California Audio Labs Tempest CD player April 1987

Martin Colloms reviewed the Tempest in April 1987 (Vol.10 No.3)

The original Tempest, produced through the end of 1986, was reviewed by John Atkinson in Vol.9 No.6. As a result of using Mike Moffat as a consultant, however, CAL introduced some changes in early 1987, mainly relating to improvements to waveshapes and to power-supply noise levels in the digital sections. The new Tempest is not designated differently from the original, and costs $2095. An upgrade from original to current can be had for the difference in their prices ($200). CAL has also introduced a less-expensive, 16-bit player, the Aria ($1495), introduced at CES in Las Vegas and shipping since February 1, 1987.

Setting some sort of reference, the Tempest proved to be an interesting addition to this group, noteworthy for its complement of 6DJ8 tubes used in post-DAC circuitry. The case is large enough to contain the original Philips/Magnavox 2041 chassis as well as the new section, complete with its subchassis and substantial power supply. The Tempest is clearly built with great care, and the main circuit board is populated with high-quality parts. A good measure of the content of a high-end tube preamplifier is undoubtedly present.

The glowing tubes are visible under the perforated top plate, and the standard of finish is high, as befits a top-line model such as this. Why should a designer add tubes to a technology that is so thoroughly solid-state from start to finish that even sound itself is stored as an array of on/off binary numbers? The answer is simply that, whatever the music source available for the audio, there always seems to be scope for improvement in the playback quality. Whatever distortion, compression, masking, etc. occurs in the recording/replicating process, there is still scope, within reason, for a better preamp or a better cable to let through more of the music. It is just such improvements that the audiophile community values.

Thus, in the case of a CD player, whatever the ultimate quality of the system, the designer's eye will alight suspiciously on the relatively low-grade audio components used in the filtering and output sections of a stock machine. His response will be: "I can do better than this! No doubt, customers will value and be prepared to pay for the obvious improvements which I can bring to this section."

CAL felt that tubes offered both the best chance of ameliorating the more aggressive solid-state aspects of CD, and the best sound per se. By and large, that supposition appears to be true; the future looks bright for this company.

On the technical side, the player requires quite complex circuitry to first accept the output current delivered by the D/A converters; a simple preamp-style, voltage-gain stage is not viable here. The automatically switched de-emphasis must be implemented, as well as the overall three-pole Bessel filtering specified for the original Philips system. To summarize, separate left and right 14-bit DACs yield 16-bit accuracy: 4x oversampling at 14 bits, plus a cunning combination of digital filtering and digital noise shaping, result in an enhancement of resolution by two bits, providing the low-level distortion and noise of a 16-bit system. That this process works has long been confirmed by measurement, where, in general, Philips 14-bit players out-resolve many systems based on the use of full 16-bit linear DACs—even Philips' own!

The combination of digital and slow-rolloff analog filtering results in a transfer characteristic which is essentially linear phase. The original Tempest, however (though not the revised), was unfortunately absolute-phase-inverting, an audible effect in a player of this class. For the listening tests, it was necessary to invert the polarity (by turning over the speaker connections) in order to match the correct phase of the majority of other players.

Sound Quality
The original Tempest placed itself (just) at the top of the tree. My initial audition in phase-inverted mode exacerbated the hint of a bright, lean quality in the original, but was put in a more balanced perspective when phase-correct. The revised version appeared to take better control of the treble balance.

Measuring the response showed that the early Tempests really did have some mild treble lift in the final half-octave, but, in my view, this would be insufficient to account for the tonal quality. That player was a little bright, with a more forward presentation—the converse of the old Meridian, for example, with its more laid-back, mildly rich character. The revised Tempest was better in this respect, and was also absolute-phase correct.

In the midrange, the CAL was unique. By CD standards, its performance was exceptional in purity and clarity, the sound being well-detailed with a good feeling of transparency.

Stereo soundstages were well defined, stably focused, and showed fine width and depth. With the "right" CDs, a close approach to high-quality vinyl replay was apparent in this area, though still better focus was possible.

The Tempest proved a little shy in the upper extreme of the midrange; a touch more authority and bite would be helpful here, but in the treble its former brightness was countered by the promising level of treble clarity, which was so much better than the usual tizzy grain heard from many cheaper players. Some care over choice of cable and system matching will help the Tempest here; conversely, with a poor system match, the treble balance could prove an irritant. For example, the Tempest will suit a system based on the Apogee Duetta better than one based on the Magnepan MGIIIa.

The Tempest's bid for audio dominance finally rests on its dynamics, which, in CD terms, are among the best. The inner shading of textural dynamic contrast, masked to a perceptible degree by many other players, stood revealed. The bass was generous, with a natural quality on orchestral instruments, though judged a touch lacking in slam and speed on rock.

CAL has succeeded in setting a new reference standard for CD sound without the specific benefits of 16-bit D/A converters or heavily rebuilt digital sections and the like. The fundamental technology is early Philips, but the subjective results show just how much can be achieved with careful post-development.

In a sense, CAL proves what we already know: an audiophile-quality analog filter section is preferable to the usual collection of ICs and electrolytic capacitors. (With their Sonographe CD player, Conrad-Johnson has achieved a similar demonstration at a lower price/quality level.)

The CAL Tempest revised offers less fancy facilities and features than usual, but makes real claims to audiophile status and deserves very serious consideration by any CD enthusiast.

With the benefit of foreknowledge, we know that a well-toleranced current-generation Philips 16-bit, 4x oversampled chassis should be capable of about a 20% uplift in sound quality compared with its 14-bit predecessor. Let us also look to the future, and see what CAL can do with one of these new chassis in a more advanced model.

The CAL Tempest is the best player for reproduction of classical, naturally balanced, program, and will give the best depth and soundstage perspectives.—Martin Colloms

California Audio Labs
SONICblue Inc.
Santa Clara, CA
(877) 563-9388