Anyone who has been into audio for a while has probably developed a preference for a certain style of speaker and a certain driver configuration. This preference might be based on personal experience, the opinion of a trusted reviewer, technical knowledge, or maybe even some efforts at DIY speaker building. There are actually lots of ways to build a good speaker, and there’s certainly room for personal taste. Still, I’m often surprised at how little most people (even some who are deeply involved in audio) really understand about the performance implications of different driver configurations.
Much of this is explained in The Loudspeaker Design Cookbook. When I first read The LDC 20 years ago, it probably increased my understanding of how speakers work by an order of magnitude. What I’ll present here is much less technical. I encourage all audiophiles and reviewers to dig deeper by picking up a copy of the LDC, which is now in its seventh edition and which was written by Vance Dickason, editor of Voice Coil magazine and one of the world’s most experienced speaker designers.
In order to understand why some driver configurations work better than others in certain situations, it’s important to understand a couple of general principles:
1) All other things being equal, the larger the diameter of a driver, the lower the frequencies it can play, and the more difficulty it’ll have reproducing high frequencies. Again, that’s a general principle, true only when all other things are equal.
2) The dispersion of a driver (the evenness with which it disperses sound in all directions) is mostly a function of its diameter–or, in the case of oval-shaped or rectangular drivers, their horizontal and vertical dimensions. A good rule of thumb is that a driver’s dispersion starts to narrow (or “beam”) at the frequency whose wavelength corresponds with the driver’s dimension. To calculate this, divide 13,512 (the speed of sound in inches at sea level) by the effective radiating area of the driver. For example, a 6.5-inch woofer with an effective radiating-area diameter of five inches (measured from the peak of the surround to the opposing peak) starts to beam at about 2,702 Hz, or 13,512 divided by five.
Dispersion is important because it gives a speaker a big, open, natural sound. If a speaker’s dispersion is poor, it’ll seem more like the sound is coming from the speaker box instead of from a real singer or an instrument. Poor dispersion in the midrange can also create “cupped hands” coloration in voices, as if singers had their hands cupped around their mouths.
You might have already figured out what the biggest dilemma in speaker design is: The smaller the driver, the broader its dispersion (see the measurements chart below, click on the photo to view it in a larger window); however, the smaller the driver, the less it can handle low frequencies. A little more contemplation might lead to the realization that “simpler is better” rarely applies in speakers. The simpler the speaker, the more likely it is to have some combination of uneven frequency response, poor dispersion, a lack of bass and/or treble response, and/or high distortion.
Let’s consider the pros and cons of some of the most popular speaker driver configurations. One note: There are innumerable factors that influence speaker performance, including driver design and materials, enclosure design and materials, enclosure bass loading, crossover slopes, etc. However, those are outside the scope of this article, which deals specifically with driver configurations. For a more thorough understanding of the inner workings of loudspeakers, check out the Loudspeaker Design Cookbook or another in-depth reference work.
One-way (single driver, full-range)
Some audiophiles and reviewers gravitate to single-driver designs because they believe that eliminating the crossover circuit to divide the audio signal into bass and treble produces greater sonic purity. The problem is that this comes at a cost, which is often extreme. Larger full-range drivers have uneven high-frequency response and extremely poor dispersion in the treble. Smaller full-range drivers, like the 3.5-incher used in the Role Audio Sampan FTL speaker recently reviewed here, may have a fairly smooth treble response and even dispersion up to somewhere between 5 and 7.5 kHz, so they can sound good–but because those drivers’ resonant frequencies tend to be 100 Hz or higher, they will offer little or no bass response. Compared with a decent one-inch tweeter, their treble response won’t be as smooth, and their dispersion won’t be as broad.
That said, in low-cost products, full-range drivers are sometimes the best choice. I fondly remember the great-sounding, inexpensive Cambridge SoundWorks systems designed by the legendary Henry Kloss, which featured 2.5- or three-inch full-range speakers in cube-shaped enclosures, augmented by a cheap but effective bandpass bass module. I have many times advised manufacturers of low-priced soundbars to use single, full-range drivers instead of separate woofers and tweeters; at those prices, it’s almost impossible to find a decent-sounding tweeter and to include a suitable crossover circuit.
Speakers with small full-range drivers, like the Sampan FTL or the Kvart & Bolge Sound Sommeliers (shown here) have their charms, often delivering dead-flat response and beautifully consistent dispersion from the lower midrange through the lower treble, which translates to smooth and natural voice reproduction. However, they don’t sound as airy or spacious as a speaker with a typical one-inch tweeter, and they don’t play real loud or deliver much bass. But they can be augmented with a subwoofer.
I don’t see how speakers using larger single drivers can be considered high-fidelity products. Some audiophiles like them, but from what I can tell, their attraction is more philosophical than musical. The frequency response and dispersion anomalies these speakers introduce are easy to hear and to measure; so, in my opinion, there’s no credible way to call their sound natural or neutral. Some of these speakers add a super-tweeter for more extended treble response, but their overall frequency response will still be rough and their dispersion narrow in the lower treble and upper midrange.
There are also full-range electrostatic and magnetoplanar panel speakers, such as those made by MartinLogan and Magnepan, respectively. These are a completely different matter because they radiate sound backward and forward, which helps counteract the dispersion problems caused by the large radiating surfaces of the panels. Many of these speakers use curved panels to broaden their dispersion. They can’t deliver deep bass at high volumes or the kind of focused imaging that a good dynamic (i.e., cones’n’domes) speaker can produce, but still many audiophiles and reviewers consider these speakers to be among the best available at any price.
Speakers combining a single woofer and a single tweeter are the most common, and some audiophiles and reviewers consider them the best, often citing the “simpler is better” maxim. Once again, that notion doesn’t apply. Yes, two-way speakers are typically simpler than three-way models, but that simplicity necessitates a compromise that tends to occur at frequencies between two and four kHz, where the ear is most sensitive.
The problem with two-way speakers occurs in the crossover point, the frequency at which the sound is handed off from the woofer to the tweeter. As we discussed above, the larger the woofer, the more its dispersion starts to narrow at higher frequencies. So as you go from the midrange to the treble, the dispersion narrows as it nears the crossover point, then opens wide again as the sound transitions to the tweeter, which is typically one or 0.75 inches in diameter and thus disperses sound broadly well up into the top octave of treble (between 10 and 20 kHz).
One solution is to use a smaller woofer, in which case you sacrifice bass response. Or you can move the crossover point lower, so the woofer isn’t active at high frequencies. But then you start to place excessive stress on the tweeter, which probably doesn’t have enough radiating area or excursion (front-to-back motion) to reproduce those lower frequencies.
Obviously, many speaker designers have made this compromise successfully because there are innumerable two-way speakers that sound great. With a few exceptions, I prefer two-way speakers with one-inch tweeters and woofers no larger than 5.25 inches, with crossover points down around 2.2 kHz or so. Of course, any two-way speaker can be augmented with a subwoofer.
There are two exceptions here, though. First is that some high-frequency drivers have enough response down into the midrange that they can be safely crossed over to a woofer at a lower frequency, typically between 800 Hz and 1.5 kHz. These include horn tweeters using compression drivers (not dome tweeters with a horn-shaped waveguide), some ribbon tweeters, and electrostatic and magnetoplanar panels. That’s why two-way design works well in, for example, horn speakers from JBL and Klipsch, as well as two-way electrostatic speakers from MartinLogan.
The other exception is that large woofers with low moving mass and high sensitivity can produce natural-sounding midrange and can be crossed over at relatively high frequencies. Often, the woofer will have a pleated surround (the part that connects the cone to the basket) instead of the more common half-roll surround. With one of these, you can get full-range sound from a two-way. Examples include the JBL M2 professional monitors (shown here) and the DeVore Fidelity Orangutan speakers.
Most of the speaker designers I know consider a three-way design to be the best all-around choice. A three-way design gives you the advantage of the small, full-range speakers described above: dead-flat frequency response and consistent dispersion through most of the vocal range. That’s because the crossover points are usually around 300 to 600 Hz between the woofer and midrange and 2.8 to four kHz between the midrange and tweeter. You get the broad dispersion of a one- or 0.75-inch tweeter with no concerns about tweeter distortion or failure. The designer is also free to use a larger woofer (or two or three) to get deeper bass response. A well-designed three-way speaker can deliver high power handling, extremely natural sound with no major tonal balance anomalies, and measured performance that’s close to perfect.
That’s not to say three-way speakers are perfect in every way, though. Speakers that use a midrange driver that’s the same size or just slightly smaller than the woofer are common; they will play louder than speakers with smaller midrange drivers, but they typically don’t deliver that consistently broad midrange dispersion. Also, I’ve heard some three-way speakers where the midrange driver was crossed over to a large woofer at too high a frequency, which makes voices sound unnaturally bloated because they’re coming from a big woofer.
It’s also tough to make a three-way speaker at low cost. The designer has to add the midrange driver, a separate internal enclosure for the midrange driver, a larger enclosure overall, and typically two or three extra inductors, two or three extra capacitors, and an extra resistor. Multiply the cost of these extra parts by five or six times to get the increase in retail price, then add a little more for costlier shipping, and you start to understand why there are few three-way speakers priced below $400 per pair.
A common variant on the three-way speaker is the two-and-a-half-way speaker. This usually employs one tweeter and two or three matching woofers. The woofer closest to the tweeter is crossed over to the tweeter as normal. The woofers below it are low-pass-filtered, typically at 300 to 800 Hz, so they augment the bass output of the top woofer but mostly stay out of the midrange. If you ran all the woofers in parallel, crossing them all over to the tweeter, they would interfere with each other in the midrange, producing vertically narrow beams of sound (called “lobes”) at certain frequencies. While two-and-a-half-way designs will give you more bass than comparable two-way models, they still present the same challenges as two-way speakers when it comes to selecting the right crossover point between the woofer and tweeter.
Four and More
Once you get into higher-priced models, four- and five-way speakers become common. For the most part, these are basically three-way models with a large woofer added for deeper bass. They have all the advantages of three-way models with even more bass, but at higher cost. A four-way design is especially beneficial for speakers with first-order crossovers because of the strains they place on their drivers.
I’ve just scratched the surface of this topic here, and I encourage you to learn more about it by reading the LDC or another good reference book on speakers.
• First-Order Crossovers: Panacea or Problem? at HomeTheaterReview.com.
• The Pros and Cons of Multiple Subwoofers at HomeTheaterReview.com.
• How to Choose a Subwoofer for Surround Sound or Stereo at HomeTheaterReview.com.