Skip to main content

The Brave New World: Loudspeakers

As I noted a couple months back, we tend to devalue the more familiar and prosaic items in our technology toolkit. It seems to me that we’re particularly indifferent to loudspeakers. We fuss and worry about tiny audio differences, while failing to notice that those tiny differences are absolutely swamped by the limitations of our loudspeakers. If 0.3 dB at 15 kHz makes a big difference, why doesn’t 12 dB at 300 Hz make an intolerably huge one?

Further, we don’t seem to even notice that the quality of the loudspeakers changes dramatically if we move them even a foot or so in our room. Hell, we don’t even bother to qualify our statements about the audio detail we can hear with some sorely needed caveats about loudspeaker performance. We treat loudspeakers like black boxes that are pretty much all the same – kind of like patchcords, except that some are bigger, play louder, play lower, cost more.

Our advertisements often tell us a lot about what we really value. With that in mind, take a look at the gear ads in any pro audio or video magazine. In all the advertisements flogging gear, how often do we see good loudspeakers, well installed, as an element in the advertising message? Not often, I can tell you. I just grabbed an old TV Technology out of the pile for a random check (I happened on the Feb. 7 issue, in case you’re interested). Only two ads in the entire magazine had loudspeakers visible in the graphics (one was an ad for a large audio gear supplier, the other for a console manufacturer).

In both cases, the loudspeakers were small "near-field" monitors, installed poorly. From my experience as both an engineer and studio designer, I can tell you that the facilities pictured in both ads would probably have really poor audio playback.


My longer-term study of ads also suggests this is typical. There are ads showing a guy/gal with a guitar, a girl/guy, a computer on a desk and two tiny multimedia speakers, composing rock ‘n’ roll hits. Right! Or showing a 96-input megabuck console, with the only speakers in sight being two Yamaha NS-10s perched on the meter bridge. Oh boy! The message is clear: Loudspeakers and their installation don’t really matter.

This isn’t true, of course. In fact, loudspeakers are the single most important piece of equipment used in audio production.

There are several aspects to this, all of them hard to deal with.

First, the bandwidth and efficiency of loudspeakers are pretty well constrained by their size. We can’t make woofers large enough and tweeters small enough to really extend these limits very far. So, figure that 40 Hz to 15 kHz is a reasonable bandwidth for a moderately sized, good loudspeaker.

Second, the performance of a loudspeaker is intrinsically related to the room it is in. In fact, the room and the speaker’s position in the room may be the most significant determinants of the quality of the loudspeaker. Floyd Toole has pretty definitively shown that speaker position is more important than speaker performance to critical listeners. And those of us who try to measure these things have found that you can’t even just set up two pairs of speakers next to each other for meaningful comparison. The loudspeaker in one of the positions will almost always be favored.


Third, our loudspeakers and rooms aren’t really where playback happens. In fact, the multiple millions of loudspeakers in the multiple millions of homes and cars of our clients’ listeners and viewers are the real points of playback. Our loudspeakers/rooms are irrelevant by themselves, and only valuable insofar as they can serve to predict the essentially infinite range of playback possibilities enjoyed by our great washed and unwashed public.

Finally, what we really want is for loudspeakers to sound like other sound sources in other rooms, such as orchestras, pianos, singers and rock bands in concert halls, clubs, stadiums and, occasionally, the Mormon Tabernacle. We don’t want them to sound like loudspeakers in our room.

This is obviously an impossible situation. It is equally obvious that the only way to do anything productive is through a willing suspension of disbelief. We can’t face these problems directly and literally, because they are in fact insoluble! We have to sort of, well, ignore them. In general, that’s what we do, dithering about dither instead. It’s called rationalizing.

The sad truth is that we cannot reasonably predict in any scientific way, using our loudspeakers, how the music will sound for our end users. Further, we have trouble even getting "good monitoring" (whatever that means) in any viable production room. And we can’t get loudspeakers to sound even close to the way other musical instruments sound. Hell, we have trouble making two loudspeakers similar enough that we can’t casually tell ’em apart. Man, we are definitely whistlin’ through the graveyard here!


So, a brief reality check …

A loudspeaker is a box with several transducers mounted in it. It is supposed to generate sounds with wavelengths ranging from half an inch to 60 feet (or, wavelengths much smaller than the loudspeaker to wavelengths much larger). It is supposed to do this at levels approximating the levels of live performance, both acoustic and electric.

This is really difficult. To begin with, the radiating areas of the various transducers are too small to move enough air to generate sound pressure levels approaching live performance. Meanwhile, the varying sizes of the transducers ensure that we cannot have anything remotely resembling constant power output across the frequency spectrum.

Further, the behavior of the various transducers is sufficiently different so that loudspeaker performance at the so-called crossover frequencies, where the sonic outputs of any two adjacent transducers are equal, is wildly erratic.

All this is sufficiently intractable so that we’ve almost completely given up on a couple of other major issues, such as directivity and polar response (you mean the loudspeaker should have flat response off-axis? What have you been smoking?).

Further, our consideration of small room acoustics has devolved into a kind of acoustical denial and mysticism, where nobody can really understand the principles of control room design because almost none of it stands up to scientific scrutiny. Why, it’s psychoacoustic! Which suggests, of course, that it’s really acoustics for psychos.

What makes it worse is that this has been going on for years. Loudspeaker design hasn’t changed a lot in the past 30 years, except for some better materials, better measurement techniques and sufficient amplifier power. It is stable, mature technology (which is also a good part of why we ignore it).


All this is beginning to change. A number of major manufacturers are beginning to rethink the problem. Some fairly significant recent research, plus the arrival of Surround Sound, has forced us to think a little more carefully about what it is that we’re trying to accomplish. Interestingly, I’m involved in this effort, and actually have a real commercial interest in loudspeaker development (uh-oh – a disclaimer – more about that later).

So, over the next couple of issues, I’d like to take a close look at this mundane little black box we call the loudspeaker, plus the room it’s in, how we humans hear things, and what we can do to make the whole system work better. There are some big benefits to be had, and I think they are just around the corner!

Thanks for listening.

Dave Moulton owns two Golden Retriever woofers that have remarkable fidelity. You can complain to him about anything at his Web site,