Solid State Preamp Reviews

How can a component with such clarity and transparency avoid sounding sterile and analytical? I don't know. When I reviewed the Krell KBL preamplifier last year in Stereophile, I came to the conclusion that it was "ruthlessly revealing" of all source components. In one sense this was a compliment, since it retrieved so much more information. However, it also raised questions about the KBL owner's ability to enjoy the music emotionally. As I've mentioned before, live music isn't always pleasant. Massed brass and string instruments can sound downright nasty when pushed too far, and the human voice can also sound mighty brittle. But the warm sound of human flesh plucking a harp or guitar, or the glorious sound of a full chorus singing pianissimo, can make your hair stand on end.

The fact that the Cello delivered all of this much more realistically suggests that it wasn't doing anything at all, while the others were producing specific colorations that selectively filtered the original musical signals. There's nothing wrong with a specific sonic signature, unless it begins to get in the way of the music. If I can't hear the difference between the Vienna Grosser Musikvereinsaal and the Amsterdam Concertgebouw (as is the case with the No.26S), a significant element of the performance is, in my opinion, lost. Or if the trombone sections of the New York and London Philharmonic orchestras sound alike, you have to question a component's tonal accuracy.

Another kettle of fish
Equalization certainly isn't necessary with the Palette preamp, but it's very effective when used with restraint. I've found that the majority of high-quality classical discs require no EQ. But some, due to either gross anomalies in the monitors and microphones used, or sheer engineering incompetence (I've certainly played my share of those sessions), sounded considerably better with careful EQ. Some pop discs, mastered by engineers who leaned heavily on the sonic Tabasco, can also be made to sound more believable with a little EQ touch up. Levinson points out, and I agree, that the zero EQ setting will probably differ from bypass by a small but audible amount. For that reason, I recommend that bypass mode be initially used before you start experimenting with the Palette's EQ. Also, boosting or lowering any of the frequency controls will affect the overall volume of the equalized signal. It is imperative that the equalized signal be adjusted via the two input controls to match the bypass signal level.

What does this equalizer do to the music? It depends how you use it. As I noted earlier, if you view this component as simply a fancy tone control to add a bit of "zip" to your system, you'd be better off staying away. But I've discovered that boosting the 20kHz control by only half a dB or so can restore a surprisingly large amount of ambience to acoustically "dry" recordings produced with hypercardioid microphones. Conversely, decreasing the 120Hz control can eliminate some of the upper-midbass "roar" often created by widely spaced omnidirectional mikes. I don't know what monitor speakers were used for final mastering of Jennifer Warnes's Famous Blue Raincoat (Cypress CD661111-2), but I'll bet they had an upper-midrange dip and a high-frequency peak, because there's certainly been a substantial amount of compensatory EQ. By gently lowering the 500Hz and 2kHz bands by 0.5dB, I can get rid of a slightly hollow upper-midrange ring. It's also possible to restore more of the harmonic color in Warnes's voice by boosting the 20kHz control by just 0.5dB.

This slight adjustment opens up the sound and allows the ear to more easily follow many of the less obvious musical lines. Individual tracks laid down in the recording sessions are now clearly discernible, and even the amount of reverberation assigned to each track is apparent.

Careful adjustment of the six EQ bands can really bring the overall musical presentation into focus, particularly when there appear to be some distinct foreign colorations. For example, the performance of Franz Liszt's Variations on the Figured Bass of the Cantata "Weinen, Klagen, Sorgen, Sagen" for organ (Marie-Claire Alain, organ; Erato ECD 88241) definitely benefits from a gentle drop at 120Hz (-0.5dB) and boosts at 5kHz (+0.5dB) and 20kHz (+1.0dB). It appears that Erato, in their usual form, opted for pairs of omnis in the Cathedral of Orleans, with a resultant upper-midbass rise and high-frequency dip. This adjustment opens up the soundstage, places the various ranks of the organ in proper perspective, and clearly delineates the swell organ behind the principal. (With the EQ, I can even hear the swell shades opening and closing!)

An even more dramatic improvement came with another Erato recording, one in which I performed (Shostakovich's Symphony 7, Mstislav Rostropovich, National Symphony Orchestra, Erato 2292-45414-2). The use of several "mirror-image" pairs of widely angled B&K omnidirectional mikes created a blurred, washed-out image with an absolutely horrendous upper-midbass boom and midrange squawk. By severely cutting the 120Hz (-2dB) and 500Hz (-1.5dB) controls, and boosting the signal at 5kHz (+0.5dB) and 20kHz (1.0dB), I could make this atrocious recording of an excellent performance at least listenable. It's important to note that this is the first equalizer I've heard that didn't change the overall characteristic of the musical performance. All of the musical voices retained their harmonic identities, and in proper ratio to one another. Soundstage, however, was greatly affected by overdoing the mid- to upper-frequency adjustments. The whole orchestra ended up in my lap or was moved out into the next room when the 500Hz and 2kHz controls were cranked up or down too far. More than 2dB in either direction was treading on thin ice. A modicum of taste, therefore, is advised.

But just as many recordings don't require additional EQ, such as the spectacular performance of Sir Granville Bantock's Hebridean Symphony (Vernon Handley, Royal Philharmonic, Hyperion DCA66450) (footnote 11). The overall sonic impression can be more vivid and exciting with added EQ, but I found that any EQ got in the way of the music. The same was true of Antoine Brumel's Mass for Twelve Voices (Paul van Nevel, Huelgas Ensemble, Sony SK 46348). All of my attempts to improve this gorgeous recording resulted in sonic and musical disaster. It needed no help; the wonderfully transparent Palette, without EQ, provided all that was necessary. Of course, these observations, valid in my case, may not necessarily work for you. That's the major strength of the EQ argument: it's now possible to musically optimize just about any recording, in any system.

Shortcomings
In spite of the Palette Preamplifier's excellence, I had three major complaints and one minor quibble. The first complaint concerns the lack of absolute polarity switching capability. Many digital recordings suffer from reversed absolute phase, which can cause significant musical damage (especially obvious when played through this preamp). The Theta Pro processors have polarity switching but many similar products (such as the Krell 64X) do not. I feel it to be unacceptable for any high-quality line-level preamplifier to be offered nowadays without such a function.

My second gripe concerns volume linearity at the bottom of the scale. At very low levels (between zero and one) the right channel disappears. Levinson claims that an equally good-sounding volume pot linear down to zero level would be prohibitively costly. He also suggests that since the B&W 800s are so sensitive (4 ohms, 93dB/W) and the output of the Theta Pro D/A converter so high, the low-level problems I get with my system would not be of concern in other situations. He's got a point, except for the fact that the problem does exist in my system. It should be remedied.

The Palette Preamplifier's lack of channel-balance capability and stereo/mono switching is, in my opinion, a serious oversight. Interestingly, one can adjust the left and right input levels, but only when EQ is engaged. If the listener chooses to bypass this function, there is no ability to trim left and right balance. If all rooms were perfect, this wouldn't be a problem. But in the real world, a balance control of some sort should be available at all times. The ability to sum both channels into a mono signal is very important, since it allows the listener to check for balance and phase coherence in the recorded material.

The fourth and least distracting problem involves the use of balanced three-pin Fischer terminators vs industry-standard XLRs. I'm not going to argue Cello's claim that the Fischers sound better, since I don't have any way to do a blind A/B comparison. But for the sake of the people who purchase Palettes, who will be forced to have their cables reterminated with Fischers (unless they also purchase Cello Strings), it would be helpful to maintain the industry standard and keep confusion to a minimum.

Conclusions
As Stereophile's Musician in Residence, my mandate is to evaluate those products that go the extra distance in the quest for musical honesty. Of all the reviews I've submitted, this one has been the most difficult to write. The closer something comes to the absolute, the less clear become the distinctions between real and reproduced, rendering conventional descriptions and comparisons irrelevant. At the same time, the review process has been more than the usual exercise in component evaluation. It has been a revelation. Although I certainly knew that a preamplifier could make a significant difference, the degree of change far exceeded my wildest expectations.

Similar to the B&W Matrix 800 loudspeaker, the Palette goes far beyond anything else I know of in closing the gap between live and electronically reproduced music. Setting a new standard for soundstage dimensionality, it gleans multiple layers of previously undetected spatial information from the recorded material. Much more than a preamp with equalizer, it is a refined musical reference tool. As Al Merz, one of my National Symphony colleagues, eloquently stated, "with this preamp you hear the musicians...the others give you someone's idea of how they think the music should sound."

The Cello Palette Preamplifier isn't perfect. Addition of a polarity switch, a better matched, more linear volume control, and channel balance and stereo/mono switching capability, would all be nice. And the $6500 price tag is a bit steep. But it still sounds better than anything else I know of, in spite of the shortcomings and cost. This is not an audiophile product, in the strictest sense of the word; it addresses the finer points of musical reproduction usually ignored by the sound freaks. It has no detectable sonic signature, nor does it grab you by the throat, screaming, "I'm a preamplifier!" If you're expecting a gut-busting, all-you-can-eat buffet of sonic fireworks, or a designer's fantasies of how the music should be presented, look elsewhere.

In a nutshell, the Palette Preamplifier gives the listener a glimpse of what performers experience every day on stage: total immersion in the music. Should you shell out the bucks and buy one? Just like everything else in this world, it depends on your priorities, and the realization that even the best is imperfect. Would I put my money on the line for this preamp? Yes, indeed.

---

Footnote 11: Granville Bantock (1868-1946), an English composer well known at the turn of the century, has fallen into undeserved obscurity. The Hebridian Symphony, as well as the other works on this disc, represent some of the most richly melodic, harmonic, and well-orchestrated pieces in the contemporary repertoire. A fabulous recording engineered by Tony Faulkner, this CD is a "must buy."—Lewis Lipnick