Audio Kinesis Swarm Subwoofer Awarded 2019 Golden Ear Award by Robert E. Greene

Recognizing member and contributor @audiokinesis for this award!!!
Ag insider logo xs@2xdavid_ten

Showing 14 responses by audiokinesis

Thank you very much, david_ten!!

The Swarm’s first awards were four years ago, in 2015, so I have been fortunate that the concept apparently has a good shelf life. Credit to Earl Geddes, whose idea I’m using (with his permission).

"Again, what about the swarm is unique other than being sold four woofers and told to place them a certain way?"

Totally valid question. As you have figured out, the concept can be implemented in many other ways.

The basic idea is to have a lot of bass sources distributed far apart around the room. Todd Welti and Earl Geddes were the first to advocate this specific approach to the best of my knowledge; Todd investigating symmetrical placement strategies and Earl investigating asymmetrical ones. They developed their ideas at approximately the same time but were unaware of one another’s work.

I don’t think any one feature or characteristic of the Swarm is unique, but I think the combination is probably unique, at least in its general price ballpark:

- The four passive subs are optimized for lower in-room extension than you normally see from 10" subwoofers (20 Hz ballpark), in part because we know in advance that we’ll be using four of them so we don’t need to optimize for max SPL.

- The single shelf-mounted amp that drives the four subs has some nice features, including a 4th order lowpass filter, a single band of parametric EQ, and a continuously-variable 0-180 degree phase control. And since it sits on a shelf, it’s less likely to fail (better cooling and less vibration) and easier to replace if it does. A second amp can be added for some performance improvement.

- The native response curve of the individual subwoofers is the approximate inverse of "typical" room gain. The individual subs have a pluggable port, which can transform them into low Qtc sealed boxes, better suited for small rooms. In my modest-sized living room, the ports are sealed in three out of the four.

- It’s easy to reverse the polarity of one of the subs, which usually improves the in-room smoothness (and smooth bass = fast bass).

- The woofers I use have unusually powerful motors. Their combined motor strength is greater than any single home audio woofer I am aware of (including the discontinued monsters from TC Sounds, Aura, and Acoupower), and to the best of my knowledge surpassed only by the most powerful 21" prosound subwoofers. Not sure "motor strength per dollar" is a thing, but if it were, the Swarm would score pretty high.

- WAF isn’t too bad. With the port and input terminals on the bottom of the box, and the woofer on a side, those sides faces the wall and then the subs just look like four wooden blocks, each with a footprint of one foot square and height a bit less than two feet.

In other words, I didn’t just package four subs together and concoct a marketing scheme. I started with Earl’s concept and made my design choices with a specific end in mind.

That being said, there is nothing about the Swarm that a hard-core DIYer couldn’t equal or surpass. Here is the amp that I use, in case anyone wants to give it a go:

Leland Crooks of Speaker Hardware offers 12" passive subwoofer kits that I designed, four of which would outperform the Swarm.

" Would this work if the speakers weren’t on t floor but up higher say close to the ceiling or in the ceiling facing down?" 


What would be really sweet is, three subs near the ceiling and the fourth one closer to the floor. This would usefully distribute the sources in the vertical dimension, complementing the distribution in the horizontal plane. But if that’s not feasible, four on the ceiling is just as good as four on the floor. 

Also, it is not necessary that all four subs be identical. Earl’s personal first-generation multisub system included a single massive ubersub that went considerably deeper than the other three.

" If I didn’t already have 8 subs... "

Your eight subs intelligently distributed are probably twice as smooth as my four.

In general, the in-room smoothness goes up, and the spatial variance (variation in frequency response from one location to another) goes down, as the number of intelligently distributed sources increases. So two subs are twice as smooth as one; four subs are twice as smooth as two; and eight subs are grounds for divorce in most jurisdictions.

So I stopped at four.


Rooze wrote: 

" I can’t get beyond the notion that some subs just don’t sound very good, and multiplying an average sounding sub by four surely doesn’t equate to good quality bass.... how can it?" 

I think the suitability of the subs for this application matters. Back to that in a moment.

The biggest issue at low frequencies is room interaction. Compared to the few dB difference between good subs, the peak-and-dip patterns that rooms impose on subs are huge. By spreading multiple subs around the room, we get four significantly different room-interaction peak-and-dip patterns. The more different peak-and-dip patterns we are summing, the smoother the net result.

Now ime subs with huge bottom ends are generally not the best choice for a distributed multi-sub setup. My understanding is that your Hsu’s have pluggable ports, and ime that can be very helpful in getting the in-room bottom end balanced right (it’s what I do too). If you had four Hsu’s with pluggable ports, I’d say start out with the ports plugged in all of the subs except for one, the one nearest a corner. One sub in one corner is okay, but I suggest no more than one. If the Hsu’s have phase adjustment, put the one farthest from the main speakers in reverse polarity, unless that’s the one with the ports open.

Then if you need more low end, unplug one port (your choice which one). If you need less low end, plug ports in that sub nearest a corner. I have zero experience with a quartet of Hsu’s, but this is what I would try. The ability to plug ports in the Hsu’s may make them actually more suitable fort this application than most subs.   I'm not suggesting you buy two more Hsu's - just tossing out an idea of how Hsu's might be used. 

Avanti1960 wrote:

"I have successfully integrated one and two subwoofer solutions into my system when the main speakers had a natural roll off (acoustic crossover) that made for a near ideal low pass frequency relative to the subwoofers. However I was not so successful with a recent set of main speakers that had extended low frequency output to 20Hz from the transmission line port."

I’m not aware of anyone using a Swarm with main speakers that go down to 20 Hz. Overlapping that much with the mains wasn’t part of my game plan. What be inclined to I’d try is, reducing the output from the transmission line terminus (probably with open-cell foam) so that the subs can be turned up louder without making the low end too thick.

"I am now on a new set of main speakers and although I love the bass, measurements show that it reaches a certain output and levels off while the rest of the frequency spectrum can get louder. I miss the louder bass with more impact as the volume increases."

I’m not clear about what the issue is. Is the bass region compressing? Can you describe in more detail what the measurements show?

"Would the 4 sub array lend itself to better integration with main speakers that are near full range (avoiding bass quality and blurred midrange issues) without high passing them?"

So far we’ve been able to integrate well with unfiltered main speakers that go down as low as 30 Hz. Sometimes with overlap, sometimes not. Sometimes using the single band of parametric EQ and/or the phase control to reduce the amount of reinforcement in the region of overlap. "Blurred midrange" makes me think the subs were being rolled off too high and/or too shallow, and may have been too loud.

"In my two-sub setups I found that I needed to adjust the phase of each individually for best integration. Why wouldn’t I need to do this with the swarm system? "

I’m not an expert on optimizing two-sub setups, so I can’t thoroughly compare and contrast. But in general we want the in-room bass energy de-correlated for best results, and that includes for blending with the mains. You were introducing some de-correlation by having two subs instead of one, and some more via the different phase settings. The Swarm introduces a lot of decorrelation via the widely-spaced asymmetrical placement and reversing the polarity of one of the subs. Some Swarm systems use two amps, usually set 90 degees apart in phase.

I haven’t had anyone yet tell me that they can’t get a good blend between the Swarm and their mains, whether they are using one amp or two. I think the decorrelation that a distributed multi-sub system brings to the table help make that blend less problematic to achieve.

But you can certainly use multiple powered subs of your choosing and thereby be able to adjust phase and frequency and gain of each independently.

Hans asked:  "What are some ways you can suggest to position (suspend/hang?) the subs near the ceiling?" 

Thus far, to the best of my knowledge, all of the Swarm units that have ended up near the ceiling have been placed on the top of bookshelves or something similar.   I don't have any ideas for how to suspend subs from the ceiling... no GOOD ones, at least! 

Audiomaze wrote:  "Can this work with passive subs and a powerful amp driving them in series." 

Yes.  I use series-parallel connections for the four 4-ohm subs to present the amp with a 4 ohm load.  

" If I understand correctly, one of the ceiling subs can be in a corner with polarity reversed. " 

I normally suggest one sub in a corner (but no more than one), and that a different sub be the one with its polarity reversed... but, once your check clears, you can try whatever you want! I think it is possible that, in a small room, reversing the polarity of the one in the corner could work well.

I just wanted to make it known that, in addition to Jim Romeyn's DEBRA which differs from my Swarm in footprint (his is rectangular, mine is square), Vinh Vu of Ginko Audio offers a slimmer and trimmer Swarm-type system as part of his Sextet, which was introduced at Capital Audio Fest a couple of weeks ago: 

Vinh is using two of the subs as speaker stands in that system, which works very well with his patented Arches resonance-control support system.  The Arches add stability without adding visual mass, as you can see at the link above, and can be used on a regular Swarm or Debra set as well, especially if two of them are going to double as speaker stands.  Jim and I will be offering the Arches as an option. 

Just for the record, I don't get a kick-back from Sextet subwoofer sales. Vinh paid me in Arches for my design work (it was a collaboration but I did the math with the same target curve in mind as for the Swarm and DEBRA). 

"Does anyone here have experience pairing a Swarm/DEBRA with relatively full range this instance the Vandersteen Treo’s?"

Installo43, one of my customers has Vandersteen Quatros.

Before getting the Swarm, he had been using one sub and a Meridian processor with 16 of its available 60 digital filters professionally calibrated to smooth his in-room bass response.

He replaced his sub with the Swarm, ran the Quatros full-range with their filters set to neutral, and hired the same (highly experienced) technician to make measurements and calibrate the filters on his Meridian processor. The technician found that the ONLY adjustment needed was to the level of the subwoofer amp. No further equalization was needed, either from the Meridian processor or from the Vandersteen’s analog filters (well, the Meridian did equalize the two surround speakers a bit, but that doesn’t affect what the Swarm was doing). He wrote that the Swarm "had rendered both the Meridian Room Correction and the Vandersteen’s analog equalizers unnecessary" for his room.

Quoting from his e-mail:

"The technician, J___ S____, who has performed dozens of these calibrations, said he has never seen anything like this. The room is rendered literally flat in frequency response and spatial energy distribution - sonically the room disappears. We played one of Kal Rubinson’s recommended demo discs, the 100th anniversary for the San Francisco Symphony Orchestra using the John Adams piece, "Short Ride on a Fast Machine", and you would swear you were in the hall. You can "feel" the ripple in the tympani skins! Very impressive."

I DO NOT claim that these results are "typical", but apparently it is possible to use the Swarm with a set of Vandersteens.

got a dog in the fight
Installo43, yeah I don't think I could buy a testimonial like that one if I wanted to!
In your situation I think we'd use essentially the same approach... run the Treo's fullrange, and use the Swarm to fill in the bottom octave or so. 

I have owners of speakers like Wilson and Magico who run their main speakers fullrange and then use the Swarm for the very bottom end.
@noble100 wrote: "I was just going to contact you with a question I have about whether room bass treatments negatively or positively effect the performance of the Swarm/Debra bass systems in a given room." 

In my opinion room treatment in the bass region is virtually always beneficial, as the improved damping reduces the magnitude of the peak-and-dip swings.


" The bottom octave is generally considered to be 20Hz to 40Hz, the second octave 40-80Hz and so on. "
This is what I had in mind when I was talking about "the bottom octave or so".  

I don't normally recommend running the Swarm higher than about 80 Hz because the ear may be able to detect the locations of subs that are well away from the main speakers, and that would be distracting.  That being said, there are work-arounds if the need arises to cross over higher up.


Luisma31 wrote: " no no, with your midbass driver I think the 80 hz works wonderful, actually I run mine 60 below, don’t you dare touch my crossover Duke, sounds perfect the way it is." 

Thank you Luis! 

I was referring to the hypothetical situation where an 80 Hz crossover was not high enough.

One time we were integrating a Swarm with a system that was horn loaded down to 80 Hz, but it turned out that the response started rolling off at about 150 Hz and then the rolloff accelerated below 80 Hz. So we placed two Swarm units near the main speakers and drove them with one amplifier and rolled them off at about 140 Hz, and then the other two were placed much further from the main speakers and were rolled off around 60 Hz. We used the parametric EQ’s to fine-tune the blending.

Something like that would be the work-around if the crossover needed to be higher than 80 Hz. 


"I guess speakers that are weak between 80 and 140 are not very common ..." 

I agree, not very common. This was the only such case I have been involved with.