Bravo!: the 1998 Santa Fe Chamber Music Festival CD Page 2

For the close mikes, environmental noise would not be a problem, as it would be suppressed in comparison with the direct sounds of the instruments and voice. The piano was a magnificent New York Steinway D, its soundboard autographed by all the pianists who had performed on it, from Vladimir Ashkenazy to András Schiff. I picked up its sound with an ORTF pair of cardioid mikes placed about 15" above the piano strings and looking at the width of the soundboard. The outputs of these microphones were panned in the mixdown to full right and 20% left, so that their piano image coincided with that produced by the main pair of microphones.

For the violins in the Elgar and the violin and viola in the Mozart, a B&K cardioid mike was placed for each in the Pinchas Zukerman position, looking down at the body of the instrument from about 4' above. An AKG dynamic mike was used for the viola in the Elgar, while the quirkily named Mitey Mike II—a tiny instrumentation omnidirectional microphone—was pressed into service for the cello, about 12" from that instrument's right-hand F-hole.

Very low-noise, transformerless solid-state preamplifiers from Millennia Media were used for the piano, violins, and viola, as well as for the main pickup in the Neikrug work. For Heidi's voice, the B&K omnis fed a tube preamp, again from Millennia—the Forssell M2a—while a Bryston solid-state preamp was used for the viola in the Elgar and the cello. Analog/digital conversion was provided by two dCS 24-bit two-channel converters, a Manley 20-bit two-channel converter, and the four 20-bit converters used in the Nagra-D recorder. All were word-clock-synchronized to the Manley at the CD's 44.1kHz rate, so that the digital data could all be properly aligned for the mixdown when it was uploaded to the digital audio workstation.

This meant that two disparate four-channel recorders could be used for the sessions: an open-reel Nagra-D and a Tascam DA-38. The latter uses Hi-8 videotapes and is normally an eight-track machine running at 16-bit resolution, but for this project it was converted to a four-track, 24-bit machine using the PrismSound MR-2024T "bit splitter." (For an upcoming Stereophile CD, I used a dCS 972 digital/digital converter in combination with the PrismSound box to use the Tascam as a two-channel, 24-bit recorder running at 88.2kHz sampling.)

Once the recorded performance had been assembled from the takes chosen by Marc Neikrug for the Elgar Quintet and Pueblo Children's Songs, and by Philip Traugott for the Mozart, my production philosophy was basically the same for all three works: mix the close-miked, individual sounds of the instruments and voice into the sonic picture as captured by the main pair, adjusting the level of each and its position in the soundstage to reconstruct what a listener in the "sweet spot" in the St. Francis Auditorium would have heard.

For the mixdown, the eight tracks, all at 20/24-bit resolution and 44.1kHz sample rate, were kept in the digital domain using a Sonic Solutions digital audio workstation with 25 gigabytes of hard drive. I used the Z-Systems rdp-1 to apply as much filtering of the LF noise as I felt subjectively acceptable when the distant microphone tracks were uploaded to hard disk. I also used the Sonic's DSP to synthesize a 9Hz notch filter to reduce the level of some subsonic ringing from the Shure mikes—even with wind screens, the flow of cold air from AC vents above the stage was causing some flapping of the mike diaphragms. If you have reflex speakers, you will see the woofer cones moving in response to the residual 9Hz signal.

The two-channel master sounded too dry if the distant mikes' contribution was low enough to keep the residual air-conditioning noise at a suitably low level. I therefore used a judicious amount of artificial reverberation from a Lexicon digital processor. Again, the signal was kept in the digital domain and the reverberation parameters were carefully adjusted to match the natural ambient sound of the hall, to help blend the sounds of the close-miked instruments with the balance as picked up by the main microphone pair.

Close mikes, multitrack recording, equalization, artificial reverberation—as I said above, this is a big philosophical change from the purist, documentary approach I prefer for classical recording. It is actually the way rock recordings are produced. But, given the changes in the hall's conditions and intrinsic sound, it was necessary.

For the Mozart Quartet, I chose a somewhat close balance, the brilliance of the sound matching the virtuosity of both the scoring and the playing. (My thanks to BMG's Philip Traugott for his invaluable help and suggestions regarding the balance in the Mozart.) The crystalline clarity of the sound of Marc Neikrug's Steinway is well suited to what might be regarded as a small-scale piano concerto. For the symphonic-scaled Elgar, I decided on more of a contribution from the main pair of omni mikes; hence, this work's soundstage is not so precisely defined as the Mozart's. The heavier-set, more Romantic nature of Elgar's orchestration benefits from the correspondingly lusher balance, with the downside that slightly more of the low-frequency rumble can occasionally be heard.

In between these two larger-scale works is placed the delicious miniature of the Pueblo Children's Songs, balanced at its natural level, with Heidi Grant Murphy's liquid soprano set against the tapestry of Marc's piano, its conventionally sounded notes punctuated by "stopped" notes.

Bravo! has been the most intensive of all of Stereophile's recording projects, but I believe the sonic results reflect the superb musicianship captured on those magical Santa Fe evenings in July and August 1998.—John Atkinson