Wilson Audio Specialties Sabrina loudspeaker
Stereophile 05/2016 - Robert Deutsch
For the original text, click HERE
I first encountered the work of Dave Wilson in the late 1970s. He was then a recording engineer responsible for some great-sounding records, including pianist Mark P. Wetch's Ragtime Razzmatazz (LP, Wilson Audio W-808), which quickly became one of my favorite system-demo records.
Then Wilson turned his attention to designing loudspeakers. His first model was the Wilson Audio Modular Monitor, reviewed for Stereophile by its then-publisher, Larry Archibald, in August 1983, who described it as "the most enjoyable speaker system I've listened to, and significantly valuable as a diagnostic tool." At $35,000/pair ($83,577 in today's dollars), the WAMM may have been the most expensive speaker then on the market.
The next speaker from Dave Wilson was the Wilson Audio Tiny Tot (WATT), later joined by its bass module, the Puppy. The WATT/Puppy combo became one of the most popular—and most widely imitated—high-end speakers ever made, and underwent a number of revisions over the years. The last version, System 8, was reviewed by Wes Phillips in our June 2007 issue. Since then, Wilson Audio has continued to produce new speaker models, concentrating on the range between the WATT/Puppy's replacement, the Sasha W/P ($33,550/pair) and the Alexandria XLF ($210,000/pair).
At roughly 40" high by 12" wide by 15" deep, the Wilson Sabrina ($15,900/pair) is about the size of the original WATT/Puppy, but in almost every other way it's different. Most obvious is that the Wilson Sabrina is a single-box speaker, not a two-box affair like the WATT/Puppy. According to Wilson Audio, the inspiration for the Sabrina was the top-of-the-line Alexandria XLF, the objective being "to take the wealth of knowledge and experience contained in the XLF and distill it down to its essence." And while the Sabrina can't be called inexpensive, it costs less than the Wilson Sophia 3 ($22,500/pair)—which was, until now, the company's entry-level floorstander. What has always kept me from considering reviewing a Wilson speaker were the prices. Even Wilson's "bookshelf" speaker, the Duette 2, costs $22,500/pair with stands. For me, $20,000 is a kind of threshold: I'd have to think long and hard before spending that much on a pair of speakers—and the same goes for reviewing them.
But what if "the Wilson sound" were available for significantly less than $20,000? That's the promise of the Sabrina—and listening to a pair at the speaker's Canadian debut convinced me that this was one Wilson speaker I needed to check out.
Description and Design
In a video on Wilson Audio's website, the Sabrina is shown next to the Alexandria XLF, highlighting the former's relatively small size. The contrast is indeed great, but that's because the Alexandria, with a height of over 70" and a per-channel weight of 655 lbs, is huge: a giant among loudspeakers. Yes, the Sabrina is much smaller, but compared to "normal" floorstanding speakers it's not that small. What helps make it look relatively small is the fact that its cabinet is slimmer at the top (about 6.5" by 6") and slopes back. But the Sabrina is exceptionally heavy for its size: each speaker weighs 93.8 lbs.
Because Wilson Audio has always believed that form should follow function, their speakers have never been the most aesthetically refined. Some have been described as looking like giant robots. Judgments of appearance are subjective, but I find the Sabrina very attractive: I think it has the most pleasing proportions of any Wilson speaker I've seen. It's available in three standard and two upgrade colors, the latter at a $1000/pair premium. (The review samples were in Titan Red, one of the upgrade colors.)
On the face of it, the Sabrina is a fairly straightforward three-way loudspeaker, with no exotic materials used in the drivers, and no powered subwoofer. As always, the success of a design depends on the implementation, and the attention paid to details—and from everything I've seen and know about Wilson Audio, attention to detail is perhaps their greatest strength. The contribution of each component is carefully evaluated, its effect on sound quality being the criterion.
The enclosures of Wilson's more expensive speakers are made of Wilson's proprietary X-Material, a phenolic composite. As a cost-saving measure, only the Sabrina's baffle and bottom panel are made of X; the rest of the cabinet is constructed of high-density fiberboard. Although HDF doesn't match the extreme nonresonant properties of X-Material, it is, by definition, more dense and hard than the commonly used medium-density fiberboard (MDF). The cabinet is assembled by hand, glued with proprietary adhesives, hand sanded, gel coated, painted with multiple layers of automotive-grade paint, then polished and buffed to the same high standard as all Wilson speakers. There are two rear-firing ports: the lower serves the woofer (which has its dedicated internal chamber), and the upper serves the midrange. The tweeter has its own rear-wave chamber, to isolate it from the other drivers' outputs. The baffle is covered with dense felt, with a cutout for each driver.
The Sabrina's three drivers are variations on the ones used in other Wilson speakers. The 1" doped-silk dome tweeter is based on the Convergent Synergy tweeter first used in the Alexandria XLF. It's crossed over to a 5.7" midrange unit with a paper-composite cone, which in turn hands off to an 8" woofer, again with paper-composite cone; the crossover frequencies are in the regions of 1.6–1.9kHz and 290–350Hz, respectively. According to Daryl Wilson, son of David and the person most responsible for the Sabrina's design, "both of these crossover points are higher than we typically cross over our drivers, but work beautifully with the driver complement/configuration in the Sabrina." Each crossover is built individually to match the reference crossover within ±0.2%. The design of the Sabrina crossover uses proprietary measures to reduce distortion, with the primary design goals being bass performance and dynamic contrast. The crossover topologies and specific component choices were made with a combination of computer modeling, acoustical analysis, and listening to recordings of live, unamplified music.
Like all Wilson Audio speakers, the Sabrina does not accommodate biwiring—which, in Wilson Audio's view, would degrade the sound quality: each enclosure has only one pair of terminals.
Wilson's larger, heavier, more expensive speakers are shipped in wooden crates; the Sabrina comes in just under UPS's maximum shipping weight, so Wilson doesn't have to use a more expensive shipping company and are thus able to pass on this saving to the consumer. My review samples arrived on a shipping pallet, packed in heavy-duty cardboard boxes. The speaker is covered with a layer of protective film that must be removed—a tricky process that involves pulling the film gently downward and outward, large sections at a time. The user is warned that if the film is removed at any temperature other than "room temperature," or if the film is torn too aggressively, or without sufficient care near the edges, the paint can be damaged. Also, the protective film should not be left on for an extended period of time, and should not be exposed to a heat source or direct sunlight.
The Sabrina comes with a deluxe three-ring binder that includes detailed unpacking and setup instructions, but Wilson Audio strongly advises that you leave these tasks to an authorized dealer "trained in the art and science of the Wilson Audio Setup Procedure." In my case, this involved the Toronto area's Wilson dealer, Audio Excellence, and the setup expertise of Wilson's director of sales, Peter McGrath. Given McGrath's long experience in setting up Wilson speakers, I was content to leave the Sabrinas' initial setup to him—with input from me. Like me, McGrath likes to set up speakers to produce a wide soundstage. His setup of the Sabrinas was similar to what I've found optimal with other floorstanding speakers: along the long wall of my listening room (16' long by 14' wide by 7.5' high), roughly describing an equilateral triangle with the listening position, and toed-in to almost fully face that position (see photo of my room). When I'm seated, my ears are about 37" from the floor, which is within about half an inch of the level of the tweeters. (Given the Sabrina's backtilt, determining the exact height of the tweeter involves some eyeballing.) The speakers were a little farther apart than I place my Avantgarde Uno Nanos, and closer to the front and side walls. The Sabrina is supplied with heavy-duty spikes, as well as aluminum discs to be placed under the spikes to prevent damage to hardwood floors such as mine. I used the discs.
The positions of the driver modules of Wilson's larger speakers can be adjusted, to time-align the drivers' outputs for arrival at specific listening distances. For the Sabrina, no such adjustment is possible, but the driver array and the sloping baffle have been optimized for what Wilson calls a "typical listening room," and the tilt of the speaker can be adjusted by varying the length of the front and/or rear spikes. The speakers were initially set up with a slight additional backtilt, which is recommended in small rooms such as mine, to produce at the listening seat a soundstage of sufficient height. I later tried decreasing the tilt, and preferred the result: the focus was improved, and the soundstage was still high. Of course, rooms and listening positions vary; what worked for me can't be assumed to work elsewhere. While I wouldn't describe the Sabrina as exceptionally "tweaky," it definitely benefits from careful setup.
Except for grilles that have been engineered to be part of a speaker's overall acoustical design, I've yet to encounter a grille that does not degrade its speaker's sound in some way—and this was true of the Sabrina's. The sound was just a bit less open with the grilles in place, though the effect was quite subtle—and certainly not as marked as I've heard with some other grilles. I did all of my critical listening with the grilles removed.
Whether or not audio equipment needs to be broken-in is a contentious issue: a 2005 discussion of the topic at Stereophile.com had over 60 responses. The Sabrina's owner's manual states that Wilson Audio subjects all midrange drivers and woofers to a 12-hour break-in before the drivers are tested, calibrated, and matched. They acknowledge the benefit of additional break-in, and suggest that it will be 90% complete after 24 hours of playing. But being an audiophile means being concerned about that last 10%. On the subject of break-in, I tend to be on the "more is better" side, and use various break-in/system enhancer CDs (from Monitor Audio, Nordost, and Purist Audio), as well as music played at fairly high levels. I do believe that the Sabrinas sounded better with additional break-in, the sound becoming generally more relaxed, and losing the slight edginess that I heard from the speakers at first, just out of the box.
The first power amplifier I used with the Sabrinas was a McIntosh Laboratory MC275LE, which I reviewed in a Follow-Up in the October 2012 issue. This tube amp has a conservatively specified power output of 75Wpc, which seemed a good match for the Sabrina's recommended specified minimum amplification of 50Wpc and sensitivity of 87dB. However, I was a bit concerned about the fact that, also according to Wilson's specs, the Sabrina's impedance drops to 2.53 ohms at 139Hz. The MC275LE has separate taps for 4, 8, and 16 ohms; the Sabrina's nominal impedance is 4 ohms, so at first I used the Mac's 4 ohm taps. This seemed to work well, the amp driving the speakers at normal as well as very loud, "audiophile demonstration" levels, and the bass seemed fine. However, during setup, Peter McGrath questioned the use of the 4 ohm taps. He said that, in his experience, Audio Research tube amps worked better with the Sabrina through the ARCs' 8 ohm taps, and he wondered if that might be the case with the Mac. We tried it, switching between the two sets of taps, and we both preferred the sound through the 8 ohm taps, which I used for the rest of my listening. (In a similar comparison for my September 2012 review of the MartinLogan Montis, I'd also ended up preferring the Mac's 8 ohm taps.)
Although the tubed McIntosh worked well with the Sabrinas, it's probably atypical of the amplifiers that most people would use with these speakers, and probably not optimal for extract low bass from them. I had two other amps on hand, but both were also tubed, and less powerful than the MC275LE. What I needed was a solid-state amp with more power than the MC275LE. As per Stereophile editorial policy, the amp would have to be one that has been reviewed in the magazine—a known quantity. Looking through the list of potentially suitable amps, the one that struck me as a good bet was the Theta Digital Prometheus, reviewed in the March 2015 issue by Larry Greenhill, who praised it for its "huge dynamic range and bass impact." He said the Prometheus was one of the best-sounding amplifiers he'd heard in his listening room, and it was his Editor's Choice for 2015. The claimed output of the Prometheus is 250Wpc into 8 ohms, 500Wpc into 4 ohms, or 850Wpc into 2 ohms—I thought it would be well able to take the Sabrina's impedance dip in stride. Jeff Hipps of Theta Digital kindly arranged to loan me a pair of Prometheus monoblocks.
There was a hiccup in pairing my Convergent Audio Technology SL1 Renaissance Black Path Edition preamp with the Prometheus. The CAT's standard, minute-long automatically muted turn-on period was apparently not long enough for the Prometheus; when the SL1 unmuted itself, I heard a soft, low-frequency "motorboating" sound. All I had to do was to keep the CAT muted for an addition minute after turn-on, and everything was hunky-dory. The combo was then dead quiet, with pristine clarity, and no digitalesque artifacts that I could hear. (The Prometheus uses Bruno Putzeys's NCore NC 1200 class-D amp module, with a linear power supply created by Theta's David Reich.) Just as LG had said, the Theta had great dynamics and bass impact, both of those qualities being superior to the Mac's (although I preferred the MC275LE's tonality). My descriptions of the sound of the Sabrinas represent a kind of averaging of the sound with the two amplifiers, with amplifier-specific distinctions as noted.
Dave Wilson's stated aim in designing the original WATT/Puppy was to produce "a compact and simple loudspeaker that could provide a degree of musical enjoyment rivaling—and even surpassing—much larger systems." Listening to the Sabrinas—even before the break-in period had passed—I was immediately struck by how "big" they sounded—or, in terms that more clearly reflect Dave Wilson's aim, how much they sounded like live music rather than speakers. And their resolution was first-rate: Playing familiar records often resulted in responses of "Hey, I haven't noticed that before."
Having recently upgraded my phono system, I've been playing LPs more frequently. As a kind of tribute, one of the first recordings I played through the Sabrinas was Concert, Dave Wilson's first recording of a pipe organ (Wilson H-1-77). I started with Johann Gottfried Walter's Concerto 3, played by James Welch on a tracker-action organ. There was a great sense of space, the sound of the organ filling the venue in a manner both powerful and delicate. I think of the pipe organ as a "slow" instrument, with no real transients—compared to, say, a piano—but with this recording played through the Sabrinas, the organ sounded unusually "quick," with a lot of dynamic variation.
Further listening revealed that one of the Sabrina's strengths was its ability to faithfully reproduce dynamic variations. Piano recordings demonstrated this particularly well. Sonata, pianist Robert Silverman's recording of solo works by Liszt (CD, Stereophile STPH008-2), engineered by John Atkinson and Robert Harley, has a wide dynamic range: The sound of Silverman's piano goes from whisper quiet to blow-the-house-down loud, with everything in between. Just for the fun of it—don't try this at home—I first played the recording at what would be for me a normal volume for this sort of music, and then at a volume two clicks higher on the CAT preamp's volume control—an increase of about 3dB.
The peak volume, measured with the iPhone 6 AudioTools Analog SPL meter app (C weighting, high microphone level, low-pass filter disabled) was 95.7dB. This SPL measurement is not calibrated to professional standards, but it gives at least some measured indication of the level. To my ears, it was very loud: Other than to demonstrate a speaker's dynamic capability, I wouldn't want to listen at this level. The Sabrinas—for this test, driven by the Theta Prometheus—took it all in stride, with no audible distortion. Reproduction of dynamic contrasts is one of the strengths of horn-based speakers such as my Avantgarde Uno Nanos; the Sabrinas came closer to the Avantgardes in this respect than any other speaker I've reviewed.
And the bass? Well, that Wilson pipe-organ recording has lots of it, including several pedal low Cs (32Hz). I expected a speaker of the Sabrina's size, with just a single 8" woofer, to merely hint at these notes, or present them only as harmonics. But no, there they were: clean, and at levels that, while not quite room-shaking, were certainly more than enough to provide a solid musical foundation. The Avantgarde Uno Nano, which has a powered subwoofer section with twin 10" drivers, and the GoldenEar Technology Triton One, with its powered DSP-controlled subwoofers and passive radiators, go even lower, but the Sabrina was not far behind. The bass was tuneful, and transients, such as those in recordings of bass drums or timpani, had appropriately quick onset and very little overhang. Wilson identifies its woofer as having been first used in the Alexia and modified for use in the Sabrina. Whatever those modifications were, the effect is that the driver now provides greater extension and power-handling capability while retaining the quickness of a normal 8" driver. Quite a feat.
Furthermore, the crossover from woofer to midrange—often a difficult area—sounded seamless. This was evident in the reproduction of such instruments as cello, double bass, and bass guitar, and in recordings of male voices. I've heard some otherwise-fine speakers that made baritones sound more tenorish than I know those singers sound in real life, with less chest resonance. A recording I've played quite a bit lately—and that I chose as one of my 2016 Records to Die For—is Frank Sinatra's Ultimate Sinatra collection (4 CDs, Universal B00224360-02), which chronicles Old Blue Eyes' career. In his early days as a band singer, Sinatra's voice was decidedly lighter—he took some high notes that could have come from a tenor, but there was still a chest resonance. The Sabrina reproduced this very faithfully.
In fact, the reproduction of voices, male or female, was another of the Sabrina's strengths. As a standard for a recording of a woman's voice, I always go back to Sylvia McNair's Sure Thing: The Jerome Kern Songbook (CD, Philips 442 129-2), which also serves as a test of a speaker's reproduction of the midrange and treble. The midrange-tweeter crossover had a smoothness similar to that of the woofer-midrange blend, and the Sabrina's treble was revealing without being exaggerated, which would have shown up as "spitty" sibilants. Sibilants sounded overemphasized only when they'd been recorded that way. (The highs were sweeter with the McIntosh MC275LE than with the Theta Prometheus, but I happily listened to the Wilsons with each amplifier for long periods, without having the urge to switch to the other amp.)
The soundstage thrown by the Sabrinas was wide and deep, with images on it precisely defined. The deepest soundstage and most precisely defined images I've heard from a pair of speakers under review were with Fujitsu Ten's Eclipse TD712 Mk.2, which I reviewed in January 2007. The spatial definition produced by that single-driver speaker was uncanny—but the Eclipse couldn't match the Sabrina (or any number of high-quality, multi-driver speakers) in maximum loudness capability and bass extension. Compared to other multi-driver speakers, the Sabrinas' imaging was excellent. When I played the "Depth of Image" tracks from The Best of Chesky Jazz and More Audiophile Tests, Vol.2 (CD, Chesky JD68), the Sabrina's reproduction of them was comparable to that of the GoldenEar Triton One—and the GoldenEars did better on this test than any other multi-driver speaker I've heard in my listening room.
Colorations from the Sabrina's cabinet? There weren't any—at least, not that I could hear. John Atkinson may be able to pick up something with his accelerometer, but the Sabrina had the most sonically inert cabinet of any floorstanding speaker of my experience, surpassing even the otherwise admirable—and, of course, much cheaper—GoldenEar Triton One.
My dedicated listening room began life as one of our house's three upstairs bedrooms. When reviewing a speaker, I go to considerable lengths to set them up just right, but much of my listening is not actually done in that room. Instead, it involves what's been called the Listening In Another Room (LIAR) test. Right now, having listened to the first track of Sure Thing in the listening room, I'm sitting on a loveseat in our living room, down on the main floor, my laptop on a small table in front of me as I work on this review, keeping an ear out for Sylvia McNair. She seems to be upstairs, singing. The extent to which this illusion is maintained is, for me, one of the tests of speaker performance. I wouldn't want to judge a speaker solely on the basis of how it sounds outside the listening room, but the LIAR test does have the advantage of eliminating the effects of minor variations in the positions of speakers and listener. The Avantgarde Uno Nano, whatever its shortcomings, performs very well on the LIAR test, and the Sabrina was very much in the same class. A variation on the basic LIAR test is to consider whether listening downstairs makes me want to go upstairs. And now, if you'll excuse me . . .
Who will buy?
From the start, beginning with the WAMM and the WATT/Puppy, Wilson Audio Specialties has been driven more by technology than by marketing—with technology applied in the service of music. One assumes their speakers have never been designed to hit specific price points, but to reach certain levels of performance, upon which they're priced according to their design and manufacturing costs, with an eye to economic survival and continued technological development. Wilson's expectation was apparently that consumers would appreciate what the speakers had to offer, and would be willing to pay the price—and in that Wilson has been extraordinarily successful.
Did the development of the Sabrina mark a departure? Did they set out to design a full-range speaker that has the essence of the Wilson sound, at a significantly lower price? If so, Wilson has succeeded spectacularly in producing a speaker that, while of relatively modest size, indeed has all the hallmarks of the Wilson sound: wide frequency range, high resolution, real bass, a striking absence of cabinet colorations, and a particular adeptness in communicating the dynamic contrasts of live music.
Who will buy the Sabrina? Adrian Low, the owner of Audio Excellence, told me that a number of his customers who own larger Wilson speakers have moved to smaller homes and found that their Wilsons are now too big for them. Downsizing to the much smaller Sabrina would involve few or no compromises in sound quality. Another group of potential buyers are audiophiles who have long admired Wilson speakers but never been able to afford them. Finally, there are those who didn't consider themselves to be in the speaker market at all but who, on hearing the Sabrinas, said, "I've got to have these speakers!"
I'm inclining toward membership in that last group myself.
Wilson Audio Sabrina specifications
Wilson Audio Sabrina associated equipment
Wilson Audio Sabrina measurements