Floorstanding Loudspeakers

"A good big'un will always beat a good littl'un" is as true in the world of loudspeakers as it is, say, in the world of car engines. Once you've driven a car with a big, torquey, low-revving V8, small-capacity, high-revving engines, no matter how quick they can make the car go, seem buzzingly inelegant. Similarly, as I've listened to a number of tiny boxes in the last few months, I felt the need to spend some time with full-range speakers. Over the next few months, therefore, I'll be reviewing a handful of speakers with more-than-minimonitor pretensions.

The subject of this month's review, the Thiel CS2 2, was introduced in 1991 as a replacement for the Kentucky company's CS2, favorably reviewed by Anthony H. Cordesman (at that time a Stereophile contributor) in Vol.8 No.6 (1985), and by Larry Archibald in Vol.12 No.1 (January 1989). The usually fussy Sam Tellig wrote last April (Vol.15 No.4) that the CS2 2 sounded so good that he "couldn't find anything wrong with them" and rated them "a steal."

Superficially, the CS2 2 resembles the CS2 in being a floor-standing three-way design with a sloped-back front baffle and a crossover exclusively featuring first-order, 6dB/octave slopes, something that has been a feature of Thiel's "Coherent Source" loudspeakers since 1978. In the details, however, it is a completely different beast. Rather than the fabric-dome HF unit, the 2 2's tweeter is the same 1" aluminum-dome unit, custom-made for Thiel by Vifa in Denmark, as used in Thiel's flagship CS5. This has a wide roll surround and a short voice-coil in a long magnetic gap to obtain sufficient dynamic range—the unit offers up to ±1.5mm excursion—to allow the use of a first-order crossover filter. The midrange driver is a smaller-diameter 3" cone unit, handling just two musical octaves, from 800Hz to 3kHz. Because of the shallow filter slopes (see Sidebar), however, this driver must be well-behaved at least an octave to either side of its limited nominal passband.

The woofer is an 8" cone unit with a unique double diaphragm. A conventional cone with a slight flare is bonded at its center and circumference to an exterior straight-sided cone (fig.A), resulting in a diaphragm that, though it has increased moving mass, is much stiffer than usual. This results in good, resonance-free behavior up to 3kHz; as the woofer in the CS2 2 only handles the range up to 800Hz, its rollout above that frequency is clean. Much attention has also been paid to the woofer's motor system. A long coil moves in a short magnet gap to ensure consistency of drive force for coil positions well away from the center. The magnet has specially shaped pole pieces that keep the magnetic field symmetrical about the coil's center position; it also ensures a constant amount of iron within the coil regardless of its position. In addition, there are copper rings around the magnet's center pole. All these techniques make for a dramatic reduction in distortion.

Fig.A Thiel CS 2 2, profile of woofer diaphragm.

The CS2's reflex port has been replaced in the CS2 2 by a passive bass radiator. This is an elliptical 6" by 9" flat diaphragm mounted just below the woofer which acts in just the same way as the tuned slug of air in a port to extend the speaker's response downward in frequency. Unlike a port, the radiator blocks internal cabinet resonances from reaching the outside world.

As implied in the Sidebar, a first-order crossover involves rather more than a single series or shunt element in the feed to each drive-unit. The CS2 2's crossover involves 26 elements implemented with 35 parts, including custom-made polystyrene capacitors, and air-cored inductors wound with low-oxygen copper. Electrical connection is via a pair of 5-way binding posts underneath the speaker's base.

The CS2 2's cabinet is also surprisingly complex for a speaker in this price range. The 2"-thick front baffle is sloped back to align the drive-units so that they sum correctly on the listening axis. This black-painted baffle is fabricated from two layers of 1"-thick fiberboard, contoured with computer-controlled woodworking equipment to surround the midrange and woofer with a shallow flare and to present no sharp diffractive edges anywhere in the vicinity of the tweeter and midrange unit. The baffle's rounded edges are continued in the grille frame; the speaker is intended to be used with the black cloth grille in place. The cabinet is finished in real wood veneer—Thiel responsibly makes use of farmed exotic woods—and is extensively braced to minimize panel resonances. The midrange unit is loaded with its own braced cylindrical sub-enclosure.

All things considered, the CS2 2 is superbly constructed and finished. If its sound even halfway lives up to its finish, Thiel has a winner on its hands. We shall see if it does.

Sound
From the very first note of the very first record played via the Thiels, I was bowled over by the ease of the speaker's sound. Whereas I had found the earlier CS2 to have had rather an unforgiving, uptilted treble, the new speaker sounded effortlessly smooth throughout the midrange and treble. This was provided I didn't listen on my feet. Standing upright at the listening position with pink noise playing revealed the speaker to sound hollow and sucked-out on this unrealistically high axis. Slowly lowering myself revealed a comb filter coloration—the kind of phasing sound effect used on the Doobie Brothers' "Listen to the Music"—that moved down in frequency until I was just below the Thiel's midrange axis, when its effect was minimal. As the brain is excellent at detecting rates of changes in phenomena, this is actually a good way to determine the optimum axis for a speaker with first-order crossovers: play pink noise and move up or down until you no longer hear the suckout change its frequency with your position.

On that rather low optimum axis, which, with my listening chair, entailed me having to prop the front of the speaker up slightly with a German Acoustics brass cone, the speaker still sounded a little on the bright side, though not to the degree that it interfered with the music. Remember that brightness is more a function of a component's low-treble performance, not its extreme highs. In fact, all the way through my listening notes I made no critical comments whatsoever about the Thiel's top two octaves. The CS2 2's tweeter didn't call attention to itself in any way, neither adding an excess of sibilance nor sounding dark or rolled-off.

Substituting the balanced Levinson monoblocks for the unbalanced Audio Research amps made the speaker sound more bright, which limited the ultimate loudness despite the solid-state amps' greater power. The residual grain on my Dream of Gerontius recording on the second Stereophile Test CD was also accentuated a little through the Levinson, though whether this was due to the amplifier or the AudioQuest cables is hard to say. The bass also sounded less rich with the Mark Levinson No.20.5s, which was a benefit on LP-sourced material. CDs, however, sounded best with the Audio Research adding a little low-frequency flesh to their bones.

The bass of the CS2 2 was in general excellent, having good extension and reasonably good control. Kick drum on Robert Harley's drum recording on Test CD 2 was tight, with appropriate weight. There was no difficulty in hearing when bass guitarists using 5- or 6-string instruments made use of the contrabass register, Jimmy Johnson dropping an octave on the word "long" to drive the chorus along in the second cut on James Taylor's 1991 LP New Moon Shine (Columbia 46038). The pedals on Vol.4 of Jean Guillou's stroll through Bach's complete organ works (Dorian DOR-90151) also had a satisfying purr.

However, this CD did reveal the Thiel's Achilles' heel—a limited dynamic range in the bottom audio octave. Though the F-Major Toccata reproduced cleanly to levels approaching 100dB, the following fugue (which appears to have some formidable subsonic content) bottomed out the passive radiator at the same level, forcing me to jump up out of my chair and reduce the volume a couple of notches. Checking the woofer and radiator performance with both pure and 1/3-octave warble tones at about a 2W level (giving in-room spls between 84 and 92dB, depending on frequency) revealed clean, effectively distortion-free sound down to 31.5Hz. Below that frequency, however, the passive radiator huffed and puffed, even at this low-power input. Unlike the air in a port which has a large excursion ability and only limits a speaker's dynamic range when its velocity becomes so high that the wind noise swamps the music, a passive radiator can only move so far and no more.

I must stress that it was only on organ recordings with extreme recorded pedal levels that I noticed this problem. Nevertheless, it is something which prospective purchasers of the CS2 2 should check out. It may well be that the larger and more expensive CS3.6 would be better suited to some audiophiles' needs, particularly if they want to regularly play Guillou's Pictures at an Exhibition organ transcription (Dorian DOR-90117) at high levels. The mondo pedal tones in track 2 of this disc resulted in extreme woofer excursions at around 12V RMS input (nominally 36W), even though the big, higher-frequency climax at the end of the track was handled cleanly at 17V or so. To put this limitation into perspective, I was never bothered by it on orchestral music. Even my copy of the classic Solti Rheingold (Decca SET 382/4), which I was pleased to find still free from ticks and pops, sounded out without problem.

I did notice that the midrange unit underwent relatively large excursions during high-level organ pedal passages, due to the limited rollout of the first-order crossover allowing bass leakage into the unit. Nevertheless, I couldn't hear any untoward effect on the music, the ear being quite insensitive to Doppler distortion.

Moving up in frequency, and again making the proviso that the listener must sit on or just below the midrange axis, the midband was superbly clean and free from coloration. Again, there is hardly any mention of midrange problems in my listening notes. Piano, for example, which tends to throw problems at the top end of a woofer's passband into sharp relief, was very evenly reproduced throughout the upper midrange and low treble. The Thiel is one of the best speakers I have auditioned in this respect, regardless of price. There was an occasional suggestion of lower-midrange congestion, though this was very music-specific. Though the bass guitar on the Test CD 2 was very clean, for example, Richard Lehnert's voice announcing it acquired a little bit of a boxy quality. And some of the more densely scored passages in Brahms's third piano sonata on Stereophile's Intermezzo album (STPH003-1) acquired a little trace of extra mud.

When it came to imaging, the CS2 2 was a champ. Central images were tightly defined in space, without the splash to the sides at some frequencies that afflicts lesser speakers, while images could be heard well to the outside edges of the speaker positions with appropriate recordings. Depth, too, was well-defined, and good recordings were reproduced with an excellent sense of space. The choir on my Gerontius recording, made with a SoundField microphone, was placed well behind the orchestra and baritone soloist, as it should be, while the drums on Test CD 2 spatially sounded like a real set of drums. Only on the psychoacoustically treated LEDR height test tracks on the first Chesky Sampler CD (JD37) did the Thiels offer less then excellent performance, the images stubbornly refusing to move significantly above the speaker positions.

It was in the area of midrange and treble transparency that the Thiel perhaps scored its most longlasting impression. Whether it was the edits in the aforementioned Rheingold that producer John Culshaw had thought would go unnoticed; the edits in the aforementioned Intermezzo that I had thought would be imperceptible; the differences between the Super-Bit-Mapped "Sony Legacy" reissues of Miles Davis's Kind of Blue and the Brubeck Quartet's Time Out and the earlier "Columbia Jazz Masterpieces" CD releases; the disconcerting lurch in tape speed (did someone lean against the analog machine while mastering the CD?) 50 seconds into the von Suppé Poet and Peasant overture on the Alexander Gibson Concert Tour album (footnote 1) (Chesky CD62)—all stood revealed by the Thiels. Such was the overall clarity of the Thiels that I found them invaluable in auditioning all the rehearsal, session, and concert tapes we had made of pianist Robert Silverman for his next Stereophile CD. I could hear every little detail of the sound, from the difference in perspective of the piano given by the differing microphone arrangements we'd tried, to the small audience, traffic, and distant plane sounds that would perhaps go unnoticed though lesser speakers, and, most important, to the musical differences between the different performances.

Yet this revelation of tiny details in the sound wasn't achieved in the thrust-under-your nose manner of "professional" monitors (like the PAS Studio Monitor 1 reviewed by Dick Olsher elsewhere in this issue), nor was it in the over-etched, "ruthlessly revealing" manner of, say, the original MartinLogan CLS or even Thiel's earlier CS2. The CS2 2 was always listenable, its clarity never achieved at the expense of the musical balance. If you think about it, this is about the highest praise a speaker can get.

Summing up
This was a hard review to write. Normally, the critic ends up with a shopping list of faults that he just has to assemble into some kind of logical order. In the case of the Thiel CS2 2, however, it seemed that the longer I listened for flaws, the less I heard—an enjoyable task for the music lover, but a frustrating one for the professional reviewer. At its price, the CS2 2 is a sonic bargain. Beautifully finished, with a sound that is smooth, tonally well-balanced, and extended at both extremes, it is musically one of the most satisfying loudspeakers I've heard. Its main drawbacks are the slight tendency to brightness, which will lead to care having to be taken in choosing a matching amplifier—the Thiels and the Audio Research Classic 120s were a marriage made in heaven—and the limited dynamic range in the low bass. Unless you mainly play organ or synthesizer music, or want your high-end speakers to double for high-level disco party use, however, this should not be a problem. The CS2 2 is a "good big'un" for less than you would pay for a good pair of minimonitors with their stands. Highly recommended.

Footnote 1: Just before the cello enters with the "I've Been Working on the Railroad" theme. I grew up on these performances, recorded by the premier English RCA team of Chuck Gerhardt and Kenneth Wilkinson, which were originally released in the early '60s as part of the Reader's Digest's Adventures in Light Classical Music boxed set—so I regard them as definitive. Buy this Chesky, as well as the others in the same set: René Leibowitz's Evening of Opera (CD61) and Portrait of France (CD57). I must respectfully disagree with J. Gordon Holt's rather fainthearted praise of the latter in this month's "Record Reviews" section, therefore. It is much better than he says, even Bolero.