Does size matter?

March 25, 2013
 by Paul McGowan

When it comes to subwoofers, the answer’s yes; size matters greatly. If you have a couple of 8″ drivers in a box you’re calling a subwoofer, you’re going to be disappointed if you follow the setup steps we’ll be discussing – because you won’t get the results we’re getting. Big is better as long as you’re not asking it to go too high in frequency.

I am always reminded of the classic film Crocodile Dundee (the first one) when the local tough tries to rob Mick Dundee with a small knife and Mick whips out his big bad blade and says to the tough “that’s not a knife, THIS is a knife” with a knowing smile shared between two tough guys. So let me just start of saying to those of you with small subs, “that’s no subwoofer, THIS is a subwoofer”.

The simple fact is surface area is a big deal for moving air and achieving low frequencies. The lowest notes on a pipe organ,B̦sendorferImperial Grand Piano and a tuba is around 16Hz Рa frequency considered below the threshold of human hearing. These instruments do not create single notes at this frequency but what they do generate is full of higher frequency harmonics Рthe harmonics giving the instrument its sound signature Рand those higher frequencies in combination with the low fundamental give a full visceral feel to the note. When played in a live space, any one of these instruments will produce sound you both feel and hear Рyour stereo system should do the same.

To hit realistic in-room levels of undistorted clean bass at 15Hz is quite a feat. Carver’s Amazing subwoofer claims to go down to 18Hz with a 12″ woofer, the same claim is made with my current sub the Martin Logan Descents that have three 10″ woofers, but I suspect these claims are made in a close mic setting.

The problem with too small a surface area has to do with coupling the air in the room – which at very low frequencies is not linear and as the frequency goes down, the surface area of the driver has to go up rather dramatically to get what you want delivered in the room.

If I were building a subwoofer I’d start with an 18″ woofer, or at least two 12″ drivers. The total surface area of two 12″ woofers is 226 inches, while the total surface area of an 18″ driver is 254 inches – so nearly the same – my three 10″ drivers are about 240 inches. A single 12″ woofer has only 113 inches and two 8″ drivers about 100″.

Here’s something else to consider – a single 18″ woofer is better, in my opinion, than two 12″ woofers or three 10″ drivers because a single piston couples more effectively than multiple smaller drivers. Sub manufacturers like using multiple smaller drivers for three reasons: they are easier to source, have less mass so it’s easier to get a faster transient response and their higher frequency performance is better than a single big woofer.

I would counter these arguments by suggesting the sourcing of a driver should be based solely on performance, not economic considerations. I would also suggest that given a proper motor assembly one can have any transient speed desired and lastly, in a good full range setup who cares about going higher than 60Hz? Ok, I got that off my chest and, since I don’t make subwoofers, let’s move on to the real world of what most of us are likely to own.

So, I imagine nearly none of us have an 18″ powered subwoofer around and in all likelihood, we are probably going to have a single 12″ woofer per side. Oh, did I say per side?

Yes, sorry, but the myth of the single sub being adequate is plain incorrect. Stereo subs are the only way to go – unless you simply cannot afford to have two. In that case, I fully understand the limitations we all face in this hobby so I’ll try my best to help you make a single sub work as well. But by all means, if you’re financially capable of having a stereo subwoofer system, then that’s the best case for your musical setup.

Lastly, make sure you have a great subwoofer. I’d sooner have one great subwoofer than two POS.

Subscribe to Paul's Posts

12 comments on “Does size matter?”

  1. Paul,

    I knew you were going to say we needed two subwoofers. Oh well. I think the most interesting bass I have heard was at an Easter Sunday service at Westminster Abbey a number of years ago. In addition to the organ 32-footers crank’n out the bass line on the hymns, the organ and members of the London Symphony brass section performed Marche Triomphale by Louis Vierne. No way a couple of 18 drivers going to reproduce that! All you need is a stone quarry, a thousand stone masons and about 80 years, plus one very large organ (that didn’t come out right). Nothing like transepts and dozens of stone columns to set up some nice reverb and echos that take 15 seconds to decay.

    Paul, I am hoping you will address cable quality issues for line level feeds to subs. I believe you will recommend against line level inputs, but that is what I have. Does it pay to use a special subwoofer cable, or will any good full range interconnect do?

    1. Line level is a really tough one – it depends on the preamp or DAC driving the cable – which is usually long. If the unit is capable of driving the long cable you’re fine, if not, then a high level input is probably better.

  2. A properly set up audiophile system can reproduce this natural reverberation quite well – with the caveat that all of us would choose attending a live concert in Westminster Abbey! It would certainly engage senses that our system can’t : )
    Paul, you are correct as far as my own experiences tell me – two 18″ subwoofers seem to couple with the room better than two 12″ subs. You may want to delve further into the size of the room in this equation – the larger the volume of air in the listening room, the more piston area required. AS SPL stands for sound pressure level, I’m sure that someone has done the math to establish the proper square inch piston area required to achieve and maintain the appropriate pressurization per cubic feet of air. I’ve never looked into it – would it be covered under fluid dynamics?

  3. Once again I think we are going to disagree. IMO the design of a subwoofer involves multiple factors and the size of the drivers is only one of them. I’ve yet to hear an 8″ subwoofer that’s worth a tinker’s damn. But I’ve heard sound columns of half a dozen 12″ woofers used for sound reinforcement in auditoriums that would make awful high fi subwoofers as well. In theory you could build small subwoofers that will perform very well. What is a woofer or subwoofer? It’s an air piston driven by a linear motor. The problem is to get it to move enough air and to operate without a resonant peak over a desired range of frequencies. To move enough air it doesn’t matter if it’s a ten square inch area driver moving over a stroke distance of one inch or a one square inch driver moving over a stroke distance of 10 inches if everything else is the same. But between the two choices, the single larger driver moving a shorter distance is more practical. However the same result can be obtained by using multiple small drivers IF you know how.

    In the early 1950s it was accepted that to produce low bass notes you needed a very large enclosure. Bass speakers in those days were horns, bass reflex, and infinite baffle designs. Many looked like clothing closets. In fact some DIY speaker enclosures were clothing closets. And then alone came a speaker system with a 12″ woofer (some call it 11″ but it’s defined by the distance between opposite screw holes) in an enclosure less than two cubic feet that blew the rest of the industry away including the mighty klipschorn, the largest Bozaks, Altecs and EV theater speakers. It could reproduce the lowest audible octave without audible distortion the others could not reproduce at all. It became for a time the world reference for undistorted low bass output and the choice of many serious audiophiles, music lovers and was used in laboratories where such equipment was needed.

    I’ve mentioned in a previous posting why my design of choice is the now out of general favor acoustic suspension design. What a coincidence in timing a debate rages on at another site about bass. Here is a column by John Marks with some explanation;

    Not being an audiophile anymore (for nearly 40 years now) I don’t follow the equipment of the moment on the market but reading this I was surprised to learn that the same Johnny-one-note problem that bedeviled ported designs in the 1950s and 1960s is still with us today. This is the result of the spring constant “k” in Newton’s second law being a strong function of frequency. Air slides back and forth though the tuned tube at one frequency easily, efficiently, but not at all easily or efficiently at other frequencies. It is a resonant air column in the true sense of the word just like a pipe in a pipe organ or any horn instrument. Marks points out two other aspects of the resonant column that have not been overcome (the physics never changes.) One is the lack of damping “b” in Newton’s equation. This means when the applied voltage to the woofer goes to zero, it still keeps resonating for awhile, it doesn’t stop quickly. Energy is stored in the air mass in the port and there’s no frictional loss dissipating it. The energy stored in the air mass also presents a problem of back EMF for the amplifier to damp out by driving the woofer turning it into a generator. This can be considerable, much worse than the mechanically heavily damped acoustic suspension speaker with a larger cone mass.

    John Marks further adds;

    “The advantages of a sealed box are so compelling that, when he was visiting me to hear and measure the ESP Concert Grand SIs, John Atkinson was moved to query, “Why doesn’t everybody use sealed-box designs?” He then answered his own question: “Because nobody has ever gone broke selling a loudspeaker that booms.””

    Can multiple small acoustic suspension woofers equal the low frequency bass output in terms of distortion, FR extension, flatness of FR, and total output if correctly designed and built? Experience re-engineering original Bose 901 leads me to the conclusion that the answer is an unqualified yes. Size is only one factor and if other factors are wanting, size alone cannot overcome them.

    1. Thank you Mark, as always, well thought out and illuminating. Let’s start with what we agree on: sealed boxes. Ported subwoofers simply do not cut the mustard. On this we agree 100%.

      We can also agree that multiple woofers coupled together can be effective although I would maintain that to my last thought here, they aren’t quite as effective as a single cone of the same size.

      On the issue of size I think you may be neglecting one thing – or if not, then my info is wrong. To my knowledge, air is resistive only above a certain frequency and reactive below that frequency. Below the crossover point it becomes exponentially harder to couple the air to the piston – hence a larger surface area woofer has an advantage over a smaller area.

      1. One problem with small woofers is that their Xmax is usually limited. Even if multiples are in aggregate of comparable surface area to one large woofer, they generally can’t move the same quantity of air. Original Bose 901 has the aggregate surface area of about a 14″ woofer. (I measure the effective piston area of a woofer from the inner edge of the suspension on one side to its opposite side, the angle of the cone irrelevant because it’s only effective in the direction perpendicular to the direction of travel. A 12″ AR woofer actually has an effective radiating area of 16pi square inches, Bose 901 about 20pi.) Small woofers usually have a fairly high Fs also but in an acoustic suspension design this can be mitigated in several ways, one of which is to increase the effective mass of the cone, the other to electrically equalize the output which works up to the driver’s maximum electrical capacity or Xmax. But AR 12″ has an Xmax of about 3 times Bose 4″ CTS drivers. As I pointed out in a prior posting, Bose deliberately pushed the F3 of the system high making even less efficient at very low frequencies than it had to be. Be assured enough of them adequately equalized can produce as much bass as is needed for any application down to 23 hz as measured by CBS Labs I think (or was it Julian Hirsch?) Also consider a high quality headphone. It can produce sound down to 20 hz without distortion. Suppose you had thousands of them playing at once. A small very long throw woofer just isn’t a practical thing to build.

        Insofar as the fluid dynamics of air is concerned, I’m not entirely knowledgeable at high pressures and I suppose I should learn more about it since acoustics is actually a application of the fluid dynamics of air. All I know is that both nitrogen and oxygen behave at normal temperatures and pressures as nearly ideal gases. Can a small throat create a lot of low frequency sound? Just look at the relatively small throat of a jet engine. Even 10 miles up it can create a sonic boom that will shatter windows for miles around. Of course it takes about 2x 25,000 horsepower to do it. That’s about 36 million watts. I don’t think even your class D amplifier will put out that much power 🙂

    2. I basically agree with you about closed boxes. They are better controlled than relex designs. But most listeners are used to less than optimally designed reflex designs. And a good reflex may be number 2 to a “good’ closed box but properly designed it ain’t so bad.

      My own woofers are 18” drivers in a box designed to be a good reflex(bessel function which is minimum time delay). with -3 db without room lift at 23 hz. I used them this way for years with great satisfaction. And I can’t stand sloppy bass at all. But I think I got into the discussion about your woofers and closed boxes with John Marks. And i recalled my designer friend made my boxes to be tight with the port stuffed and -3 db at about 32 hz. I tried them and loved them a little better. The already tight bass was even tighter and more defined. But there was atrade off. There was a dryness and loss of’ ‘life’ to the bass. I still preferred the closed set up and use them that way still. But the designer who agrees with waht we both hear prefers the reflex set up for the extra life even though he knows it’s a bit less accurate. Apples and oranges I guess. And I’ve discussed this with the TAD designer who uses reflex bass. He also agrees but says he gets an extra 10 hz with the reflex and believes the bass quality is the same as the closed box down to the closed box -3db point.

  4. I have a friend who is a speaker designer who likes to say if you want faster bass get a better tweeter. It’s the speed of the overtones that is the speed of the woofer. Bass fundamentals are not fast.

    My feeling is that the larger problem of bass speed is not the initial attack but the decay which almost always contains some ringing to satisfy the buyers who want to hear bass whether it’s in the soft ware or not. They feel cheated without over full mid bass(most don’t know what deep bass really is). And fast bass is really well damped bass with minimal over hand, sort of the opposite of what one first thinks it is. It’s not the guy in the 100 meter dash who gets there 1st but the one who stops in the shortest time.

    1. I think I’ll say something about this myth about speed. If two drivers of the same diameter are playing at the same loudness, no matter whatever differences there are between them such as efficiency, at every point in their travel they are moving at exactly the same speed. What can be different is what’s called group delay. This is the time gap between the instant voltage is applied to the voice coil and the time the corresponding mechanical response occurs. This is important in so called time aligned speaker because at the crossover point you have two drivers that may respond to the same signal with two different delays. With a steady state tone this will cause reinforcement and cancellations to occur. At the same loudness the FR can have a 3 db peak if they’re in phase, a zero output (infinite dip) if they arrive exactly 180 degrees out of phase. But the phenomenon is both frequency dependent and also a function of the physical location of both drivers and your ear.

      Generally the smaller drivers have shorter group delays than larger ones due to lower inertial mass.

      Strategies for correcting this are to place the tweeter further from the listener than the woofer making the distance its sound has to travel longer, selecting drivers with comparable group delays, tricky crossover network designs that usually don’t work. Digital delay lines for the smaller “faster” drivers should work. Putting the subwoofer adjacent to the wall while pulling the rest of the speaker closer to you away from the wall only makes this problem worse. Does it matter? I’ve yet to hear of even one test where identical sounds except for time alignment was the only variable were conducted so we don’t know. But even perfectly time aligned speakers will have phase interferences if they drivers are not coaxial and you are not directly on axis due to geometric considerations. As for the speaker being “fast” to stop, that’s a matter of undamped resonances which are virtually inevitable with ported designs. They are a consequence of their high Q. Their peaky bass output depends on an undamped resonance. BTW I don’t trust those measurements where they combine the output from the driver with the output from the port to get a summed composite. It will take a lot more to convince me they’re valid than I’ve seen so far but I have relatively little experience with them. I avoid ported designs.

  5. Paul,
    One of the more interesting subwoofers in the last couple of years, that flew in the face of accepted design parameters, was BG’s BGX-4850. It used 48×4′ woofers, each driver in its own sealed enclosure. 12 drivers loaded into a slot, in opposing pairs, then put in two and sometimes four different locations. (48 drivers altogether) The original design was done by Laurie Fincham. The measured room distortion was very low, depending on room placement of course. Near field distortion was very good, and it had a fairly flat response to about 20Hz.. One of the ideas on why the distortion was so low was that that the smaller cone had very little deflection. The slot didn’t seem to create port resonance.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram