Thiel CS5 loudspeaker Page 2
But I just couldn't get the speaker to sing in the bass.
Nor, ultimately, in the midrange and above, despite an astonishing lack of any coloration. My reference amplification and cables are the Mark Levinson No.26/25 preamplifier and a pair of No.20.5 power amplifiers, connected with AudioQuest Lapis balanced interconnect and AudioQuest Clear speaker cables. With just about every high-quality moving-coil speaker I have used in my current room, this setup gives the best balance between midrange bloom and bass control, between treble clarity and soundstage perspective. Yet with the CS5s driven by this combination, everything seemed just too damned polite too much of the time.
I was reminded of something the Audio Anarchist had said in the August 1989 issue of Stereophile following his auditioning of the CS5s at the 1989 SCES: "The speakers sounded processed...as if there were a lot coming between the signals and the drivers." Now negative first impressions at shows aren't worth the paper they usually aren't written on, in my experience. But this is exactly what I was hearing, even Bonnie Raitt's delicious Nick of Time album sounding too restrained, too much Eisenhower-era, narrow-lapel cool and not enough 1990s baggy-shirt boogie.
Changing cables to the Straight Wire Maestro equivalents recommended by Jim Thiel brought back a significant amount of life, but I began to realize that there was a basic incompatibility between the constraints of my room and the needs of the CS5s. To persist in the review under these circumstances seemed unwise, even irresponsible. I phoned Larry Archibald: "How'd you like to take over the CS5 review?" We haggled. I agreed to loan him the magic Stax D/A processor. The deed was done.—John Atkinson
Larry Archibald Relieves
It's both an honor and a fascination to review the best work someone's ever done, no matter what the field. This is all the more true when that person has spent over four years in the development of that work. Not only is the result likely to be something special (especially when the designer really knows his stuff—and no one would deny that Jim Thiel knows his stuff), but the reviewer gains unusual knowledge of just what matters most to the designer.
The Thiel CS5 is no disappointment in this regard. It is a demanding product, one that requires the right room, a lot of patience in terms of finding the speakers' ideal location, and the utmost in associated equipment. As with the Infinity IRS Beta, you will need the help of a skilled and devoted dealer. When his and your labors are finished, however, you'll have an instrument that does quite a few things better than any of its competitors.
My first exposure to the CS5s came at Stereophile's High End Hi-Fi Show in San Mateo in April '89; it was, in fact, their first exposure to critical auditioning outside of Lexington, Kentucky. I was even able to audition them at the Show in two situations, one arranged by dealer db Audio of Berkeley, the other by Audio Excellence of San Francisco. These auditions, using Rowland and Audio Research electronics, were most impressive, but not exactly revealing. A good demonstration is, in fact, designed only to reveal a product's strong points and disguise the weak points. (Distorted, steely recordings are one of the best kinds of test since they quickly highlight a product's exacerbative tendencies in these areas, but you'll never hear them at a good demonstration.)
My next exposure came in two locations at last summer's CES, where both Madrigal and Thiel were using CS5s. The general consensus (see Vol.12 No.8) was that Madrigal's much better room at the Nikko Hotel showed the speakers off to great advantage, while Thiel's own room at the McCormick Center Hotel, unfortunately with windows occupying an entire side wall, considerably quieted auditioner enthusiasm.
Finally, after much delay, a pair was ready to come to Santa Fe. We were all quite keyed up, as both our other experience with Thiel products and the product's positive presentation in public had prepared us for great things. Setting up the speaker in JA's listening room, however, didn't meet expectations. Not only was one of the midrange dome-drivers' voice-coil rubbing (kindly replaced by Jim Thiel, on hand to make sure the speakers were, in fact, functioning properly), but the general impression was one of restraint—excessive politeness, if you will. Nothing brought this home more than a sterile, emotionless presentation of Bonnie Raitt's Nick of Time, an album with a lot of get up and go, if not admirable sonics. So disappointed was I that I immediately went home and played Bonnie on my Mirage M-1s to see if my memory was failing me—at such an early age! Fortunately, though Bonnie wasn't as good on the Mirages as she had been live last summer in Santa Fe, she could still get the blood a-stirrin'. At JA's house, the CS5s just didn't start up, on Bonnie Raitt, or, in my experience, on anything else.
My first comment, then, would be to look carefully into your heart to see if the room you intend for these $9200, 380-lb, flat-into-the-20s speakers will support their need for breathing room. If not, take a listen to them well set up at a dealer's (with some of your favorite music), then buy something else. My own room measures 20' by 35' and has an 11½' ceiling, and in terms of setup I was stretching its limits to accommodate the CS5s, though I'm sure they will work acceptably in smaller environments (16' by 24' sounds like a practical minimum). Take care, though—this is a speaker with almost-live music capabilities; you don't want to choke them down in the wrong room.
The Thiels also demanded nontypical setup within my listening room. They ended up 9½' from the 20' back wall (farther away than usual), 10' apart center-to-center (farther apart than any speakers have tolerated to date), about 5' from the side wall (closer than usual), and 9' from my listening chair (typical). I started with them a few feet closer to the back wall and closer together, but the sound didn't quite lock in. The radically wide spacing gave stunningly coherent imaging and soundstaging, well-integrated mid- and low bass (which hadn't quite made sense in their previous position), and, to my surprise, never any problem with "hole in the middle." All other speakers I've tried this far apart would deliver anomalous center imaging on at least some recordings.
The setup in your room will certainly be different; don't be afraid to try something different from what you're used to. Setup was quite sensitive to overall geometry within the room, but not to excruciating exactness of individual speaker placement—movements of a few inches one way or the other didn't make huge differences, though I was superstitiously fastidious in observing exact setup symmetry.
Amplifier Problems: Your choice of amplifier can determine whether the Thiels deliver a dramatic, almost unexceptionable, reference-class sound or one which keeps making you think that perhaps Jim Thiel should finish the last 20% of his work on the CS5 design—definitely a disappointing result with a $9200/pair speaker! What seems to be happening here is the revelation of needs the CS5s have—not defects that need to be corrected or mollified (as you might interpret is the case with the IRS Betas), but potentialities awaiting just the right driving force for them to become actualized. This is most true at the low end. The Thiels will deliver awesome amounts of sound from the midbass on up, and prodigious levels of low bass. (In case this distinction confuses you, "awesome" is greater than "prodigious"; the very low bass capabilities of the CS5 are one of its limitations, though not one that ordinary program material reveals.) The mid- and upper bass of the CS5 can, however, sound furry, muffled, and a bit thick with the wrong amp—which is still an amp that sounds great in other systems.
I will have to leave it to engineers to account for this low-end amplifier sensitivity. JA's measurements reveal the CS5 to be a simple load from an impedance standpoint, but a cruel one. My hypothesis is that the rubber "pads" that are stuck to the cones of the two lowest-frequency drivers, presumably to control resonances, load down those drivers so that they just require one helluva lot of current to make them work under complete control. In addition, nothing in the crossover rolls off the speaker at the low end, which leads not uncommonly to up-to-1" excursions of the woofers as they try to reproduce the few hertz of record warps.
The upper-frequency sensitivity is one that seems most related simply to tonal balance (and could, probably, be accommodated simply with gentle use of an excellent equalizer). J. Gordon Holt has always been one to point out the mildly laid-back character of solid-state amplifiers. As often as I've listened with him, this character has never been particularly evident to me, swamped as it was by the much-more-irregular sound of whatever speaker it was we were listening to. As usual, JGH was right—it just took a speaker as flat as the Thiel for me to hear it. (JGH is able to hear these kind of mild tonal aberrations even through the colorations of his favorite horn speakers!)
Given that the CS5s aren't bi-ampable—so you don't have the option of solid-state on the bottom and tubes on the top—I have to recommend this as a speaker which needs solid-state amplification, but it should definitely be solid-state of the forward and lively character; in my experience, the Krells gave the best sound I heard from the speakers.
Spectral Balance: I forego the traditional review category of "The Sound" in favor of "Spectral Balance," "Dynamics," "Imaging and Soundstaging," and "Resolution" simply because, longwindedly, I have something particular to say in each of these areas.
It will come as no surprise, to those who have followed Thiel products over the years, that flat frequency response was a paramount design desideratum for the CS5. Jim Thiel has ever-more-resolutely pursued flat tonal balance, even when the flaws of systems (and of the drivers he had available) led to criticism in the press and commercial limitation in the field. Basically, a significant group of critics and consumers heard Thiels as "bright," which was often inaccurate since it was the speakers' response in the range from 8kHz–20kHz which was more extended than the competition—far above the brightness region.
The CS5 is the first Thiel which, to my ears, completely sidesteps these criticisms, while retaining extraordinarily flat perceived and measured frequency response. Just yesterday I was discussing with JA the various soundstages available from the CS5, through different speaker positioning (see discussion below), and commented that I couldn't really hear any tonal shifts while varying speaker angle up to 20°. No wonder! The frequency responses between on-axis and 15° off-axis only vary significantly (ie, by more than 1dB) above 16kHz (not a region where my still-youthful ears hear a thing).
Playing music, you will find the CS5s telling you new things about almost every record you own in terms of the precise character of voices and instruments. This is a difficult characteristic to quantify, or even to describe. It would appear that, when every frequency is handled with precise evenhandedness (from an amplitude standpoint), their relationships lock together to create a much more realistic reproduction.
Equally important is the smoothness and cleanness of this frequency response. Not only are no regions unnecessarily emphasized, there are almost no peaks or resonances to stand out. You will note that I've so far disdained the standard audiophile adjectives: warm, boomy, midrangey, bright, tipped-up, dark, lively, sweet, or dry. The CS5 is the first speaker I've heard whose apparent frequency response approaches that of an amplifier, and, in fact, my discussion under associated equipment of the speaker's amplifier sensitivity tells you that its tonal balance is perhaps flatter than that of some amplifiers (particularly into this load).
I am, in fact, at a loss to come up with an adjective to describe the CS5, other than "informative." Not only are the changes introduced by amplifiers substantial compared to the contribution of the speaker; the recording-to-recording variations are mammoth. Thiel's CS5 Owner's Manual contains the following introduction to the speaker: "The CS5 is a precision instrument designed to translate, as accurately as possible, electronic information into musical sound." I would say that Jim Thiel is justified in his presumption; the CS5 can most closely be characterized as a large-scale musical viewing lens for hearing what you have on record.
At the conclusion of such glowing praise I must insert some qualifications. The Thiel is as wide-range in frequency response as I've had in my listening room, matching both the IRS Beta and the Waveform in its ability to reach down into the 20Hz region and up beyond the range of audibility (the Waveform is, actually, slightly more extended at the bottom, but below-20 is not a region of great interest to me).
In spite of the fact that a substantial part of the CS5's cost is devoted to the range below 100Hz, this area remains its weakest—though to a degree that I might not even make this criticism were the upper frequencies less good. Low frequencies are produced with great power down to the midbass, but with less precise rendition of detail than is available from, say, the Sound-Lab A-3s that JGH so loves. Of course, the CS5s go a lot lower than the Sound-Labs, and will play a lot louder. Keep in mind, though, that this is the region where amplifier selection makes such a difference—the criticism I make here would have been much more severe with some of the amplifiers I had on hand. It's even possible that a pair of bridged Krell KSA-250s would remove some of the lack of definition I heard.