There’s no doubt in my mind that the goal of high-end audio is to reproduce the sound of live music in the home. That’s the endgame, the ultimate: Make it sound real. As if we’re really there with the musicians.
But live music isn’t an audio reviewer’s real reference. A reviewer’s reference comprises a selection of hopefully well-recorded and well-rounded audio tracks played through his or her audio system. That system -- the reference system -- is what the reviewer uses to evaluate and compare audio components. A reviewer doesn’t switch back and forth in quick succession between a live venue and a listening room, attempting to hear differences. No. A reviewer plays the same recordings over and over, comparing how they sound with the reference system intact vs. how they sound through that same reference system, one of its component replaced by the product being reviewed.
Given that, what should a reviewer's reference system be? It should be as accurate -- that is, as neutral -- as possible. I can go into the nuts and bolts of the meanings of accurate and neutral another time, but essentially, these terms mean that what comes out of an audio system should be the same as what goes into it. A reviewer should try to assemble a reference system that is as close to this ideal as possible. Among other things, this means using electronics within their safe operating ranges (e.g., amps that don’t clip with most speakers). With electronics, that’s pretty easy to accomplish these days. Pick a Bryston, a Coda, a Gryphon, or a Simaudio amp, for example, and source components that include modern D/A converters, and you’re almost there. What’s usually hardest to get right is the speakers -- typically, the most flawed link in the chain of audio components.
A real example
About five years ago, when I was considering hiring Randall Smith to review products for the SoundStage! Network, I visited him at home (he lives nearby) to listen to his audio system. His setup included some well-regarded Energy Veritas speakers and Anthem electronics, along with a honkin’-big subwoofer. Although the overall sound wasn’t bad, the bass I heard that day in Randall’s room was boomy and overblown. So I set out to fix it. I spent a couple hours measuring and listening to his system, and calibrated his in-room response below 200Hz. I finally got the system to sound and measure basically flat, then told Randall to live for a while with the changes I’d made and let me know what he thought.
He later told me that he listened to his system as soon as I’d left and hated how it sounded. To him, it was simply too lightweight in the bass, and therefore unenjoyable. To Randall’s credit, he didn’t touch my settings and, in a matter of days, began to hear things in his music that he hadn’t heard before. Midrange textures became more apparent, highs had detail that had previously been obscured, and he began to appreciate a new articulateness in the bass. His own internal reference was changing, and for the better.
A year or so later, Randall had a reference system that sounded basically neutral, and I felt comfortable sending him his first product to evaluate. Today he’s moved into a much smaller room, and his system, too, has changed. His Vivid Audio V1.5 loudspeakers, mated to a JL Audio Fathom f110 subwoofer, anchor a compact reference system that sounds wonderfully transparent and extended and, yes, neutral. I’ve heard it several times and I’ve seen the measurements. I’ve played my music through his system, and our opinions of what we’ve heard have coincided. Randall’s reference system is as accurate as I could hope for.
But what happens when a reviewer has a reference system that doesn’t provide a useful reference? Problems. Big problems. Again -- although all of us are trying to re-create the sound of live music in our homes, a reviewer’s reference is recorded music listened to through his audio system. That’s what reviewers use to judge components with, and that sound becomes what they consider to be "accurate." If there’s a flaw in a reviewer’s reference system that the reviewer doesn’t hear as a flaw, then there’s a real likelihood that when a more transparent, more neutral, more accurate component is inserted in that system, it’s the newcomer that will be misjudged as sounding inferior.
There have been several recent examples published in the audio press. In one case, a set of in-room speaker measurements revealed the reviewer’s reference system to be terribly nonlinear, and other measurements showed that the system’s performance was dramatically improved when the speaker under review was inserted. Yet the reviewer’s conclusion was that the nonlinear setup sounded most like live music. Well, no -- the nonlinear setup sounded most like what the reviewer wanted to hear . . . after all, it was his daily reference! It was what he was used to, what he liked, what he’d assembled himself -- his reference.
Reference systems don’t have to be expensive. I could easily assemble a reference system based on, say, a pair of PSB Synchrony One speakers, which retail for $5000 USD per pair and sound quite neutral and transparent. The reference system of Doug Schneider, the SoundStage! Network’s publisher, includes a pair of Revel Ultima Salon2s ($22,000/pair). Check out the measurements of the Ultima Salon2s at www.SpeakerMeasurements.com and you’ll see a well-designed speaker that’s competitive with the best at any price. For years, my own reference speakers have been the Rockport Technologies Altair or Arrakis. Compared to everything else I’ve auditioned in my room, these have provided the most neutral and transparent sound. There’s more than one way to skin a cat.
At the end of the day, a reviewer’s reference system should be a reference: It should be accurate to both the ear and the microphone. When it is, then you, the reader, will learn more about the product tested, and have more accurate information with which to research and perhaps assemble a genuine reference system of your own.
. . . Jeff Fritz