The Stereophile Test CD
"How do those new loudspeakers sound?"
"I don't know. Let me put a disc in the CD player and we'll listen."
"But how do you know what the CD should sound like?"
"I assume it's supposed to sound like the instruments and voices would sound in real life; the absolute sound, if you wish."
"But you don't know if that's appropriate. Recording engineers have an arsenal of special-effects boxes and signal processors at their disposal. And as you can see from Peter Mitchell's "Industry Update" column in this issue of Stereophile, even the microphones used at the very front of the recording chain can introduce significant coloration, let alone their effect on the soundstage."
"Rats. Let's just listen to it."
"But you don't know yet what the speakers sound like!"
A not unusual conversation, maybe even a familiar one. Yet as J. Gordon Holt points out in this month's "As We See It" ("Why Hi-Fi Experts Disagree," reprinted from a 1963 issue of this magazine), there is indeed a fundamental problem in using records whose intrinsic sound you don't know to make judgments concerning the sound quality of hi-fi components. Say you put a disc on and the sound is boomy in the upper bass. Unless you were present at the sessions and engineered the recording, you have no idea whether your system is accurately reproducing a boom that was there in the hall or was introduced by the microphones and recorder, or whether it is adding a boomy quality and is therefore inaccurate. It could also be a combination of both factors, exaggerating an existing boom.
The same problem exists when it comes to soundstaging. If your system reproduces a soundstage with distinctly unstable central imaging, is this due to an inaccurate reproduction of what should be a well-defined image, such as that produced by crossed figure-8 microphones, the so-called "Blumlein" technique? Or is it due to accurate reproduction of the soundstage produced by a spaced pair of omnidirectional mikes, a technique which, while excellent at capturing sounds with very little coloration, is notorious for producing vague, unstable imaging? Which came first, the chicken or the egg?
In practice, reviewers for this and other magazines get around this apparent impasse in two ways. One is to use recordings that they have made themselves and thus have a better idea about the intrinsic sound quality. The other is to play as many different recordings of as many different types of music as possible. If they all sound boomy in the upper bass, then the obvious culprit is the system. Occam's Razor in action!
But that's a time-consuming process. And wouldn't it be convenient for Stereophile's readers if they had access to our reviewers' recordings, so that they could check the reviewers' judgments for themselves? So that they had at least one CD where they knew that every step in its production had been carried out with care to ensure that the sound was altered as little as possible between the original and that encoded within its pits. So that they would then have a better handle on how their favorite components changed, or didn't change, sound quality, whether for better or worse.
We went some of the way toward this goal with the production of Stereophile's flute-and-piano LP, Poem. Just this one kind of music, however, will not be equally good at revealing all sonic problems. It will also not be to everyone's taste.
That's when the idea hit us. Put together a program of recordings made by Stereophile's contributors; add a track to demonstrate the kinds of colorations contributed by good professional microphones; add another track to demonstrate the audible difference between a state-of-the-art, 128x-oversampling A/D converter and the industry-standard Sony PCM1630; include a track to allow listeners to check how sensitive they are to absolute-phase inversion; and finish with some simple but useful test tones, ones that will help the magazine's readers to set up their systems. The advent of R-DAT means that the master tape could be inexpensively produced, while the advent of computer hard-disk editing of digital music data means that after the initial analog/digital conversion of any of the tracks, either from the microphone feeds or from the outputs of the analog recorder, there need not be any more quality-destroying conversions.
As you can gather, we got excited about the project, not the least because Stereophile's Technical Editor Robert Harley actually worked as a CD mastering engineer before joining the magazine and has considerable experience in assembling and editing CD master tapes. So let me hand this feature over to Bob, so that he can tell you how he turned our plans into a living, breathing silver disc.—John Atkinson