Thiel CS2.3 Loudspeaker:
Page 3
The Thiels worked reasonably well in any obvious configuration, but careful positioning paid off in terms of their ability to re-create a realistic soundstage. I ended up with them firing down the long dimension of my room, just under 5' from the front wall and 5' 6" from the side walls. The speakers were toed-in slightly, their axes crossing at about the back wall. My listening chair was about 7' from the rear wall, placing me about 10' from a line connecting the speakers; my earline was at about 40", 8" above the axis of the midrange/tweeter.

Sound
If you're someone who's followed the evolution of the CS2 and '2 2, or are familiar with the CS6, you're probably not going to be surprised by the CS2.3. Excited, delighted, enamored with, taken by, in love with? Yes—but not surprised. First of all, you probably expect flat frequency response. The CS2.3 sounds as flat as Kansas, and free—stunningly free—of obvious colorations. Inevitably, when I replaced the '2.3s with another speaker, I was immediately aware of colorations—perhaps a hollow, clarinetlike coloration that overlaid the cellos and violas, or maybe a sort of "wooooooooo" sound that seemed to tint female vocals and make them sound is if they were tethered within a specific frequency band. Reinserting the Thiels, not only were the colorations gone, but the overall effect was that everything seemed to open up spatially as well, as if the instruments and voices were now able to move and breathe.

The second most obvious characteristic of the CS2.3 was one I didn't expect, at least based on my experience with the CS3.6. For all its strengths, the '3.6 never seemed completely coherent. The wonderful 3-D images didn't mesh with the background, and the character and size of individual instruments seemed to change slightly as their pitch changed. Not true with the '2.3. In fact, coherence was one of their strong points. Once I got them placed correctly and adjusted the toe-in just so, they completely vanished—not only as a source of sound, but as a distinct voice. All that remained was a seamless, completely coherent, three-dimensional soundstage.

On the Goossens/London reading of Rimsky-Korsakov's Scheherazade (Everest SDBR 3026), the sonic picture was astounding in its coherence. The opening notes of the solo violin soared outward and painted the stage, the surrounding space, and the hall boundaries with their decay into the venue's ambience, then wrapped all around the entire orchestra when it entered moments later. It was incredible! Every section, every instrument, was located exactly where it seemed it should be—not just with respect to the surrounding orchestra, but also with respect to the hall's side walls, back walls, and ceiling. I don't think I've ever heard a large-scale orchestral recording where all the pieces fit together quite so completely.

The CS2.3s also excelled at keeping individual instruments' sizes and characters consistent up and down the scale. The violins on Scheherazade showed this off, but a better example was Dick Hyman's In Recital (Reference Recordings RR-84CD). This is a solo piano recording with a very natural balance of detail, impact, and ambience, and a middle-to-distant perspective that incorporates lots of hall sound. Other multidriver designs I've tried make the piano sound a little like three different instruments changing noticeably in character—tonal balance, image size, and handling of microdynamic transients—as it moves up and down its range. With the CS2.3, there was no perceptible change in character as the piano moved up and down; just a seamless, natural portrayal of the instrument in a real acoustic space.

If anything, the CS2.3 favored coherence ever so slightly over definition, which is something I never imagined I'd say about a Thiel speaker. They didn't seem to focus quite as sharply as the '3.6es—or, more correctly, their image edges weren't as sharply defined. There was plenty of inner detail—voices within choirs and instruments within sections were distinct in their characteristics and timing—but their edges, and the edges of the entire choir or orchestral section, weren't as sharp as I've heard.

I don't want to make too much of this observation, however, for several reasons. For one, the system and setup are big factors. I could sharpen the CS2.3's focus with more toe-in, or by moving them with respect to the walls, but at the expense of soundstage coherence and the flat response. I couldn't optimize everything simultaneously in either of my listening rooms, but that's not to say another room might not be just the ticket. Also, the associated equipment I used most heavily—VAC electronics and Nirvana cables—all favor coherence over sharply defined details.

Another reason I'm hesitant to criticize the CS2.3's focus is that the line between sharply focused and over-etched isn't hard and fast, and varies from recording to recording. On the Reiner/Chicago recording of Prokofiev's Lieutenant Kijé Suite (Chesky RC10), the '2.3's focus seemed a little softer than it should be. On the Everest Scheherazade, on the other hand, it was dead on. With multimiked studio recordings like B.B. King's "R2D4" CD, Deuces Wild (MCA MCAD-11711), it was even more arbitrary. I've heard other speakers—the Gershman Acoustics Avant Garde and Genesis 200, for example, and certainly the CS3.6—outline the images more sharply, but I wasn't at all disappointed with the CS2.3s, and certainly wouldn't trade their coherence for additional edge definition.