The Fifth Element #7 Page 2

But there is—unquestionably—something there. If your system has high enough resolving power, you'll hear it, even from across the room. Once you hear it, you can't ignore it. Unlike Gould's humming, this artifact is not part of the gestalt of the original artistic event. So be careful what you pray for.

With the help of colleagues—one in particular—in the professional audio community, I have come up with a non-disprovable hypothesis that appears to explain this phenomenon. But first, an update on the system.

Increasing the Magnification
Siltech USA's Bob Dragounis, himself having written a college paper on Gould's Goldberg reconsiderations, was quite eager to join in the detective fun. He sent down a 1m pair of Siltech's Forbes Lake Signature Series interconnects. The various cables in Siltech's new ne plus ultra line are named after silver and gold mines, at least one of which one would need to have inherited to regard the Forbes Lake's $3500 price tag with anything approaching equanimity.

The next morning I was listening to some chansons from the early 1400s, and although I was not really paying attention to them, I suddenly had a clear mental image (and emotional reaction) of how I don't like it when choruses today make women whose voices are low enough to sing the tenor parts wear tuxedos. Telling Freud to take a hike, I took a Jungian approach (synchronicity) in trying to figure out why that thought had arisen unbidden. Turning my attention to the disc—which I have owned for at least 10 years—I suddenly realized that I was not hearing a fruity-voiced tenor but rather a slightly hooty contralto.

Always as a last resort, I consulted the liner notes. Bingo. With Siltech's Forbes Lake interconnect in the system, there was a new degree of resolving power, which worked on a subconscious level (footnote 2).

Adding $5000 worth of discrete runs (as distinct from biwiring with one cable) of Siltech G3 100 and G3 120 speaker cables added a slight degree of treble extension vis-à-vis Nordost's Red Dawn (a totally unfair comparison in terms of price), and removed a little whitish grain in the lower treble.

Given that nothing exceeds like excess, who was I to turn down US importer Philip O'Hanlon's offer of the loan of a $25,000 pair of Halcro dm58 monoblock power amplifiers? In they went.

Wowee. Zowee. "Amazing" was what my wife said.

So that I could fully utilize the Halcros' commanding reserves of power, the Goldbergs had to give way to Mahler, or at least Jamiroquai. And there was more than just raw power on tap: the reprise of the flute solo in the last movement of Jerry Bruck's recording of the Mahler-Wheeler Symphony 10 brought tears to my eyes. But even on the Goldbergs, the Halcros' speed, phase coherence, and tonal trueness unearthed more information, primarily room ambience (footnote 3). The Halcro amps sound musically authoritative (in that regard, they are among the top three in my experience), are remarkably uncolored but very liquid, and excel at re-creating a three-dimensional space. They catapult to the top of the Class A list in my book.

Case Closed?
Most of the explanations for the ghost-orchestra phenomenon that have been suggested don't hold up. CBS's 30th Street Studio was formerly a church. There was one big room, and no other place in which to record. (The studio closed in 1981. Gould's Goldberg Variations was the last music CBS recorded there.) So the "sound leaking in from another session" (modified Grassy Knoll) theory doesn't work.

Re-use of a used and erased master tape (Rosemary Woods I) doesn't make sense, either—the video deck would not sync up to an incompletely erased scan with a new scan on top of it, let alone recover music from underneath.

Re-use of a used and incompletely erased analog tape as an editing mule (Rosemary Woods II) looked good for a while, but a new pancake of ?" AGFA cost only $50 back then, the crosstalk is on only one channel, and I couldn't correlate the comings and goings of the ghost orchestra with likely musical edit points or, more important, Gould's humming, which displayed no jumps (footnote 4).

An informant who sometimes worked in that studio—in the same time frame, when it was also rented out to other labels—provided the most likely explanation. My source says that CBS's engineers back then were robustly immune to constructive criticism, to the point of disdainfully refusing to investigate the friendly heads-up that one channel of their eight-channel board was wired out of phase with the rest.

My source (who is not General Alexander Haig) tells me that CBS didn't like the engineers in the tape operations room to be "idle" when a session was going on, so they had to keep themselves dubbing safety copies of recently edited master tapes. My source does not know for sure, but suspects that use of patch bays at both the control-room and tape-ops ends (instead of going for direct, isolated connections) resulted in one channel of tape dubbing leaking into the adjacent channel of Gould live, so the crosstalk is on the original master. Case closed? (footnote 5)

Denon (not Telarc) was the first label to make and release LPs from all-digital sources, but for many years Denon always simultaneously backed up on analog, as that was the "best practice" for the time. So, my questions: Are there unedited Gould Goldberg II pure-analog safeties somewhere? Are they crosstalk-free?

Of course, with a nod in the direction of St. Teresa of Ávila, remember that 1981 was when whale products had just been banned but synthetic binder substitutes were not yet reliable. The analog tape stock might now be parchment, the music a handful of iron filings...

Look on my works, ye Mighty, and despair!
The lone and level sands stretch far away.

The important thing is that Gould's phrasing of the Aria—at the beginning and, even more so, so subdued in the end—is one of the most priceless artifacts of human civilization. One you should not be without. In either format.

Questions? Comments? Expressions of disbelief?

Footnote 2: This vignette proves to my satisfaction why "double-blind testing" is a delusion, at least when it comes to audio. Experiencing art is a matter of perception. Perception is the result of the action—both conscious and unconscious—of the mind on sensation. The narrowing-down of both sensation and cognition that is an unavoidable side effect of the double-blind process seems to me inevitably to disconnect us from the riches of the subconscious mind, which otherwise might be ready to help resolve or at least shed light on such issues—as I believe in this case it did. I can imagine some audiophiles—the irrationally rationalist ones I diagnosed last time out—breaking into Marcel Proust's kitchen and saying, "Now wait just a minute, sir, and put on this here blindfold. We have some other cookies we want you to smell, for our rigorous double-blind ABX test." Yeah, right.

Footnote 3: On the grounds of sound alone, I'd rather listen to Gould's Goldbergs on CD than to Murray Perahia's on SACD, the sound of which I find glassy and closed-in.

Footnote 4: Extensive, intense listening to this SACD destroys the fiction that all of Gould's studio recordings were built up from countless takes that were later micro-edited.

Footnote 5: I did ask Sony's publicist whether Sony could shed any light on the matter. As of this writing, they could not, and it would be unfair to whip Sony for CBS's sins of 20 years ago.