Intermezzo: The Santa Barbara Sessions Page 2

While the rest of the crew set up the microphones and piano, I took the opportunity to check out the environment. Measured by means of my calibrated 28" pace, it turned out to be roughly 55' deep by 35' wide. As I claim the talents of neither Fred Astaire nor Jeff Goldblum, I had to estimate the height; it appeared to be a bit greater than the width—say about 40'. A second-level choir loft in the rear extended back over the entry, increasing the depth somewhat in that area. And an alcove in the front (estimated to be about 20' by 20') opened up the space somewhat further. The piano was installed in front of this alcove—we had to remove the first two rows of pews to make room for it. The interior walls, which appeared at a casual glance to be of solid stone or adobe, were actually made of the ubiquitous Southwestern stucco; the hollow sound audible on rapping these walls indicated a probable wood-frame construction. The acoustics of the church were highly reverberant; Kavi spread the long, soft pew cushions across the tops of the first few rows of pews to help control the echo, but it was clear that the microphone setup would have to be carefully controlled.

intermezzo2.jpgThere is no question that the key to a natural-sounding recording is in the selection and positioning of the microphones—even more so in such a reflective environment. The former was a given; Tim de Paravicini's tube microphones have proven their worth in the past—JA was so impressed by their performance during the sessions for the magazine's first album, Poem, in 1989 that Stereophile purchased its own pair, which we brought along as backups to Kavi's. Placement, on the other hand, is a process of trial and error. Kavi was very familiar with the recording site, so had a good feel for where to start, but minor adjustments were inevitable. So while our soloist, Robert Silverman, warmed up on the unfamiliar piano, and while we awaited the imminent arrival of the piano tuner, Kavi tweaked the setup until JA was satisfied that we had captured the appropriate ambience of the recording space. Because of the high reverberation of the church, the microphones were closer than you might expect—about 8' from the front of the instrument. The microphone configuration was the classic crossed figure-8s, Blumlein-style, with a slight vertical distance—about 2"—introduced between the capsules which Kavi felt would "open up" the image a little.

By 8:00pm the church was a beehive of activity, none of it, unfortunately, involving actual recording. Kavi was still making minor mike adjustments, and the "control" room itself, a small service room about 20' from the main auditorium, was calmly chaotic. JA was monitoring the results of Kavi's placement efforts. Bill Reed was aligning the Ampex. Yours truly was adjusting the Nakamichi (footnote 2). JA had set up our "PA" system—a single Cambridge SoundWorks Ambiance loudspeaker driven by a Creek amplifier—to communicate with the soloist. At the loudspeaker end of this chain, Bob Silverman was working with our piano technician, Kirk Taylor. LA decided that a repast was in order (the evening sessions did have one major disadvantage—they eliminated the normal dinner hour) and went out to bring back sandwiches.

By 9:30pm, suitably fortified, we returned to the evening's problems, which, it turned out, had just begun. Bob Silverman was not entirely happy with the "feel" of the piano, but more significantly, engaging the soft pedal resulted in a small but clearly audible squeak. Bob proposed that we temporarily lock-in the soft pedal for that evening's session, record the quieter portions of the Intermezzo, then attempt to fix the squeak the next day. A not entirely satisfactory prospect, but perhaps our only chance to get anything "in the can" tonight.

But some fellow named Murphy was not to be so easily bought off. About this time we became a bit concerned that there was too much hiss coming from the tape. Should we mike closer and lose a bit of the hall ambience, or go to the quieter Milab microphones (used on one selection for Stereophile's first recording) and lose transparency by abandoning tube mikes? JA was determined to use the latter; so began a series of experiments. First, we replaced one of Kavi's mikes with one of Stereophile's pair to determine if ours might be quieter (footnote 3). Now we noticed a substantial level difference between channels—a good 8dB. What had happened? Both of Stereophile's mikes were now installed to determine if they differed in gain from Kavi's slightly older pair. The level difference remained. It was eventually tracked down to a problem with a level control in Kavi's preamp. By this time it was 11:05pm.

Footnote 2: This particular model was designed, in my opinion, the way a cassette deck should be—with accessible front-panel adjustments for bias and Dolby level for each channel, and separate controls for standard, high-bias, and metal tape—12 adjustments in all. And all of the potentiometers are recessed, so once set they can't easily be disturbed. A tweaker's delight.—TJN

Footnote 3: As pointed out by LA in his "Final Word" for March 1990 (Vol.13 No.3), these were not quite identical to the pair used in our first recording project; they are now based on the EF86 tube. Kavi's prototype mikes and Stereophile's pair of production mikes are built around the 6DJ8.—TJN