Duet: And Two to Carry Your Soul Away Page 4

Notes on the Engineering: John Atkinson

The more recording I do and the more microphones I acquire, the further I seem to move away from my purist roots. Once a diehard Blumlein devotee, who would countenance only the theoretically correct pure-amplitude stereo imaging proposed in the early '30s by the visionary English engineer Alan Dower Blumlein, I have been increasingly prepared to sacrifice imaging specificity in favor of a sound that is true to what you hear in the concert hall. What is the point of capturing the optimal imaging on a recording if the instrumental sounds are changed? Conversely, what point is there in capturing instrumental sounds with the utmost accuracy if the soundstage is destroyed in the process?

Given that the real microphones engineers have to use are all imperfect, then why not compromise both aspects of reproduced sound equally? Choose mikes and a way of using them that captures an image that, while not being as precisely defined as that offered by a Blumlein-stereo recording, still sounds acceptably real, while preserving the sounds with as little coloration as possible.

You can follow this personal voyage on Stereophile CDs: my Chopin recording on Test CD 1 (STPH002-2), captured with the Soundfield mike in crossed-figure-8 mode, exemplifies the purist imaging approach; the Concert piano recording (STPH005-2), recorded with two widely spaced omnidirectional mikes, represents the polar opposite. All the more recent CDs, from Festival (STPH007-2) through Encore (STPH011-2), have been made using a central ORTF pair of cardioid microphones and a pair of spaced omni mikes, my goal being to combine the imaging precision of the cardioids with the low coloration and extended bass of the omnis, and to preserve some of the bloom typical of the live sound.

I feel we have been successful in arriving at this goal, but while preparing for the Duet sessions, I was reminded of something the legendary English engineer Tony Faulkner had told me well over a decade ago when I was producing the first Hi-Fi News & Record Review Test CD: Spacing a pair of omni mikes about 27" apart seemed to produce a magic combination of imaging and sound quality. In fact, Tony had used this array as the main pickup for his Handel Entrance of the Queen of Sheba cut on the HFN/RR Test CD. And, as Audio Engineering Associates makes a 27" version of their short stereo bar I had already used for Rhapsody and Encore, I bought it for these sessions.

A couple of test recordings revealed that the imaging was still somewhat phasey with omni mikes spaced 27" apart—though the superior midrange tonality and the extended low frequencies of the B&K 4006 omni were both readily apparent. So I decided to add a central, forward-facing B&K 4011 cardioid mike, to angle the B&K omnis out by about 20 degrees, and to use the latter mikes' nose cones. The cone mounts in front of the electret diapharagm; it makes the mike's response flat to off-axis sources in a diffuse soundfield, and gives it an exaggerated top end for sources directly in front of it.

With the benefit of 20/20 hindsight, I realized that I had assembled a version of Decca's famous "tree," with which Kenneth Wilkinson had made so many great-sounding recordings in the '50s and '60s. The Decca tree featured the legendary Neumann M50 omnis, which were reasonably directional in the highs, whereas I had used B&K mikes. At least, at around two grand apiece, the B&Ks are affordable—a mint-condition Neumann M50 will set you back major bucks!

As the mike array used three of the Nagra-D's four tracks, I added a second central B&K 4011 cardioid facing to the rear of the hall and angled so that the violin was in the mike's null. I could derive the rear-channel feed from this mike's output if we ever decided to issue a surround-sound version of the CD. I could also fine-tune the recording's direct/reflected sound ratio by mixing in a little of this track in post-production.