It’s interesting to see the benefits of a distributed multi-sub system growing in acceptance in the nineteen years or so since Todd Welti and Earl Geddes, completely independent from one another, first began advocating for such. Disclaimer: I was an early adopter, and have been manufacturing a multisub system based on Earl’s ideas (with his permission) since early 2006.
One relatively minor point which I believe is worth mentioning has to do with low-end extension. My experience has been with four subs so that is what I will be referring to in this post, but the principle I’m going to describe applies whenever you have two or more subs spaced fairly far apart:
Four subs intelligently distributed around the room will often go a bit deeper than a single such sub would. Let me explain -
At the upper end of the bass region, say 80 Hz or so, the outputs of the four subs will tend to combine in semi-random phase. This is because the distances between the subs, and their distances to room boundaries, are comparable to the wavelengths or large fractions of the wavelengths. The in-room response smoothing we get from a distributed multisub system is because the outputs of the subs do not sum coherently.
Down at lower end of the bass region, say 20 Hz or so, the distances between subs and the distances to room boundaries tends to be much smaller fractions of a wavelength, so the outputs of the subs tend to sum approximately (though not precisely) in-phase. This of course varies with room and set-up specifics.
Theoretically, four sources combing in-phase will sum to 6 dB louder than four sources combining in semi-random phase. (For two sources, the theoretical difference would be 3 dB). Real-world the difference with four subs distributed around the room is less than the theory predicts, but we often get a 2-4 dB boost at the low-end from the more-in-phase summation at the bottom end of the bass region.
So the implication is, four small subs intelligently distributed will have a little bit more bottom-end extension than the performance of one individual sub predicts. By how much depends on the specifics, but 3-5 Hz seems to be in the ballpark.
One other small consideration that may not be obvious:
Since we tend to set the level of our subwoofers based on what sounds the most natural, if there is a significant peak in the bass region we will usually set the level just below where that peak is too obtrusive. So if we can tolerate a 3 dB peak and a single big sub has a 6 dB in-room peak, we’ll probably set the level where that 6 dB peak is only sticking up by 3 dB relative to the broadband SPL, which leaves the rest of the bass region 3 dB lower in level than it should be.
On the other hand if our subwoofers’ in-room peak is only 3 dB (regardless of how we ended up with the smoother in-room response), we can get away with setting the level of our subs 3 dB louder than before, so now there is a better level-match with the rest of the spectrum.
In other words, a distributed multi-sub system can allow you to turn up the bass region a bit louder, to a more correct level, before the worst of the peaks becomes obtrusive. This can result in better "slam" than a system which theoretically goes a bit deeper because we haven’t had to dial down the bass region a bit to prevent the occasional boomy note.
Note that the ear is particularly sensitive to small changes in SPL in the bass region, as predicted by the bunching up of the equal-loudness curves below 100 Hz or so. Therefore numerically small improvements in the in-room frequency response in the bass region tend to pay larger-than-expected dividends.
Finally, J.R. Boisclair was mentioned. Imo he is the world’s foremost expert on how to set up the four-piece multisub system I make, and to his credit he got there with zero input from me.
Duke
