Intermezzo: The Santa Barbara Sessions JA on the sound

Sidebar 4: JA on the sound

I'm fascinated by the sound of the recorded piano. While the instrument under the hands of the virtuoso—the Baldwin and Earl Wild, the Steinway and Alfred Brendel, the Bösendorfer and Ivan Moravec—is subservient to the music, there is a thread to be found running through these sounds that is pure "piano," the instrument imposing its own voice. And not only does every instrument have an identifiable sonic personality; every recording session is a unique occasion with a unique acoustic having a unique effect on the instrument's sound. If you have listened to any of Liszt's piano transcriptions of the Beethoven symphonies, you will be familiar with the way the restricted tone color more readily lays the music open for inspection. It is said that whereas English has but one word, "snow," the Inuit have 50. (Having lived in New Mexico for more than four years now, I wouldn't be at all surprised to learn that the local Spanish dialect has 50 words for the color brown.) Similarly, though piano sound may be monochromatic, that very restriction paradoxically allows a multidimensionality of musical color to be perceived.

With previous piano recordings, such as those on the HFN/RR and Stereophile Test CDs, I have gone for rather a distant image, the piano being set well back behind the plane of the loudspeakers, with a strong ambient content to the sound. For this Brahms set, however, Kavi Alexander had wanted to capture a bolder, more powerful piano sound, particularly as the instrument was exciting the church acoustic to a rather excessive degree. The microphone array was placed quite close, about 8' back just below the line of the lid. The image of the Steinway should therefore be just behind the plane of the speakers, extending from the inside edge of the lefthand speaker, which is where Robert is sitting, to the inside edge of the right. The use of a Blumlein technique means that the image should be "solid," though I confess there is a slight pulling to the sides noticeable with loudspeakers possessing sufficient resolving power.

Regarding the recording's tonal quality, those used to the typically bright sound of "commercial" piano recordings will find its treble to be a little soft. We are happy, however, that this accurately represents the true sound of the instrument in the hall at the microphone position. The complex manner in which notes decay, their harmonics intertwining, should be readily apparent, while the midrange should be evenly balanced, with no notes "jumping" forward at the listener. (Both Robert and the piano technician spent a considerable degree of time "voicing" various notes—pricking the felt pads of the hammers with a pin—so that they bore the correct dynamic relation to those on either side.) In the bass, there should be a strong feeling of weight to low frequencies, the Steinway "D" being a majestic instrument. Low bass notes should have some of the quality of a deeply tolling bell.

I think the dynamics of the instrument's sound were well captured: in the words of Alfred Brendel, its volume ranges "from a whisper to a roar." Kavi hit the ½" tape hard, reaching 14dB over the 250nWb/m 0VU mark, meaning that tape hiss was nonexistent even without noise reduction. The limiting factor affecting the recording's dynamic range is actually the intrinsic noise of the microphones rather than that of the analog tape; at a realistic playback level, this can be noticed as a soft hiss. But note that I said "realistic" playback level. This should reach, at the most, 96dB at the listening seat for the very loudest passages, such as the opening of the sonata's first movement. Not surprisingly, if you turn up the playback level so that what should be quiet passages become more "impressive"—the wonderful Debussy-esque musing at measure 19 of the sonata's Intermezzo, for example—then the microphone noise will probably be intrusive.

Don't play this recording too loud.

When Hugh Davies and I met at Doug Sax's Mastering Lab in Hollywood to assemble the master tape from the eight hours of session tapes, we tried to go for long takes to preserve, as much as possible, Robert's musical ebb and flow. Splices when necessary were made at musically natural points—I abhor the practice of arranging things so that the first half of a short phrase comes from one take and the phrase's conclusion from another. We didn't worry overmuch about inserting a measure here or a bar there to eliminate a natural noise. (Though when traffic noise intruded, we scrapped the entire take.) As a result, you can occasionally hear an extraneous noise, either the piano stool creaking or (rarely) the mechanism of the piano itself. Or even Robert Silverman's crooning baritone. As these are "honest" noises which would be heard in concert, we felt justified in leaving them in.

Enjoy.—John Atkinson

X