Stereo Subs - Do they need to be identical?

There is a reason subs can be smaller now but you still can't defy the laws of physics. A small sub in a small box requires much more cone excursion than a larger design. For a small sub to equal the output of a large one it takes much more power and also equalizing to go as low. The driver also has to capable of much greater movement. Computer aided design and servo control has been a blessing when it comes to sub design. Also digital amps allow high power with low heat and low price. Crossing over as low as posible is the secret to seemingly fast bass response not the size of the sub.

If you can "hear" the sub, as opposed to "feeling" it, then different sizes should be discernible to your ears as differences in tone or volume. Therefore, I would prefer to go with identical subs. If you do use non-identical subs, I think that the crossover frequency should be quite low so as to avoid hearing any differences. Keep in mind that the open E string on an electric guitar is somewhere in the mid 30's in terms of Hertz. I wouldn't want to crossover any higher than that.

As far as sub speed goes, my experience is that the higher the crossover frequency, the slower and more sluggish it makes the main speakers sound. How fast a sub is depends on a lot of design aspects in addition to cone size so I wouldn't feel comfortable making any generalizations. Speaker designers can tell you a lot more than I can on this point so I'll defer to them.

>There is a reason subs can be smaller now

1. Solid state amplifiers technologies which generate less waste heat so you can run modest sized (and less expensive) heat sinks and stick them in a small box like class D (switch between full on and off so the output device is only dissipating heat when it switches) or G/H (switch between output rails to limit power dissipation) amplifiers.

2. Better understanding of math and computer modeling. 4th (3dB) and 6th order (I haven't looked at bandpasses) designs can be more efficient over a limited (real sub-woofers don't cover much over an octave) bandwidth and we don't need to rely on trial and error to build ones with relatively flat response.

3. Better motor designs help some. You can get more force for a given current + power to make drivers work in smaller boxes. It just might take 40 pounds of driver to get there.

>but you still can't defy the laws of physics. A small sub in a small box requires much more cone excursion than a larger design.

Nope. Output is purely a function of the air you move. With the air spring formed by a box and driver cone increasing in stiffness as size goes down, it just takes more force, current, and therefore voltage + power to get there.

At a fixed frequency, a given driver surface area will require the same excursion to reach a certain output level whether in a .5 cubic foot box or 4 cubic foot box; although 100W may be more than enough in the 4 cubic foot box while 800W isn't (thermal compression is a much bigger issue) in the half cubic foot box.

>For a small sub to equal the output of a large one it takes much more power

Right.

>and also equalizing to go as low.

You can get there with a combination of mechanical parameters and equalization, although if you've pushed the resonance down through a lot of mass the sub-woofer is going to be inefficient through it's entire operating range (at the top, power requirements are dominated by the current it takes to accelerate the moving mass). If you want a small box it's better to have a sub woofer that's inefficient in the last octave where there isn't much musical content but 10dB better in the second octave.

>The driver also has to capable of much greater movement.

Same movement either way. Only a resonant device (port or passive radiator helps). You'll need a more expensive passive radiator (driver without a motor) because of space constraints in a small box.

>Computer aided design and servo control has been a blessing when it comes to sub design. Also digital amps allow high power with low heat and low price. Crossing over as low as posible is the secret to seemingly fast bass response not the size of the sub.

It's the transfer function. The same transfer function (amplitude and time domains; distortion spectrum within reason; where the power response is coming from at high frequencies because of HRTF) sounds the same regardless of how you got there

You can get "fast" bass from a lack of low to mid bass. With power response summing +3dB over a single speaker when you feed the same signal to a stereo pair at higher frequencies and 6dB at low. Sound sources spaced out from a wall have nulls at their quarter wave length spacing (70Hz for 4' to the front wall). A low cross-over means you have the SBIR null in the main speakers' pass band and are no longer getting the artificial boost at low frequencies.

Of course, neither "fast" nor "slow" bass is what you get with real music like a dude going to town on an upright bass in your favorite corner bar. Real bass just is.

I would say that... it depends.

In my opinion the goal is to get a smooth in-room response in the bass region, and a smooth (perceptually seamless) blend with the main speakers. Depending mainly on the frequency response characteristics of the subs and their low-pass filter, their positioning, and the rolloff characteristics of the main speakers (whether or not they are high-pass filtered), it's possible that in some situations identical subs would give a better blend, and in some dissimilar subs would give a better blend. In general, I would expect dissimilar subs to be a bit smoother. Let me explain why:

In my opinion, most subs are not designed with room gain correctly taken into account. Here's a link to a graph that I think was generated by Martin Colloms, depicting "typical" room gain (roughly 3 dB per octave below 100 Hz):

http://www.speakerbuilding.com/content/1020/rge.gif

So if you take a sub that's "flat" to 25 Hz (anechoic), in-room it may well be +6 dB at 25 Hz - which will sound heavy and boomy.

Now suppose you have two subs, one of which is flat to 40 Hz, and the other flat to 25 Hz. The summed response of these two subs may well come fairly close to approximating the inverse of room gain.

If you start out with subs whose anechoic frequency response is approximately the inverse of room gain, then I think two such subs would be smoother in-room than two dissimilar ones.

Another factor to consider is where the subs will be placed, and the slope of their low-pass filter. Placing the subs far apart (but not in corners) is usually beneficial if smooth bass is the goal, but the farther the subs are from the main speakers the more important it is that they have a steep-slope lowpass filter so that lower midrange energy doesn't come through loud enough to give away their location. I personally favor highly asymmetrical placement, but if the crossover is high and/or the slope shallow that might not work well.

Duke

Using two different size subs can be very helpful in evening out room nodes. Place them at opposite ends of the room. With monitors as mains, I would use two identical subs. Proper integration is way more than twice as difficult, but the effort is worth it.