How to Choose a Loudspeaker
So you're ready to purchase a loudspeaker. You've done some research, read a few magazine reviews, and now perhaps accompanied by a significant other, you are ready to visit a dealer's showroom. Before you pull out your wallet, you should definitely read this article. The intent here is to elucidate the pitfalls and conceptual traps that can make the process of choosing a speaker pretty frustrating. The knowledge and confidence gained thereby should allow you to cut through the fancy ads, the conflicting opinions, and the occasional pile of bullshit.
1. The Contestants:
There are more competing brands out there than you can shake a stick at. For reasons not entirely clear to me, far too many folks get the itch to design and market loudspeakers. Almost anybody can hang up a shingle pronouncing himself "loudspeaker designer." There's no board of examiners or certification process to complete. Granted, speaker design, isn't as hard as brain surgery, but even my Aunt could pass her self off as a designer. I can just imagine her squeaking: "A couple of drivers in a box, what's the big deal?"
In the past 15 years of audio writing and reviewing (first for Stereophile and now for Fi Magazine), I've come across some pretty bizzare and (let's be perfectly honest about it) incompetent designs. There are designers who have no clue as to what the basic design criteria should be and others who even lack basic design tools such as a microphone. Most of the crap originates from small companies, whose products are poorly distributed. But before you shun the "Davids" in favor of "Goliaths," you should know that many significant products and ideas do originate from small companies. For example, see Dan Sweeney's review of the Clef loudspeaker from Speaker Arts ($1,200/pr) in the September issue of Fi Magazine. The well known brand names often play it safe because stockholders demand growing profits, which discourages risk taking. The end result is typically mediocrity.
There's an implicit risk in dealing with products from small, start-up companies, which is often not discussed or ignored in the high-end arena. These companies may not be around five years from the date of purchase when your gear may need servicing. That five-year warranty you were counting on may not be worth the paper it's printed on. It's a risk I'd be willing to take for the sake of years of sonic happiness. But then, someone else might not. Again, I say this not in an overt attempt to steer you clear of the little guys, but to make you aware up front of the risk involved. Ultimately, you need to be comfortable with your purchase.
Scientism, the world's dominant religion, prescribes that an audio component's ability to give you pleasure is completely prescribed by a set of measurements that aim to document certain aspects of its performance. It is assumed that if the test measurements reach a certain level of perfection, then the listener should have nothing to complain about. The listener's perception of the soundfield created by the loudspeaker is not taken into account. It was J. Gordon Holt who first raised a subjectivist red flag in 1962 when he asserted that there was often no correlation between measurements and listening impressions. Components that measured "alike" would frequently sound different. And some components that didn't measure as well on the test bench actually sounded better than those that appeared - on paper at least - to be closer to perfection. One basic problem with test meaurements is that they don't consider the annoyance or irritant effect of various speaker distortions and resonances. A microphone is dumb. It has no feelings and can't tell us if enjoys or is moved by the music. Of course, specifications have their place. They can be used to cull out obvious dogs. But after a certain point, we have to put our ears into the loop. Audio is about listening enjoyment; you need to decide if a particular speaker moves you. Ask yourself whether this is a product that could push your emotional buttons for a long time to come. This is the same sort of strategy that a musician would employ in the purchase of a musical instrument. No one would seriously consider buying a Yamaha violin over a Stradivari because the former measured better.
I would dearly love to see all loudspeaker manufacturers include a frequency response curve with their products. Some already do, but most ship expensive loudspeakers in the kilobuck range without any indication that the product underwent any quality control at the factory. As a consumer, I would like to be assured that my prospective purchase has been tested and was found to meet the benchmark within a certain tolerance. A $20 toaster comes with more QA documentation than most loudspeakers. Consumers Union has used a simple figure of merit in their evaluation of mass-market loudspeakers: the flattest on-axis frequecny response at a single measurement point. This is certainly an important criterion. I can guarantee you that on-axis colorations are generally audible. But competing designs that measure alike on-axis can still sound totally different. We forget that a soundfield is a 3-D entity - a living and breathing energy field. We place a our mike at a single position in space and pretend that this one response measurement characterizes the output of the speaker. One needs to know something about the off-axis response of the speaker, that is the power radiated into the room, as well as the size of its sweet spot on-axis to understand how it might sound in a real room. Don't be fooled by a single frequency sweep. One can almost always find a single point in the near field of the speaker (say at 1 meter) when the frequdency response looks pretty good. What you don't realize is that if that mike were moved just one inch, all hell would have broken loose. And ironically, despite all the talk about accuracy, audiophiles seem to prize euphonic colorations. A 3 dB broad peak in the upper mids or in the presence region tends to project human voice into your lap. Speakers with such increased presence tend to assume cult status. And frankly, I too would prefer a speaker with a hyped up balance to one that sucked all the life out of the midrange.
If you already own a power amp in the 50 watt category, look for loudspeaker with a sensitivity rating of at least 90 dB SPL at 1 meter for a nominal watt input. Unfortunately, there is no uniformity in testing methods, and it appears that some manufacturers find ways to overstate the case. Beware of small two-ways with high ratings - the odds are that the specification is simply science fiction. Unless you already own or plan to purchase a high- current drive solid-state amp, stay away from speakers with broad impedance dips below 4 ohms in the lower octaves. Some 8-ohm nominal impedance loads can be nasty too with music signal. It has been shown that because of simultaneous resonances in the drivers and crossover excited by a complex music signal, an 8-ohm speaker can behave like a 1-ohm resistor for a short period of time.
3. Magazine Reviews:
Beyond frequency and power response lie important performance criteria that usually sway the critical listener toward a particular model. It makes for a long list but here it goes: level and type of distortion products, retrieval of ambient information, resolution of microdynamics, dynamic headroom, retrieval of frequency modulation nuances, delineation and control of transients, rhythmic drive, definition of bass lines, bass punch, smoothness and liquidity of harmonic textures, delicacy of treble transients, clarity, focus and palpability of image outlines, soundstage cohesiveness, transparency and spaciousness, timbral accuracy, and a believable tonal balance. The problem is that most speakers only do some of these things well.
Audio reviewers are tasked with describing the sound of a product using a more or less standardized subjective vocabulary. The reception a particular speaker receives and the level of enthusiasm showered upon it by the reviewer depend in great measure on his sonic priorities and musical tastes. If I'm into heavy metal, some criteria in the above list obviously become much more important than the rest. Bass punch and dynamic headroom would probably be more basic to me than smoothness and liquidity. This accounts in part for the major differences in opinion some products generate. I've known reviewers who'd defend a flawed piece of gear till death because it made their favorite albums sound so good. Put more weight on those opinions which are based on musical tastes similar to yours.
Reviewers are also guilty of generalizing on the basis of a single data point. Having dropped a loudspeaker into his existing system, and on the basis of auditioning the product with only a single amp and in a specific room, reviewer X is ready to declare it as either the world's worst or best speaker. Realize that such reviews are driven by a very narrow context, and are far from definitive. Naturally, experience plays a major role in the value judgements bestowed upon a product. Only through first hand experience with some of the world's best speakers can one accurately rank a speaker in the big scheme of things. The bottom line is that a review is an opinion, presumably more informed than most, but it should be taken as guidance and not as an edict from the heavens above. I've been told of audiophiles who have walked into a dealer clutching a particular review and ask to buy the product without even an audition. It's time for a reality check. Don't you become mesmerized by the written word. A lot of crap has been legitimized by overly enthusiastic and inexperienced reviewers. These guys can get you not only to salivate but in some extreme cases to dream about an intimate encounter with a particular component. The impression is left that the experience of listening to a bunch of hardware transcends live music. WRONG!
4. The Grand Strategy:
There comes a time when you have to stop reading about loudspeakers and start listening. Visit dealers near you and pay special attention to the system context. Bring albums you're familiar with you, and if possible, bring your power amp to the audition. I believe in a systems approach to audio rather than a reductionist method that simplifies the process to a listing of components. Recommended component listings paint a false picture of how to assemble a system. Two class A components don't necessarily make class A sound. And a speaker that worked in a reviewer's room may not work in yours.
If your room is large enough and you have the freedon to position the speakers away from the rear wall, then a planar or omni-directinal speaker might work well for you. Otherwise, a speaker with a more directional radiation pattern would be a better idea. Surprisingly, I've found electrostatic hybrids to work well in small rooms.
Narrow your focus to speakers that you can personally audition, then narrow the field to those that might work best in your room. Try to audition competing technologies such as electrostatics, ribbons, and hybrids in addition to conventional two and three-way moving-coil loudspeakers. This may take more time, but the process is educational because it gives you the chance to experience a range of engineering compromises. An informed dealer is one of your best allys in this quest. Trust in YOUR ears, and enjoy the hunt.