Photo of Bob’s mastering room courtesy of Mary Kent, photographer

I’ve written before in this space that to me the most wondrous aspect of our avocation (apart from the music) is the way it exists at the intersection of logic and emotion, of science and art. The equipment we use is made by engineers applying scientific principles, yet its goal is to deliver sensual pleasure. Both viewpoints are valid.

A few engineers embody both approaches. MBL’s Jürgen Reis recently told Stereophile that he starts by optimizing a design objectively—getting the technical part right. Then he voices by ear. Alta Audio’s Michael Levy recently told me that he refuses to use a certain technique that would make his speakers measure better because it would make them sound less like live music. Thus do art and science interact within the audio industry’s design laboratories.

When end users—audiophiles—evaluate a system, our main criterion is its ability to convey the music’s message. I believe this is nearly universal; indeed, I struggle to imagine a valid criterion other than this. Many of us, though, also want to make sure a product gets the engineering part right, which is why Stereophile publishes measurements alongside its subjective reviews.

It’s instructive to consider how different things would be for end-user audiophiles if we approached hi-fi the same way audio professionals do—and here I mean both the engineers who develop our equipment and those who record, mix, and master our music. If we listened like they do, the experience would be different in many subtle ways and a few unsubtle ones—to wit:

• We’d spend at least some of our time listening to one speaker only. Years ago, someone realized that listening in stereo complicates loudspeaker assessment. Listening to just one speaker—no stereo, no spatial effects—makes it easier to figure out what’s going on sonically.

The one-speaker approach is supported by research showing that, generally, loudspeakers preferred in single-speaker tests are also preferred in stereo. And yet—let’s face it—the experience is impoverished compared to stereo listening in a good room. Apart from a handful of mono enthusiasts, few people listen this way for pleasure.

• We’d listen without sidewall reflections. At home and in ordinary rooms, this usually means walls with sound-absorbing or sound-diffusing panels. Among audiophiles, it’s practically dogma that sidewall reflections are bad, but the issue isn’t clear-cut. In his book Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms, Third Edition—Kal Rubinson reviewed the first edition in Stereophile—Floyd Toole writes about the problems with the “phantom center image” produced by stereo loudspeakers, arguing persuasively for adding a third speaker in the middle (but where’s the fun in that?).

Counterintuitively, such problems are mitigated by sidewall reflections (footnote 1). Sidewall reflections can also improve speech intelligibility (footnote 2), and we all know they can expand the soundstage under appropriate conditions.

This is related to a broader point. For all its many virtues, two-channel stereo is flawed. There isn’t enough information from two channels to provide a proper reverberant musical experience. Stereo needs assistance from the room. Sidewalls help create a feeling of envelopment.

And yet: Mixing and mastering engineers have long worked to avoid sidewall reflections. Mastering studios are often built with a “reflection-free zone” at the sweet spot so that the engineer can hear precisely what’s on the recording, with no enrichment.

People developing audio hardware have also been known to eliminate first reflections. In a JAES article from 1979—also cited by Toole—a group of Yamaha engineers reported that when evaluating the sonic performance of products they were developing, they listened without sidewall reflections. But when listening for pleasure, they listened with reflections. It sounded better.

• We’d listen very loud. Every recording—especially of acoustic music—has a volume level at which it sounds most real. That is not the level mastering engineers listen at, or not usually. Their goal is to hear as much of what’s in the recording as possible, so they turn it up.

In his book Mastering Audio: The Art and the Science, mastering engineer (and Stereophile contributor) Bob Katz (who, by the way, prefers working sans reflections) proposes an approach he calls “calibrated mastering.” In a studio so equipped, a –20dBFS pink-noise signal produces a level of 83dB (one channel) or 86dB (uncorrelated 2-channel) at the listening position; this level, Katz says, is “forte” in music. That may not seem that loud, but try it. I find it aggressively loud.

To be clear, this is not the level Katz proposes working at. For mastering work, he turns the volume down from there, by 6dB for source material with a wide dynamic range or by 14dB for music that’s more compressed—more squeezed toward maximum volume. While much of the music would then be softer than that –20dBFS, peak levels would be 20dB louder. In a small listening room, that’s loud (footnote 3).

• We’d listen to the same music, over and over. For critical listening, you don’t want surprise; you want familiarity. That means listening to the same, familiar, diagnostic tracks repeatedly. One fav, for the tests mentioned in Toole’s book, is “Fast Car” by Tracy Chapman—a good song the first several hundred times you hear it.

What does this prove? Just that listening for maximum information retrieval and listening for pleasure are two very different things. Yes, you, too, can listen like a pro—just don’t expect to enjoy the music as much.—Jim Austin

Footnote 1: See Shirley, et al., JAES Vol.55, p.852.

Footnote 2: See, for example, https://bit.ly/2QK8kEW.

Footnote 3: A curious fact: Perceived loudness depends on room size, even at the same SPL—but my listening room is quite large.

Click Here: south sydney rabbitohs rugby store