This particular recording project started a couple of years ago when I suggested to Ian, organist for the Bell'Arte Singers and resident organist at Knox Presbyterian Church where A Childe Is Born (AI-CD-003) was recorded, that we might undertake an organ project in his home church. My idea was to do it in surround sound, having been playing with various recording techniques to get true matrix front-to-rear separation (and rear left-to-rear right, the major failing of Dolby Surround) without needing the electronics of an encoder and the attendant distortion and unwanted extra phase shifts.

      I suggested to Ian that he think about program material that was not the kind of showoff stuff that makes it onto every other organ LP or CD, but unusual pieces that show the coloristic capabilities of the Knox organ, and added the audio instruction that at least some of these use the trumpets at the rear of the church (see photo), and also have enough deep bass pedal to maintain audiophile credibility.

Technicalities

      I've been interested in acoustics most of my life, and have been directly involved in surround sound in its various historical forms for longer than I've been writing about audio. In fact, my writing career began in 1972 when I was in the final years of managing the Queen's University radio station CFRC-AM/FM. Because I was doing some freelance writing and music broadcasting for CBC and Toronto's CJRT-FM, I caught the attention of FM Guide publisher Forbes Calder, who asked me to write about the quadraphonic craze that was taking over audio. The result was a column called QuadrAnswers, which grew into regular product reviews, and by 1973 the Audio Ideas radio show on CKFM. Next came the move to Toronto, since most of the work was there, and more broadcasting with CBC, including a series called In The Middle Of The Music, which was all quad recordings, including BBC transcriptions in SQ, and lots of LPs in SQ and QS.

      I had plenty of equipment to play with and review, as now (I remember in particular monster receivers from Sansui and Marantz). But I'd also been doing quite a bit of recording, this increased by the provision by CBS Technology Center in the U.S. of a Sony (they had close connections even then) SQ Encoder for matrixing discrete tapes and discs (CD-4); incorporating 4 channels of mike preamps, as well as line inputs, it ended up being used for a lot of recording of SQ open reel tapes, including concerts by the Canadian Electronic Ensemble, various choirs and organs, and numerous chamber music ensembles. Some of this was for broadcast, some for fun, and quite a few became LPs, though not usually identified as SQ because by the mid-70s quad had become synonymous with corporate infighting and acoustic distortion, killed in the market by the former, and rejected by the high end because of the latter.

Sniffing Around Surround

      As a comment on the reaction of high end audiophiles, I was amazed at all the nose-sniffing at multiple channels, because I had come to feel that spatial presentation was the major failing of purist audio reproduction, and sought out better sounding surround decoders, eventually getting from England the one I still use, the Integrex Ambisonic/SQ/QS/Hafler decoder designed by the Ambisonic developers, led by the late Michael Gerzon and Peter Craven. With very precise matrix phase shifts for rear channels, and no logic steering, it provided spatiality with limited separation, but very low distortion.

      This led me to think that it might be possible to achieve the counterpart of this in the record process, a fully acoustic matrix dependent only on the natural spatial acoustic heard by the microphones from a central point and the adding/subtraction process of the 4 signals following the mike preamps. This would result in a naturally derived encoded surround signal that would decode with any Hafler or similar system.

      Experiments through the early 90s resulted in what has been employed in this recording, utilizing a pair of 90o-angled AKG 460ULS cardioid mikes facing frontward, and a rear-facing AKG C522ENG stereo mike with identical capsules inside, also at 90o angles, but very close together. This way the average of the front and rear capsule spacing laterally is about the width of a human head, and if you shot a bullet through the front capsule right down the barrel of the microphone it would also pass right through the rear-faced capsule of that side, either left or right. For those familiar with coherent miking systems, this setup is quite similar to crossed figure-of-eight, or Blumlein, with the difference that the signal of each rear channel is maintained on its correct side, and can therefore be properly added and subtracted with its front channel to derive the front and rear spatial information. This is my approach to a coherent matrix array, with the rear microphone channels reversed in polarity (180o) at the microphone preamplifiers to provide an in-phase centre and resultant correct phase shifts for derivation of 4 matrix channels. These were derived at the output of the pair of Bryston BMP-2 stereo mike preamps before Meridian 607 A/D conversion. The result is what we call NaturalMatrix, an acoustically encoded surround recording that will provide separate left and rear channels on any non-DSP matrix decoding system, from Hafler to Hall mode.

Getting Into (The) Gear

      Our microphone cables were generously supplied by Kimber Kable, Ray Kimber himself an organist, and are balanced KCAG Silver in a 50-foot 4-microphone snake. I don't think I have to say any more than that this is the most transparent microphone cable I've ever heard, and will be a continuing fixture with Audio Ideas Recordings. Its resolution is extraordinary.

      Having the microphones, their deployment, preamplification, and matrixing in hand, I looked at record media and monitoring. Having been very impressed by the PMC LB1 small professional monitors as a Hi-Fi speaker, and especially knocked out by their deep bass performance, I settled on these as monitors for playbacks, and while Ian was performing, used my trusty Grado SR125 headphone, which also has outstanding bass, as well as a very speaker-like tonal balance.

      This left recording media to sort out, and I considered using something like a TASCAM DA-88 to allow later mixdown to NaturalMatrix, but rejected this on grounds of purity, wanting the surround to be directly recorded on site. Curious about the arguments touting even 15 IPS 2-track analog as superior to 16-bit/44.1-kHz digital, I decided to employ my trusty old TEAC X-2000M recorder as a backup; it had served well as an analog stage for blade-editing A Childe Is Born, and if the sound was more than acceptable there, then with some really good tape, it should provide a pretty good analog recording. Given the blower noise of the organ, the noise floor was not an issue, so its dbx noise reduction was not engaged. Tape was Maxell 50-120B, one of the higher headroom tapes out there, and bias was calibrated for it very precisely.

      It may seem odd to use a TEAC DA-P20 DAT portable as the digital medium, but one must remember that it was simply a recording transport, fed a high resolution digital signal from the Meridian 607, which claims dithered resolution to below 120 dB, and is certainly the best 16-bit processor I know of. Clive Allen and I have been experimenting with digital media for years. The Debussy Preludes was recorded on Super VHS tape using the 607, but he had discovered data-grade DAT tapes to be as good in resolution and freedom from dropouts, so that's what I settled on here.

Getting Organ-ized

      To sum up this technical part of things, let me note that the contest between the analog and digital tapes was brief, and not even close. The clearly audible low frequency harmonic distortion, high frequency tizz, and midrange congestion of the analog tape (and this was without pushing the levels) made the choice easy, and the analog machine was left home for the subsequent two sessions. The data grade Maxell DATs had a clarity and openness that suggested very good data integrity and freedom from dropouts. I had planned to add a repeat analog comparison track at the CD's end, but Ian wasn't too keen on this for musical reasons, and I ultimately decided that it was a question that had been already answered conclusively. I still have an edited 15 IPS tape from the first session that sounds very good, a little warmer and fuzzier than the digital. I also have a CD-R master with the extra track; that disc ended up being a safety.

      Editing was done in the Sonic Solutions system (now all AES/EBU) at Clive Allen's Georgetown, Ontario studio, and didn't take long because Ian is such a great player, and knew the pieces very well. The majority of the tracks on the CD have no internal edits, and most of the studio time was devoted to between-track edits; since each work had different organ registration, the tonal signature of the blower noise changed when going from one to the next, and we tried, sometimes not too successfully, to blend these transitions. You will hear these changes between tracks, especially on headphone. The only failing of this wonderful instrument is its fairly high noise, and our aim was to try to make it as unobtrusive as possible.

Musical Musings

      Ian's program, The Sadler Selection, ranges widely from the more familiar, an organ transcription of Elgar's Nimrod from the Enigma Variations, to the utter pedal power of the Karg-Elert Pax Vobiscum and the Variations de Concert by Joseph Bonnet, to charming theatre organ pieces by de Severac and Bourgeois. It's about as varied as you could imagine, almost 64 minutes.

      I've listened to the CD-R safety on a wide variety of matrix systems, on CD, LD and DVD players, with various surround processors. They don't all sound the same, but they do sound very musical and enveloping, which is what you'd hear anywhere in the large space of Knox. I think it's a fun disc for audiophiles, music lovers, and, especially, for organ aficionados.

Andrew Marshall