SOUNDSTAGE! ON HIFIFeatures Archives

January 1, 2001


A Conversation with John Atkinson, Recording Engineer

John Atkinson, JA to over 80,000 readers, is the editor of Stereophile magazine and a man with a lot to say about audio. But long before Atkinson was an audiophile, he was an avid documentary recordist. Some of his earliest recordings can be found on The HFN/RR Test CD 1, as well as on each of the three Stereophile Test Discs. The three Stereophile discs are excellent tutorials for anyone interested in why recordings sound the way they do -- they feature the same material recorded with different microphones and different microphone techniques. Atkinson has also engineered a superb series of recordings for the Stereophile label, several of which I had the distinct pleasure of assisting him on. (All of the Stereophile compact discs are available at ) Recently, Atkinson recorded pianist Robert Silverman's complete traversal of Beethoven's piano sonatas for the Orpheum Masters label, a recording that I believe is one of the treasures of the year 2000, due in no small part to the full-bodied dynamic recording acoustic Atkinson achieved -- against all odds, as he explains in the free-wheeling discussion below.

WP: You're best known as the editor of Stereophile, and before that HFN/RR, and prior to that you were a professional musician. Where did the desire to be a recording engineer come from?

JA: I got into high-end audio because I was recording the bands I was playing in, and I had an interest in preserving what we were doing on tape. I was extraordinarily naïve -- remember, this was 1964 or 1965 -- so it came as a great shock when I tried different microphones that I got very different sounding recordings, even on analog mono recordings at 3¾ ips! The hot vocal mic for bands back in the Sixties was the Shure SM57 cardioid and I actually made some pretty good recordings with a pair of those. I was very surprised to find that it sounded so different from the Reslo Ribbon, which was also a popular mic back then.

John Atkinson and Wes Phillips flank jazz musician Jerome Harris at the Rendezvous recording sessions.

WP: Did your priorities change when you got a really good hi-fi?

JA: Not so much. I carried on doing documentary recording throughout the early Seventies. The band I was in at the time was using a lot of what was then very new multitrack recording technology -- the very first album I was involved in was recorded on an eight-track, and then the album after that was recorded at Abbey Road studios on 16-tracks, and then the rhythm section of that band became the house-rhythm-section for a studio in Cornwall, which at first had a 16-track 3M machine and later a 24-track recorder.

So, coming from documentary recording, I then spent the early part to mid-part of the Seventies doing the kind of recording I now find somewhat un-musical, which is layers of multitrack -- where you'd end up with something that was technically perfect and musically sterile. I didn't enjoy that kind of recording, so I think that's why I went back to documentary recording, where you labor to preserve real music, happening in real space as accurately as possible.

WP: The engineer at that Cornwall studio has come a long way, too.

JA: Yes, the Sawmill studio at Cornwall, England is famous now because bands like Oasis record there, but at the time it was very much a small-scale operation. The chief engineer at that studio was Jerry Boys, who shot to fame as the engineer for Buena Vista Social Club and quite a few other recordings that are well-regarded for their sound quality.

WP: Did your renewed embrace of documentary recording owe everything to your experiences with multi-tracking?

JA: I joined Hi-Fi News & Record Review in 1976 and the editor at that time, John Crabbe, was a total devotee to the Blumlein philosophy of recording, which is basically a two-channel recording where the amplitude information in the two channels very accurately reflects the directions of all the sound sources. There's only one theoretically correct way of doing that, which is to use coincident-crossed figure-eight mics and the downside of that is, yes you get great stereo imaging, but those mics themselves generally have poor low-frequency response and there's also a real problem in finding the best spot in the hall to get what is both the right stereo image and also the right musical balance between the instruments. If you have unlimited time, you can actually get to that spot. But -- in the real world -- you might sometimes have only minutes to set up your mics, as in some of the live recordings you and I did for the Santa Fe Chamber Music Festival.

I started out as a purist and, ideally, I would still make recordings in that way, if I always had a good hall to record in. Without a good hall, you have to use somewhat compromised mic positions regarding the accuracy of the stereo imaging, which will then give you better capture of the low frequencies and better feeling of the envelopment of the space in which the music is being performed.

WP: Serious recording technology used to be the realm of only a handful of well-financed engineers, but that's changed over time, hasn't it?

JA: Audiophiles are going to hate to read this, but it was the advent of relatively low-cost digital technology which revolutionized recoding. Analog tape recording is an art and I was never very good at it. I'm all thumbs when it comes to editing with razor blades. I still have an Ampex ATR half-inch tape machine in storage, but I don't ever use it.

If we go back to my philosophy of documentary recording, the more I can do to put off quality-making decisions to the comfort of my own listening room, the better. Then all I focus on in the recording is the placement of mics -- all other things, such as mixing, changing the time alignment of the mics, anything like that, I'd much prefer to do under low-stress conditions back home, and digital technology has been a boon to that.

In digital, the actual sessions themselves are like data-gathering, that's all. My whole focus is capturing those sounds with the mics I have chosen to use and laying them down on digital tape. Then I come back home with all the digital tapes and that's where the creative process begins. Digital technology has made that so easy.

WP: So the editing is where the art resides?

JA: Yes -- and, of course, it helps if you have a flexible digital audio work station in the form of a Sonic Solutions Sonic Studio. One of the things I do now, almost by habit, is to have two, or even more, pairs of microphones up to record the event. One pair is optimized for stereo imaging; these days I tend to use a pair of ORTF cardioids, that is two cardioid microphones with their capsules angled at about 115 degrees and spaced about seven inches apart, which gives you a relatively limited amplitude-defined sound stage, but, if your mics are seven inches apart, you get a .6 or .7 millisecond time delay between the two channels, which gives you a time-defined stereo image. That reinforces the amplitude-defined stereo image and the combination actually works very well.

A lot of Gordon Holt's classic recordings were made in exactly that manner -- some of these are on the first Stereophile Test CD. My problem with cardioid mics is that they tend not to have enough bass -- they give a beautifully defined stereo image that tends to be rather lightweight. So now, I always have a pair of spaced omnis up, which to a purist -- as I used to be -- is almost anathema because you get almost no good stereo image. However, you do get great low frequencies and a very natural midrange tonality. So my basic recording technique now is to capture four channels with those two pairs of mics and then, when I'm back home in my listening room, I can slide those pairs of tracks around in time so I can get the well-defined stereo imaging from the central pair and the full-bass, excellent midrange tonality from the spaced omnis.

My starting point in editing is always to align the tracks in time with one another. To facilitate that, I always have an assistant (or the musician) create a synch pulse when we begin recording, using a slap-stick. This gives me a time reference similar to the clapper board on a movie set. What's interesting is that what is theoretically correct time alignment may not be what is on the final mixdown.

Recording Robert Silverman's Beethoven cycle -- note how close to the walls the piano is located.

Take the Robert Silverman Beethoven set, for example, which you wrote about for several weeks ago. That was a very unsuitable hall. Given my druthers, I'd have preferred to have done the recording somewhere else, but you have to sometimes roll with the inevitable when it comes to recording venues. When I aligned all the mics so they were theoretically perfect, you were just so aware that the room was too small, too live, rather unsuitable for such a large, powerful piano. So I ended up ultimately delaying the omnis by several milliseconds, which made the acoustic more anonymous, but also more comfortable sounding, in that the early reflections of the sounds from the walls (see photo) were smeared over. That's a move away from documentary recording, but I think, in this case totally justified, because the hall was not optimal.

That's what I did for the 44.1kHz/16-bit recording for the CD. What's interesting is that I originally did the recording at 88.2kHz/24-bit, and at the very high sampling rate, the theoretically correct coincidence of the mics was acceptable because it was so unambiguous. The character of the hall was preserved with such accuracy that you could adapt to it and forget it -- just as you would in real life. When I downsampled the recordings to 44.1kHz, I also lost a quite a lot of that precise definition of space and time so you couldn't adapt to it -- it became annoying.

WP: What does that say to you about the future of hi-rez digital audio?

JA: I'm totally sold on it, myself, because there's an ease to the sound, as there is in real life, which enables you to adjust to what you're hearing and then forget about the acoustics and get into the music. Whether we'll all be experiencing high-resolution digital through DSD, as used in SACD, or high-speed linear PCM as used on DVD-Audio, for me, the jury's still out on which is better.

WP: I was reading your article on recording the Beethoven project in January's Stereophile, where you describe how you arrive to record over eleven hours worth of music in two-and-a-half days and discover that the piano is in a tiny little box of a room and that it can't be moved even six inches. What did you do?

JA: I put my head in my hands. We were under such time constraints. But the beauty of the situation was that we were using the Bösendorfer reproducing piano instead of a not-always-cooperative musician. So I got Jim Turner, the producer, to set up one of the most dynamic passages in one of the sonatas and just keep it in "repeat play" while I did a lot of listening to the piano from different places in the hall.

I worked out what I was hearing and determined good starting places for the mics. As coincidence might have it, the piano technician, Mike Kemper, came in to work on the piano and told me that I had the mics more or less where NPR's engineers had placed theirs when they'd recorded the piano for Performance Today. I started out with some of the Neumann large diaphragm mics, and I thought they might reach deeper into the piano sound. They did, in fact, but they were also more colored than my B&K 4011s, my old trusty mics. I went back to the B&Ks and did some test recordings. (B&K the measurement company sold off the microphone division and it's now known as DPA (Danish Precision Audio.))

We then made test recordings, moving the cardioids about a total of four inches further in on a line towards the piano. I thought the position closest in actually revealed the best balance between the sound of the hall and the sound of the piano. However, we then picked up a lot of piano action noise. Now if you think about it, a piano is actually a percussion instrument; you have a felt mallet hitting the string and it actually produces quite a thump. So I backed off about three inches from that position, which gave me more hall than I would have liked, but also lowered that thump to the point where we all felt it was acceptable.

I got there at 1:00PM and had all my gear unloaded by 2:00PM and we were scheduled to roll tape for 6:00PM, so you can imagine it was all pretty sweaty.

John Atkinson recording Musical Fidelity's Anthony Michealson at Blue Heaven Studio. The disc will be released in May, 2000.

WP: Ah, the glamour of recording! But, as they say, the height of art is to conceal art and the recordings certainly don't sound sweaty. I guess we should tell people how to order the set.

JA: We've just put it up on our Website, so you can order it at, where my article on recording is posted, and there's also an order form posted on the secure recordings page. You can also order it directly from Orpheum Masters.

WP: A ten CD set for only $65.00 -- that's very reasonably priced.

JA: Yes -- and it includes a comprehensive booklet on Robert Silverman's attitude toward performing the sonatas that's worth reading on its own.

WP: Yes, I regret not mentioning it in my review here -- it's a wonderful piece of writing and thinking.

JA: I have to say that, having been totally immersed in these sonatas as performed by Bob and having heard him live at HI-FI'98, he's a consummate performer of these works. He really has a passion and a fire that perhaps is out of fashion in Beethoven performance these days.

WP: People will get a chance to hear Bob perform Beethoven sonatas at Home Entertainment 2001 in New York this spring won't they?

JA: Yes, we've scheduled Bob to perform several of them on Sunday afternoon, May 13 at the Hilton Hotel on 55th Street.

WP: I'll be there!

SOUNDSTAGE! ON HIFIAll Contents Copyright © 2001
Schneider Publishing Inc., All Rights Reserved
Any reproduction of content on
this site without permission is strictly forbidden.