AT THE TIME WHEN HIGH AUDIO BEGAN, ONLY ANALOGUE SOURCE MATERIAL WAS AVAILABLE, and vinyl dominated the audio scene. The LP does not stand up very well by any obvious measurement criteria except upper frequency limit, but it sounds, at its best, remarkably good. It was this discrepancy that led high end audio away from trying to evaluate sound quality via measurement and even from trying to correlate the two.
Digital in contrast is naturally numerical interpretations spring from it automatically: Dynamic range = 6 x number of bits; maximum frequency = -sampling rate, and so on. Numbers have crept in the High End door, sneaking in on digital's shadow.
The Bryston 3B-ST amplifier is an interesting instance - it offers both superb sound quality and superb performance by numeric measures. Now I surely do not mean to suggest that the Bryston was designed by numerical measurement criteria alone. The Bryston people listen very carefully and also have collected listening impressions from their dealers and customers for many years. But they seem to have established over years of listening and measuring, a way to connect the two in practice. And to some extent, the 3B-ST represents a successful attempt to live up to the numerical criteria implied by digital standards, while at the same time offering truly remarkable sound quality from either digital or analogue sources. It is a champion in both worlds, the technical and the musical world of listening.

In the context of the attention now being paid to new digital standards, and the descriptions of the significance of numbers of bits, it seems appropriate to translate usual ideas of amplifier performance - technical measurements and, when possible, listening impressions - into digital terms. In this situation, 6 dB is 1 bit. For example, 16 bits correspond to 96dB of dynamic range or signal to noise ratio. (Actually, 1 bit is equivalent to slightly more than 6dB, but 6dB is a good enough approximation for present purposes.)

To clarify this kind of conversion to digital, let's think about noise first. The 3B-ST is extraordinarily quiet. Its noise level is circa 115 dB below full output. (Exact figures depend on what kind of weighting is used.) This means that the 3B-ST can cover the whole dynamic range of CD with ease. Indeed, its signal-to-noise ratio is equivalent to 19 bits plus (6 x 19 = 114). In listening terms, this means a silent "black" background.

Now let us take distortion. Distortion figures (THD + noise) for the 38 ST run on the order of .002% or less of the signal at nearly the rated output of the amplifier. This translates to distortion about 94 dB or more below signal. (1% = 1/100 of signal=-40 dB; 1%=1/1000 =-60 dB, the figure dropping by 20 because power in dB is proportional to voltage squared while percent distortion is given in voltages, not power.) Once again, the full range of CD is covered, or essentially so, This is truly extraordinary performance, matched by few amplifiers.

For comparison's sake, suppose an amplifier has 1% distortion + noise. This is 40 dB down from signal. But that is only between 6 and 7 bits down! So if a signal is at the peak level, 16 bits, then the error is already affecting between 9 and 10 bits. In effect, the resolution is enormously reduced. (No one thinks much of a CD player that drops 9 bits!) Similarly, 0.1% distortion + noise, 60 dB down, is 10 bits down, so at the 16 bit level, 6 bits are turned into garbage.

This might seem at first sight to an exercise in dB arithmetic. But listening to the 3B-ST, one begins to think that it is not idle at all. The amp is unusually transparent. So much so that prolonged listening with high quality digital sources has given me a new appreciation of what can be accomplished in a certain kind of crystalline clarity, a virtually complete transmission of signal. Granted the 3B-ST seems to have a slightly more prominent top end than most amplifiers. But the clarity is more than a frequency-response effect, or so it seems to me. Indeed, even when I used a Sigtech "target curve" to roll the top down somewhat, the clarity remained.

Of course, interpreting listening experience in terms of technical criteria is risky. Moreover, it is hardly clear how measured electronic performance, which invariably exceeds speaker performance in measured terms by such a large margin, is reflected in the listening experience at all. But somehow the Bryston 3B-ST together with, for instance, the exceptional Harbeth Monitor 40 speakers with (in particular) their remarkable upper frequency performance, gave a virtually unprecedented access to fine detail of sound. When the Sigtech prodded the system the last small steps to neutrality, the result was a remarkable insight into recordings, a feeling that the sonic window had become cleaner than almost ever before.

There is a school of thought that finds kind of nearly absolute transparency not entirely consonant with musical experience. This was made explicit some years ago by Jean Higara, dean of French audio critics (and designer of a line of well thought-of tube amps), who claimed that the low-order harmonic distortion of tube amplifiers was not just benign but actually beneficial that it masked disagreeable higher-order distortions in the recording. (This refers to power amps only. Low-level tube circuitry can be very low in distortion.) In my review this has been extended, without being made so explicit, to noise masking as well. Some people seem to have come to hear the noise-masked sound as natural and musical. And of course the fact is that most recordings, especially older recordings, were consciously or not, balanced and miked for playback in a masked playback environment, since until recently no other environment was possible. (I refer now to electronics; the old Quads, whatever their other limitations, were and are quite an unmasking speaker.)

Times change. As it happens, Bryston amplification is widely used in serious professional monitoring. To take a rardom example, I noted shortly before writing this review that the Chesky recording of David Chesky's Three Psalms for String Orchestra was monitored with Bryston amplifiers. Perhaps not coincidentally, this recording sounded remarkably pure and beautifu1 with the 3B-ST. Indeed most recordings showed their best. Certainly one could observe their limitations; and, sometimes in rather startling ways, problems in the music itself were made apparent that were in other circumstances not so obvious. For instance, slight irregularities in piano voicing, notes with a bit of excess "twang" or hammer tone, were unusually detectable on close-to piano recordings, and so on. But on the whole, being able to hear so far into the recordings was much to their musical advantage. The clarity was hypnotic in musical terms.

I believe a new general picture is emerging here. Audio until recently was a matter of distortion and noise masking other distortion and noise. In the absence of a source medium that even began to approach the signal-to-noise performance of the ear, the perception of quality in p1ayback was inevitably a matter of revealing a lot but not too much. We instinctively sought components that would reveal the musical wheat but hide the noise and distortion chaff. Of course, this was going on at a very subtle level, but going on it nonetheless was.

In this context, digital, deeply flawed though it was in early execution, eventually improved to the point that it became a salutary shock. In spite of the bandlimiting of the CD standard, it otherwise offered a far lower level of noise and distortion than vinyl. And it thus called for amplification that would match its unmasking capabilities, both in low noise and low distortion. At the same time, it demanded, with its uncompromising high frequencies, high-frequency transducers of the highest possible quality. Without complete success in all aspects the digital source itself, the amplification, and the speakers, especially to the top end the result could be musically less satisfying than the old systems, involving complementary maskings, which had also developed a great degree of refinement.

The 3B ST and Bryston BP-25 preamp with the Monitor 40s Sigteched (with the Sigtech also functioning as a D-to-A converter, and on occasion with the Sigtech's digital output feeding the Morch D-to-A), and especially in the context of the extraordinary ScanSpeak Excel tweeter as implemented in the Monitor 40s, signa1 for me a new era. They provide a system in which the full theoretical potential of CD digital is nearly revealed. And this potential comes quite close to meeting the capabilities of the ear itself in many respects. The new digital standards will come even closer, but really well-executed CD is already surprisingly good.

This does not mean that musical realism is complete. Stereo itself has limitations, as do microphones. And even exceptional speakers, even combined with the Sigtech system, do not attain the low distortion, low noise, and smoothness of response of the amplification. But in this total system, I do find a sense of clarity of presentation without harshness or edginess and a purity that is an unending source of musical satisfaction.

There are many amplifiers that offer distortion nominally well below the supposed thresholds of audibility of distortions of various kinds. But the Bryston 3B-ST made me wonder it these supposed thresholds have not been set far too high. The 3B-ST simply sounds cleaner and purer and more nearly devoid of grain and noise, whether signal related or otherwise than almost any other amplifier I've heard.

For all its virtues, the 3B-ST is not universal in applicability. It is not seemingly very happy with loads below 2 ohms, and with difficult loads and/or low-sensitivity speakers, it can complain. (The original Carver Amazings had it crying for mercy when I turned it up.) The 3B-ST is a rapier, not a saber. Bryston has bigger amps (or one can run the 3B-ST in bridged mode) when more muscle is needed.

I am well aware that some people will look at the 3B-ST and say, oh, just another solid-state device of moderate power and extended highs. And the 3B-ST is indeed a fairly unconventional amplifier, as I understand it, as far as general circuit design goes. But the Bryston people seem to have managed, by consistent, long-term listening tests and technical refinements, to have moved to the forefront in the particular way I have been describing. The 3B-ST is an unpretentious, small black box. It is not a designer statement, visually. And it is not even very expensive. But it goes very far in the direction of perfection in amplification, in transmitting the entirety of its input signal. Built carefully, it is guaranteed for 20 years. You are likely to want to keep it that long. When the new digital standards show up and even with CD reaching its potential, most amplifiers are being left behind. What use is 20 bits of resolution, or even just all of CDs 16, if your amplifier has resolution and S/N ratio for only 12? The 3B-ST is up to the challenge.

We invite you to experience the Bryston SST² Series amplifiers

20 Year Warranty - A Generation of Music