What can active speakers do that passive ones can’t? Impulse response, impulse response, impulse response, says Bruno Putzeys. That, and their sheer practicality, makes them the wave of the future.

This article originally contained more technical information. To read the unabridged version, click here.

“A highly evolved implementation of an established technology nearly always outperforms the first example of a new and superior one,” reads a famous engineering maxim. This explains a lot about the high-end audio psyche. Whenever a new audio technology comes out, often audiophiles will quickly dismiss it, saying “this won’t catch on,” and stick with what they know.

That’s why the average hi-fi show looks more like a museum of audio development than a showcase of new-tech. People are still debating whether solid state amplifiers can match the quality of tube amps, whether push-pull designs will ever work as well as single-ended ones, whether pentodes (including beam tetrodes, since this is a North American publication) can compete with triodes, and whether indirectly heated cathodes could ever be a substitute for directly heated ones. And that’s just the amplifier side of things. Wax cylinders, anyone?

This article is a reply to PMA Magazine Editor Robert Schryer’s question “Do active speakers have the potential to sound better than passive ones?” I worry that editors are getting wise as to what buttons to push to get an article out of me. (Oh, we know which ones.- ed.)

An Uneventful Start

The definition of an active speaker is quite simple: a multiway loudspeaker with a dedicated amplifier per “way”, with any filtering done before the amplifiers:

Historically, that was all: amplifiers and filters traded places in the signal chain, but the drivers got the exact same signals. I have no idea what kind of sound improvement people expected from this. I know of several respected designers who did the experiment and concluded that active is for the birds. Indeed, there is no point going active unless you intend to use the concept as a jumping-off point for something that isn’t possible using passive crossovers.

Economics

Probably the best kept secret in high-end audio is that amplifier circuits are cheap. Whatever your favorite amplifier topology, the circuit itself constitutes a fraction of the cost of the completed product. The chassis and power supply make up the rest.

Standalone amplifiers must comply with silly FTC rules requiring full-power sinewave operation for five minute—something that would likely destroy most speakers in seconds. This forces manufacturers to use oversized power supplies and, for class A/B designs, larger heatsinks. But if you design the amplifier and speaker so their continuous power capabilities are a match, you can save a lot of money.

Adding separate mid and tweeter channels doesn’t change the total output power, so doing so has no impact on the power supply. The differential cost of an extra channel is just a handful of parts. On the other hand, good, low-distortion passive filter coils are big clumps of copper. The raw metal alone costs more than adding an extra amplifier channel. Finally, hiding the amplifier in the speaker can save a fortune in chassis manufacturing costs.

So, from a cost perspective, active speakers have always made sense. Even if you just want to match a speaker and amplifier, doing so “actively” will be cheaper and more practical. That’s why active speakers are everywhere in recording studios—engineers need reliable tools, not a bunch of gear to fiddle with.

By contrast, the whole consumer market is built around exchangeable components that buyers are encouraged to constantly “upgrade.” You’re not sold a sound system so much as a “journey” (i.e., taken for a ride).

What Changed?

Until recently, that was how things stood. Active speakers were built like passive ones, with the electronics thrown in at the last moment. Then about ten years ago, a new generation of active speakers arrived that were conceived as active from the get-go, and that confounded expectations of what a speaker, and that confounded expectations of what a speaker was capable of—expectations shaped by the limits of passive design.

Take bass performance, for example. It’s commonly assumed that a speaker needs to be big to produce deep, powerful bass. That’s just physics, right? Not so fast.

The problem with passive speakers is that they are a tangle of intimately interacting electrical, mechanical, and acoustic elements. Just designing a bass cabinet with a flat response quickly degenerates into endless cycles of trial and error where each parameter is tweaked over and over until you find a driver, cabinet and filter that work together.

Even the design options are limited. Sealed cabinets, for instance, offer the tightest bass, but they’re as rare as hen’s teeth. Why? Because any half-way efficient woofer has such a strong magnet that the braking force of the voice coil overdampens the system. To get a nicely defined, flat response from a sealed box, you need to make the woofer less efficient—either by adding series resistance (unappealing) or using a weaker magnet (even worse for reasons we won’t go into here).

Hence, the preeminence of ported cabinets. These do work with strong woofers thanks to a second, underdamped resonance that flattens the frequency response—though at the cost of a much steeper roll-off (4th order vs. 2nd for sealed). The result? A woollier, more drawn-out transient response. The second drawback is that below the corner frequency there is no air pressure to hold the woofer back, making ported cabinets rather easy to overdrive. It’s a situation we tolerate because hardly anything else works. Ported cabinets have become so much second nature that they are rarely questioned. It’s how one does things. So, instead of rethinking the fundamentals, designers spend their time creating slight variations like transmission lines or passive radiators. And when they succeed, they’re so impressed with themselves that they fail to realize that it hasn’t really moved the needle in sound quality.

Breaking Out

An immediate advantage of active speaker design is that it decouples the electrical domain from the mechanical and acoustic ones. Furthermore, you can easily undo the effects of the remaining coupling between the latter two domains.

It is simpler than you might imagine: pick a woofer, mount it in a sealed cabinet, then measure or calculate its response. From there, take the response you want and design a filter around it that makes up the difference.

This type of equalizer is known as a Linkwitz transformer. It’s a second-order shelving filter that works by combining two elements: an anti-resonance tuned to cancel out the speaker’s natural low-frequency resonance, and a new resonance with the exact corner frequency and Q factor you want. The result? Precisely tuned bass—every time.

This represents the first fundamental sonic benefit of active speakers. The ability to achieve a perfectly tuned second-order bass response from any sealed cabinet guarantees a tighter, more articulate bass than you’ll ever get from a ported cabinet.

And the cabinet can be of almost any size. The caveat, of course, is physics: getting the same low-frequency performance from a box half the size requires four times the amplifier power—and the woofer’s voice coil must be able to dissipate that extra heat. Both amp and driver need to be up to the task.

But since we’re already in the realm of active systems, there’s an opportunity to go further. We can vary the corner frequency dynamically to limit peak cone movement. We let the bass go deep most of the time and gently adjust the corner frequency just for the duration of an excessive bass transient that is about to overload the woofer. This matters because when a woofer bottoms out, it sounds awful. Ironically, it’s these rare, extreme transients—not typical listening levels—that tend to dictate how big the driver needs to be, irrespective of how little it moves during normal use.

Admittedly, it’s a tricky assignment. But the few products that do get it right put on a spectacular show for their size, going deep while sounding perfectly clean when pushed hard.

So why all this focus on size when the subject of this article is sound quality? Simple: most listening rooms place a limit on the size of the speakers. It’s not about making big speakers smaller, but about making smaller speakers play big.

That’s exactly what today’s “second wave” active loudspeakers do. These relatively compact, sealed systems easily outdo much bigger traditional speakers in the bass department.

Crossover Filters

A curious quirk of the hi-fi market is that you will find one-way, two-way, three-way speakers etc., all inhabiting the same lofty price segment. Clearly, not everyone who buys a small 2-way speaker does so because they can’t afford something bigger. On the face of it, this makes no sense. Splitting the signal into more bands removes intermodulation between signals occupying different bands. Each driver is optimized to work cleanly in a narrow range. Multiway speakers should win hands down on all counts. Why should anyone cling to two- or even one-way speakers then?

Or take recording studios. The big monitors are used only to turn it up loud when the band comes to listen, but any actual work is done on speakers like 5″ Auratones. Mixing engineers say they just can’t hear what they’re doing on the big speakers.

Something is going on here, and it’s almost certainly the little matter of transient response. Here are some calculated step responses of ideal passive one-way and multiway speakers:

Only the one-way configuration has an unblemished step response; it makes a clean jump and then settles back. As we split the spectrum into more bands, the transient gets frayed in time more and more dramatically. Each crossover point adds another downswing to the step response and delays the bass. Technically speaking, passive multiway speakers are all-pass filters, meaning all frequencies come through equally strongly, but with an added, frequency dependent delay.

Those are musically relevant time shifts. If you think flamenco players just strum their guitars, that’s because a passive three-way speaker makes it sound like that. An electrostatic speaker will reveal a precisely timed sequence of notes instead. That’s not because electrostats are “fast” and electromagnetic drivers “slow” (they’re not), but because of the crossover filter.

Indian classical music leans on extreme precision of interplay between the musicians, and so does free jazz. The latter builds up tension by everyone intentionally playing slightly off the beat and then releases it when suddenly the band pulls together in an astonishingly precisely timed climax. Listeners with big multiway speakers just sit there waiting for a recognizable melody to emerge, completely failing to notice that the whole game is about rhythm and timing.

Western classical music, on the other hand, is written and played to be heard in large, reverberant spaces, so the additional phase shift created by even very complex passive speakers is relatively innocuous. Still—the attack of a note is what gives a musical instrument its unique timbre, much more so than the spectral makeup of the sustain. In a real human voice, you hear the vocal cords flap together clearly as individual clicks. One-way speakers preserve this, giving a sense of intimacy that is lost as soon as you add even a single crossover point.

Finally—and I think this should be obvious—there’s the matter of dynamics. When an impulse is spread out over time, its peak amplitude decreases. A big five-way speaker system will happily rattle the rafters, but it won’t deliver a spontaneous punch.

This is vexing: to capture the full drama of a live performance, you need a powerful speaker—usually a multiway design. But while these speakers offer impressive power and cleanliness, they take away in timing. In the end, music lovers tend to choose the speakers that preserve the qualities they value most in their favorite music. Show me your speakers, and I’ll tell you what music you listen to.

Read part 2 here.

This article originally contained more technical information. To read the unabridged version, click here.