The general principle I learned from Earl is, the more intelligently-distributed bass sources within a room, the smoother the in-room bass. Two subs are potentially twice as smooth as one, and four subs are potentially twice as smooth as two. A dipolehas roughly twice as smooth in-room as a monopole in the bass region, so four intelligently-distributed monopole subs are theoretically comparable to two dipoles.

If you do a casual survey of relevant posts by Maggie and Quad owners, I think this is what you will find: Those who have tried a single sub usually go back to using no sub, and those who have tried two subs usually keep them in the system. I think this is because two subs exhibit less in-room smoothness discrepancy relative to a pair of dipole mains than does a single sub. So don’t fall into the trap of thinking "I’ll try one sub and if it’s an improvement then I’ll add another." One sub probably won’t be a worthwhile net improvement.

So to get back to your question, I believe two intelligently-positioned subs would be smoother (and therefore potentially "tighter") in-room than just one. Some EQ or other adjustability might be called for, because the amount of boundary reinforcement varies significantly from one room to another. If the two subs have continuously-variable phase controls, that might be sufficient adjustability: Set their phases 90 degrees apart to begin with, and adjust their relative phases from there (along with your adjustments of level and frequency).

I don’t mean to dismiss the qualitative differences between different subwoofer models, but I think the room-interaction advantage of two small subs intelligently distributed would probably outweigh the benefits that the single larger (more expensive) sub has to offer.

Duke

distributed multi-sub advocate

Thank you millercarbon for your kind words.

Gdnrbob made an insightful observation: "...it just makes sense that multiple subs will nullify/even out room nodes. Any of the subs listed would do that, when 4 or more were hooked up."

Yup. That’s why I don’t claim that my four small subs are the one and only way to do it. For example Millercarbon is building his own four-sub system, using a better woofer than the one I use.

And the subs don’t have to be identical either. So you can start with what you have and add on.

Duke

ieales wrote: "Some liked the Bose 901 spread. Others thought it appalling."

I assume you are implying that a distributed multisub system is analogous to a Bose 901.

It is not.

The reason is, the way the ear/brain system perceives room reflections at low frequencies is different from the way it perceives reflections at mid and high frequencies.

At mid and high frequencies, early reflections primarily widen (or "spread") the image and degrade clarity, and often cause coloration; while late reflections done right enhance ambience and timbre with essentially no detrimental side effects.

At low frequencies, the ear/brain system cannot distinguish between the first arrival sound and the reverberant energy. I can explain this in more detail if you would like. But the implication is that the reveberant field is virtually all that matters in the bass region.

"Not all of us require homogenous bass through out the listening room. Some want it dead nuts accurate in one small area."

I think you’re using the word "homogenous" to imply that smooth bass throughout the room is somehow undesirable.

If your bass is "dead nuts accurate" in one listening area, then it is "homogenous" in that area, presumably from a combination of steps you’ve taken, probably including EQ. A distributed multisub system can expand that area, which is desirable to some people but obviously not to you.

The two different approaches are optimized for two different priority sets, but ime "accurate" and "throughout the room" need not be mutually exclusive.  Assuming budgets are not unlimited, it makes sense to optimize for your priorities.  That being said, I have customers who have gone from single-equalized-ubersub to a Swarm (EQ use optional) and preferred the latter even in the sweet spot. 

"If I play just the subs rolled off @35Hz on programme with very little real bass, I can still localize the subs and off centre bass blindfolded. Is that possible with ’homogenized bass’ from a swarm?"

Some upper-frequency energy will come through the subs because the electrical filter rolling off the top end of the sub is not a brick wall. This upper-bass/possibly lower mid energy is what gives away its location. In the absence of much louder energy in that region coming from the main speakers, of course the sub’s location can be heard. But most of us listen with the main speakers on, and the loudness discrepancy is great enough to mask the subs’ locations.

Duke

"Everything HiFi is a compromise."

Amen brother!!

And I agree with you that 105 Hz is usually too high for something like a Swarm.  I try to avoid letting the subs run up any higher than 80 Hz for exactly the reason you describe.

There is a technique that works in situations where the subs have to kick in fairly high.  I had a customer whose 107 dB horn speakers shelved down significantly around 150 Hz, before rolling off at about 80 Hz. (His midbass horn was imo too small to hold up well down to the manufacturer's claimed 80 Hz low end.)

What we did was, use two amplifiers.  We placed two of the Swarm units along the front wall near the main speakers, and those two units went up to about 130 Hz (which blended well with the downward-shelving of the mains at about 150 Hz).  The other two Swarm units went on the side and back walls, and were rolled off above 60 Hz or so (blending well with the ballpark 80 Hz rolloff).  The lowpass filters were all 4th order.  It did take a little while to dial in all of the settings even with measurement equipment, but when we were done it worked well enough that he took out his checkbook.  He was in a position to easily spend up to twenty times the price of the Swarm, so that was a nice outcome.

Duke