Cello Palette Preamplifier
The Palette Preamplifier is a recent addition to the Cello line. Intended as a consumer version of the combination of the much more expensive Audio Suite preamplifier and Audio Palette equalizer (which together, including common power supply, now sell for approximately $30,000!), the Palette Preamplifier is the first Cello component intended exclusively for home use. Whereas the Audio Suite and Audio Palette use discrete circuitry and balanced inputs and outputs throughout their signal paths, the Palette incorporates integrated circuits, single-ended inputs, and is housed in a single chassis to keep costs down. While Levinson readily admits that the Palette doesn't quite have the resolution and tonal accuracy of the Audio Suite/Palette combination, he feels that it performs much better than its four-fifths-lower cost would suggest. Since I haven't had more than a few brief auditions of the Audio Suite/Palette, and those in unfamiliar systems, I can't comment on this. However, I can say that the Palette is an extraordinary preamplifier.
The Palette Preamplifier is visually similar to the more expensive Cello components: tall and boxy, with double rows of circular control knobs on its face. The top row of six knobs is dedicated to basic preamplifier operation (input selection, tape monitor, left and right input balance, EQ in/out, and output level), with the lower six knobs reserved for selected equalization points: up to ±22dB at 20Hz, ±12dB at 120Hz, ±6dB at 500Hz, ±6dB at 2kHz, ±12dB at 5kHz, and ±22dB at 20kHz. The rear of the unit offers five pairs of single-ended line-level RCA inputs, two pairs of single-ended RCA record outputs, two pairs of single-ended RCA outputs, and one pair of balanced outputs via Fischer connectors (Levinson claims that Fischer connectors sound much better than the more common industry-standard XLRs). The outboard power supply is hard-wired to the main chassis via two three-screw terminal strips.
The front panel is finished in brushed aluminum, with engraved black lettering and an inset, centrally located power-on LED. All 12 control knobs appear to be machined from the same brushed aluminum, and are all the same size. The bodies of the Palette and its power supply are finished in a gray matte similar to that found around military bases. Not exactly the sexiest audio component in the world, its appearance might kindly be described as "industrial" (footnote 2).
All input and output connections, input selector and tape monitor switches are pcb-mounted, the switch shafts connected to the front panel via extenders. The entire circuit is mounted on an upper printed circuit board, with the equalizer mounted on a lower board. Litz grounding wires are used to tie the various ground points to a central point, and a solid-copper, nickel-plated ground plate ties all the connector points together. The unit's input impedance is very high: 1 M-ohm (megohm). According to Levinson and Tom Colangelo, chief design engineer at Cello, one of the limiting factors in preamplifier design is the relatively low input impedance (10k-50k ohms) presented by various products on the market. According to them, most line-level units (CD players, processors, etc.) will not drive such low impedances without audible distortion; the addition of low-impedance balanced inputs compounds the problem. By designing for such a very high impedance, Levinson claims that the Palette Preamplifier presents essentially no load to line-level source components, with much better sonic results.
Levinson and Colangelo came upon this revelation quite by accident while copying analog master tapes onto DAT. The DAT copies were sonically inferior to the masters using conventional loading, so they upped the input loading to 1 M-ohm. This resulted in an immediate improvement, with the DAT copies "sounding virtually identical to the 30ips masters."
To equalize or not: a sound philosophy
Mention the word "equalizer" to audiophiles and they'll run screaming from the room. A dirty word, "equalizer," conjuring up visions of those horrible things indigenous to Circuit City that sell for a cool $99.95. But not all equalizers are created equal. In fact, all recordings are equalized to some extent, because every microphone commercially available has an equalization curve built in. Most recording mixing consoles have EQ capability, and engineers regularly utilize outboard equalization (either in the analog or digital domains) for final mastering. It may come as a surprise to audio purists that the majority of their treasured recordings contain some degree of post-production EQ. So why is it such a heinous crime to use high-quality EQ in domestic playback situations? Because most audiophiles believe, as I used to do, that any manipulation of the frequency balance in the original source will seriously compromise the sonic and musical results.
Over the past several months, while preparing this review, I've had the opportunity to spend some time talking about music and sound with Mark Levinson, also an accomplished musician who specializes in the double bass and various esoteric stringed instruments from the Far East. Like other musicians I know who are into high-end audio (more than you might think), Levinson considers the areas of tonal accuracy and transparency to be of utmost importance. Until I had the opportunity of acquainting myself with the Palette, I would have considered the idea of putting an equalizer into my reference system totally nuts. But I'm certainly not above entertaining another philosophy, especially for improvements in the musical integrity of the recorded performance.
Levinson suggests that EQ is a useful tool in musical reproduction because it addresses the number-one obstacle: tonal-balance problems in the program material. We all recognize that recordings vary tremendously in sonic quality. But why? Well, of course, there are different halls and different performers. But there are also different microphones (with strong sonic colorations and frequency aberrations (footnote 3), and studio monitor systems (including amps, cables, and speakers) that vary significantly in tonal balance. As Levinson states, "one system might be up 8dB in the high frequencies where another is down 7dB. If a recording is optimized for one system, it may not sound 'right' on another. Since even a dB or two can make a big difference, such big disparities pose a formidable problem that cannot be overcome by changing cables, processors, and equipment. The only logical answer is to give home listeners a user-friendly and sonically transparent instrument with which to correct tonal imbalance."
In short, he is proposing a new definition of the "purist preamp" (footnote 4); one which gets the listener closer to what the musicians themselves intended you to hear. The point is well taken, since no audio system can produce a natural sonic result with all recordings. Many recordings that initially sound "bad" on one system may sound much more natural on another. Levinson claims that Cello's original goal was to offer a combination of two products with this unit: a very high quality preamplifier, and an equalizer "for those who refuse to compromise their music."
A note of caution: Indiscriminate use of EQ can result in some bizarre sonic byproducts. Overzealous adjustment of the midband frequencies (500Hz and 2kHz) alters soundstage dimensionality, creating a too forward or distant perspective, depending on how far up or down the controls are adjusted. Careless use of the 5kHz and 20kHz controls will do similar damage to the natural harmonic structures of musical instruments and human voices. For that reason, the Palette is probably not the product for the sound freak who would use this feature to make a circus out of the music, cranking up the bass and treble and blowing the windows out of his listening room! But if you're searching for ways to get closer to the real thing, the Palette, judiciously used, can help restore musical equilibrium to recordings that may not be compatible with a specific playback system.
Some of you reading this will consider any serious discussion of EQ in Stereophile to be heresy (footnote 5). Certainly what I'm discussing here requires one to totally rethink the basic premise of musical reproduction. It's no secret that the various components, cables, speakers, etc. we use in home systems sound dramatically different. One of the most significant audible differences (forgetting the various mechanical designs) between two like components is the tonal balance, as Levinson suggests. This balance dictates a great deal of the way we psychoacoustically interpret such things as soundstage dimensionality and harmonic textures, just as we do with live music in the concert hall. The Palette enables its owner to render that balance more real-sounding.
Footnote 1: The Mark Levinson brand name is now owned by Madrigal Audio Laboratories. Readers should note that there is no connection between Mark Levinson products and Cello products.—Lewis Lipnick
Footnote 2: Lynn-Jane, my wife, actually finds the Palette rather attractive. She notes that its plain, self-effacing looks don't call attention to the fact that this is just another expensive "audio toy for boys."—Lewis Lipnick
Footnote 3: Prior to joining the National Symphony, I worked as a recording engineer. I would carefully choose specific microphones according to the various hall acoustics, musical material, and frequency range (ie, free-field vs spot miking, vocal vs instrumental, low- vs high-frequency propagators).—Lewis Lipnick
Footnote 4: For those students of high-end audio history, in 1974 Mark Levinson introduced one of the first—perhaps the first—commercially available preamplifier lacking tone controls. The JC2, as it was called, created a minor revolution throughout the industry, since the idea of offering a so-called "purist" preamplifier without tone controls was considered much too avant garde for the establishment. According to Levinson, available technology at that time prevented the design of acceptably high-quality EQ circuits. Not until the mid-1980s (according to ML) was it possible to design an acoustically transparent analog equalizer.—Lewis Lipnick
Footnote 5: To be fair, it's not just the image of equalizers as cheap pieces of junk which leads audiophiles to reject the entire concept. Rather, with the kind of narrow-band equalizers commonly available, it's impossible to "correct" program material for any specific tonal change that might have been imposed by the engineers, even if it were possible to know what those changes had been. Why, then, put another device, one that will need to be as sonically pure—and therefore as expensive—as a high-quality preamplifier, in the signal chain if all it can do is introduce arbitrary changes in tonal color? The equalizer concept offered by Mark Levinson in his Cello products, based, I believe, on an idea by veteran engineer Richard Burwen, is to offer the user broad, overlapping EQ bands that stand a much better chance of actually compensating a recording for the slings and arrows of outrageous tonal change it has endured on its way to the listener's home.—John Atkinson