What process did you use to integrate multiple subwoofers for 2 channel listening?

@tvad, thanks for the thumbs-up! Not that this will ever be a question on Audiophile Jeopardy, but the the first generation Swarm came out way back in 2006.

@mapman wrote: “One other thing. I know audiokinesis is big on speaker dispersion patterns and how that affects the sound.”

Ha! Yes I do think in terms of radiation patterns quite a bit. From a psychoacoustics standpoint the room interaction situation is very different for different parts of the spectrum. Our brains have time to process the direct sound separately from the reflections at shorter wavelengths, but not down in the bass region, so I use different approaches to room interaction - i.e. different radiation patterns - in different frequency regions, the details being beyond the scope of this thread.

@hilde45 wrote: “The problem I’m not able to solve, and perhaps the room’s physics make it impossible, is a +5db or more bass hump from about 73-92 Hz."

Is this bass bumpage present throughout the room, or is it mainly present in your normal listening area?

Approximately where is the crossover frequency between your subs and main speakers?

What controls do you have on your subs?

How much freedom do you have when it comes to the locations of speakers and listeners in your room?

Thanks,

Duke

@hilde45 , thanks for your answers. You’ve obviously worked very hard on this and your results really are quite good, which makes that 88 Hz peak stick out like a sore thumb all the moreso!

Might it be a floor-to-ceiling mode? Do you have a lower than normal ceiling?

At any rate, here are a few ideas:

- I agree with @erik_squires that THIS is the sort of thing EQ is good at fixing (bass traps too, but I think EQ would be more cost-effective).

- You might try placing the Salks (very nice speakers, by the way) about six feet out from the wall behind them, which will result in the low frequency reflection off that wall arriving ½ wavelength behind the front wave at 88 Hz, hopefully providing some useful cancellation.

- Place one of your subs near the Salks. Run it up high enough that its output overlaps with the Salks in that 88 Hz region, and then adjust the phase control until the sub is CANCELLING the Salks enough in that region to make the peak live-with-able. If you can’t get the cancellation you need going from 0 to 180 degrees, then use the speaker-level inputs on that sub (if you aren’t already) and reverse the polarity of the signal going to them, such than now your phase control is effectively doing 180-360 degrees. You may need to do this overlap/cancellation trick with more than one of your subs, and/or you may need to use the parametric EQ in one or more of your subs to enhance the cancellation effect.

- The contribution from the Salks across the bass region may result in too much net bass energy before you can fully take advantage of what the subs can do. If so, stuff some open-cell foam in the slot ports of the Salks to disrupt the airflow, turning the box into an aperiodic enclosure. If this is a net benefit, you can look forward to further hours of audiophile fun as you fine-tune the foam-in-the-slots. 

- Another way to reduce the contribution of the Salks in the bass region is to wire a small capacitor in series with the amplifier’s input, perhaps by performing amateur surgery on an inexpensive interconnect. You can use an online first order highpass filter calculator, substituting the amplifier’s input impedance where the tweeter’s impedance would normally go. You will probably have some trial-and-error chasing down the best capacitor value.  (Maybe your system already has an easier way to roll off the bottom end of the Salks, but if so I overlooked it.)

- As a ballpark rule of thumb for a distributed multi-sub system (I avoid using the otherwise-convenient term “DBA” because it can also mean “Double Bass Array”, which is something very different), you want to get the subs spaced as widely apart as you reasonably can while distrbuted in as many dimensions as you reasonably can. My mentor Earl Geddes finds two subs are much better than one, and three subs are much better than two, and four subs are an improvement over three but not by as much. So imo you are definitely playing the distrubuted multisub game already! If you can elevate one of the subs so that it is closer to the ceiling than to the floor, you get bonus points. This is one of the things that @hleeid did to get good results in his 11 x 13 foot room; I think he actually elevated three of the four Swarm subs, as that was more practical than the other way around in his room. In your case though, I would still keep at least one of the subs near the Salks if you try that overlap/cancellation thing. For any of the subs which are not near the Salks, and especially if they are physically closer to you than the Salks, you want to roll off their top ends lower and steeper, so they don’t betray their locations by passing upper-bass energy loud enough for you to hear. If that’s not possible, then imo ideally you’d want those other subs to be farther away from you than the Salks so that their output is not the first-arrival sound.

- In my experience it is usually beneficial for there to be some deliberate phase variation between the subs, so don’t be afraid to twiddle those phase controls. The effects of the phase controls are more subtle than the effects of the subs’ volume and low-pass frequency controls.

Hope there’s something useful in this.

Duke

@hilde45 wrote: "...the orange "morning" line is speakers only.  Improvement from "Morning scan, without subs" to "Afternoon scan, with subs" mitigates the dip significantly at various points, e.g. pulling up a null by 10 db at 129 hz, by 8 db at 241hz."   

Maybe I'm just slow, but eyeballing the curves I would have thought that Orange was WITH subs, and Green was speakers ONLY.   

Anyway assuming that's the only difference, AND that the microphone locations and other measurement conditions are identical, the curves imply that the subs are active pretty high up.  Where are you rolling off the top end of the subs?  

Incidentally the dip around 270 Hz and peak around 540 Hz look to me like floor bounce artifacts (perhaps modified by the ceiling bounce) and imo should NOT be EQ'd away.  

Duke