The Pioneer Paradox and Euphonic Distortion
I read Stereophile Magazine—but it’s fair to say I’m not a high-end audiophile. I read the magazine for its excellent
reporting on new recording technologies (and related media). When it comes to many of the equipment reviews and
columns, however, I read Stereophile
as a joke book, and as a running mostly-unconscious commentary on foolishness and gullibility.
One issue in the spring of 1999 particularly struck me for two or three reviews
of very expensive audio components. All of the components were praised to the skies, with no doubts as to their
musicality and true high-end performance. And in each case there was a set of graphs showing actual measured performance,
with the measurements typically done by someone other than the reviewer.
What struck me was how badly each of the components measured. Not one of these
components offered high fidelity if that phrase means accuracy: not one would consistently pass signals without
distorting them.
I have no doubt that each of these components makes pretty music: in every case, the distortions would tend
to soften the edges of recordings, making them less crisp but, well, prettier. To my mind, these components are
like the Vaseline smeared on camera lenses when used with certain aging stars: they cover up some of the flaws
and make it all more attractive.
Did any of the reviewers say that? Of course not. Years ago, one of Stereophile’s editors commented that a whole
class of expensive low-power tube amplifiers was really not amplifiers so much as non-adjustable tone controls.
Many years before that, Stereophile’s
founder (the redoubtable J. Gordon Holt) used the term "euphonic distortion" to describe changes in a
recorded signal that may make it more attractive—but that nonetheless distort the original.
I’m not an expert audio reviewer, I haven’t heard any of these devices, and I’m
not ready to say that reviewers don’t hear what they say they hear. But I do wonder.
Knowing the Source
I wonder, in particular, because Stereophile and its competitors in the True High End absolutely despise double-blind reviewing and
assert that you can’t possibly hear a component properly under blind conditions. Many of the reviews include loving
descriptions of the build quality of a particular device, conversations with the designers, musings on the other
fine products of the company, and other aspects that couldn’t be part of a blind review.
When I read the two or three reviews that eventually generated this essay, I
thought: "How would these items have been judged if they had had Pioneer or Sony or Yamaha nameplates on them?"
(Not Pioneer Elite or Sony’s audiophile line, but ordinary Pioneer or Sony equipment—the kind despised as "mid-fi"
by all True High End people.) I believe there are three answers to that question:
- "Who has time to bother with that unmusical crap?" No high-end journal
would review mainstream audio equipment: it’s beneath them.
- If (by some mischance) such equipment was reviewed, it would be savaged—not
only for the ordinary build quality but for the obvious ways that it distorted the signal, even if that distortion
was superficially pleasant.
- Except at the very lowest price levels, I doubt that Sony, Yamaha or Pioneer
makes audio components that measure as badly as these high-end devices.
If I’m right about the second point, then consider what would happen if Stereophile Magazine did all of its sound reviewing with components in black boxes:
holes for the control knobs and a third-party universal remote control, but otherwise an unidentifiable box? I
realize the reviewers would never stand for it—but, logically, they should welcome the idea. After all, the only
thing that should matter for
a review of the sound quality of equipment is the sound quality.
This isn’t a call for double-blind reviewing. The reviewer should be able to
spend as many months as he (or, rarely, she) sees fit, discovering every nuance of the sound of a component. But
that sound quality should be judged independent of the maker of the equipment or its price tag, and I will assert
that it almost never is. A $10,000 CD player that applies its own special "time-balanced" equalization
curve to compact discs is a triumph of engineering and musicality when it comes from the right company; a $100
CD player that’s down one-half deciBel at 20KHz is a piece of garbage if it comes from a mainstream audio company.
Multiphase Reviewing
An objective equipment review would have three separate phases, with evaluations
written at each phase. The first and longest-lasting phase would be the "Pioneer phase," in which the
equipment’s price, maker, cabinetry, internal design and all that was entirely concealed from the reviewer. The
reviewer would be expected to judge the device strictly on the basis of how well it performed—and would not be
allowed to rewrite that judgment based on later phases. I’m asking here for "objective subjectivity"—an
evaluation based on how the component sounds, but without any external issues to warp that evaluation.
The second phase would be the measurement phase. Ideally, that would also be
done without knowing the maker, price, or internal design. I don’t claim that any set of available measurements
will accurately portray the sound quality of a piece of equipment—although I will claim that, when measurably defective
equipment is highly rated, there are inherent questions that aren’t being answered.
Finally, you’d have the unveiling: a description of the actual item, which could
include comments on fit and finish, design characteristics, and all that. There could even be a third commentary
based on the revealed aspects of the component—but only after the objective judgments had been offered.
I believe that a series of such three-phase evaluations would be quite revealing,
but I don’t believe it will ever happen. I do believe that the reputation of a component’s maker, the nature of
the component’s design, and a component’s price tag all do influence the sound as it is understood by the
reviewer.
(I’m aware that this evaluation process wouldn’t work for the extreme edge of
the True High End, the little wooden blocks, stones, and other devices that seem to belong in the psychoceramic
area of audiophilia.)
Euphonic Distortion
I won’t get into the LP-vs.-CD issue here. I suspect that Li’l Mikey Fremer knows
what he’s up to: in the June 1999 issue, applauding a case in which a symphony orchestra was recorded using analog
tube-based equipment, he notes that "most classical producers probably prefer, or think they prefer, the dead-quiet noise floor of digital."
Aha. I see the hint of an admission there: that analog is "better than digital"
not because it’s somehow uncovering more music—but because it’s adding to the signal, specifically the background noise that’s inherent in all analog tape recording
and all LP recordings. That noise is "part of the music" to many people, even though it’s an artifact
of the recording process.
That, in a nutshell, is euphonic distortion: in other words, pleasant lies.
Am I opposed to euphonic distortion? Not particularly. When I listen to music
on my PC, I usually turn on the Dolby Surround feature, even though that’s actually distorting any music that wasn’t
recorded with Dolby Surround. I know it’s distortion; I also find it pleasant under those circumstances.
I am
opposed to claims that euphonic distortion represents higher accuracy or that it’s somehow "better than accuracy."
If a classical producer wants to have appropriate background ambiance for a recording, that ambiance can and should
come from the recording venue itself; there’s no good reason for the recording chain to add its own ambiance.
Violins should be musical. Clarinets should be musical. Steel drums should be
musical. Auditoria should be musical. CD players, amplifiers, and speakers should be accurate
in reproducing that musicality—with preamps and control centers to let
listeners add their own "musicality" to taste.