subs and small room

Small rooms are the hardest to get smooth bass in. If you are really serious about getting good bass in a small room, there is a solution but I must warn you - at first glance is seems insane.

First a bit of background. You may have noticed that moving the speakers, or moving a subwoofer, changes the bass at the listening position. This is because how the low frequencies interact with the room depends not only on the room's dimensions, but also on the location of the low-frequency source(s) in relation to the room's boundaries as well as the listening position.

Using multiple subwoofers will result in smoother in-room bass because where the response of one sub is "zigging" due to its room interaction, the response another sub is "zagging" - so the net result is smoother than any one sub could be. This isn't my original idea; I learned it from Earl Geddes, but it is also supported by Todd Welti.

"With enough subwoofers, it is theoretically possible to cancel out all modes in a room." - Todd Welti, "How Many Subwoofers are Enough". In a small room, the problem is modes that stick out like a sore thumb, and using multiple subs is the solution. Small subs though - we don't need big ones.

"The spatial variations, and to a certain extent the frequency response variations, will go down (get smoother) as 1/N, where N is the number of independent sources." - Earl Geddes, Speaker Asylum Post 10-30-05. This give us an idea of how much improvement to expect from each additional sub.

As a practical matter, Welti suggests four subs as enough to get a significant smoothing of the bass. Here is a paper he wrote on the subject:

http://www.harman.com/wp/pdf/multsubs.pdf

Whereas Welti recommends symmetrical placement of the subs, Geddes advocates asymmetrical placement:

http://www.gedlee.com/downloads/sub%20study%20.pdf

I suggest using four small subs that have a steep-slope low-pass filter (4th order, rather than the more common 2nd order) to allow greater flexibility in sub placement. If you can't hear the location of the subs, you have more freedom to experiment. I use a four-sub setup in my listening room when I'm using speakers that call for subs (like hotrodded Maggies).

Duke
dealer/manufacturer

Duke offers a lot of "sound" advice. 2 subs are almost minimum, but you can sometimes get away with one sub depending on where it's crossed over and the use of a good parametric EQ. Quite often in small rooms the crossover point may be as low as 35 Hz, which may be lower than the first 1/2 wavelength of any axial mode (roughly 17 feet). Thus you really aren't exciting any of the modes in that application.

We used a Talon Thunderbird and Rives sub-PARC at RMAF and won an Audio Oasis award. This was nearly the case in that we did not really need the parametrics in the sub-PARC because we crossed over so low.

Rives, thanks for the kind words - you could so easily have accused me of double-talk!

Seriously, my apologies for the double-post. I thought I was editing (instead of double-posting), and then my connection failed and I couldn't get back online.

Duke

Tobias,
Its not frequency doubling that occur with room modes but amplitude doubling or cancellation i.e peaks and dips.

Duke,
Multiple subs would only work in a large room. In a small room this would only create nodes (peaks/dips) at mid-bass and higher frequencies. So the result without any other form of control e.g EQ as suggested would be mid-bass prominence.

Samuellaudio,
Can I suggest you plot your room (Rives CD2 plus RS SPL meter) to know where you stand currently and then do it again with a sub i.e a loaner. You will then know for sure how much the sub affects the room.

Athipaul, I don't see how multiple subs in a small room "would only create nodes (peaks/dips) at mid-bass and higher frequencies".

The subs are not contributing at mid-bass and higher frequencies, so they do not "create nodes". They do smooth the in-room response in the region where they are contributing, as I will explain.

The problem in a small room is not too many room-induced peaks and dips; the problem is too few! In a large room you will have more room-induced peaks and dips bunched up closer together, and the ear tends to average them out across 1/3 octave intervals (called "critical bands) so the individual peaks and dips don't stand out - just the broader trends. In a small room, the room-induced peaks and dips are too far apart for the ear to average them out so they tend to stick out like a sore thumb. But by using multiple subs spread around in a small room (each contributing its own unique peak-and-dip pattern at any given listening position), we approximate the more dense peak-and-dip pattern that exists in a large room, and the ear's smoothing mechanism can work in our favor. I have done some simplifying in this one-paragraph explanation; you might want to take a look at Welti's paper.

Equalization is not necessarily the panacea it seems to be at first glance. Equalization can smooth the response in one listening position, but actually make it worse elsewhere in the room because the peaks and dips move around as the listening position changes, so from a different listening position the equalization may well be boosting a peak and/or cutting a dip. Multiple subs gives a more uniform response throughout the room, so that any equalization still called for will probably be beneficial throughout the room rather than in one location only.

Duke