Bullock Plays Bach – Part 2: Recording Bach

Most classical recordings are made in prime acoustic spaces like churches, recital halls, concert halls and the like. The conventional approach seems to be to set up two microphones (or one stereo microphone) more or less where a listener’s ears would be, the musicians do their thing, and what the mics hear in the room is what goes onto the CD. There’s typically no signal alteration after the fact such as equalization, compression, level adjustment or what have you, the idea being to capture the “authentic” sound as “purely” as possible. (Editing, though – lots and lots of editing, to fix mistakes or imperfections in the performance. That’s apparently not considered cheating, whereas all the other stuff is. More about that in a bit…)

And most non-classical recordings are made a very different way: in studios that are designed to be as acoustically “dead” as possible, soaking up all the echo and reverberation that one would hear in a church or a concert hall and leaving nothing but the initial sound produced by the instrument. This in theory gives the engineers and producers a sonic “blank slate” to sweeten any way they choose after the initial recording: through adding electronic reverb that can be completely controlled, equalizing (raising or lowering certain frequencies that aren’t in balance), compression (flattening out the dynamic range somewhat, so the loudest parts and the quietest parts aren’t quite so extreme), and even overdubbing additional instruments later on, taking advantage of multitrack recording and mixing capability. If what’s being recorded is an ensemble, from duo on up, the musicians are often separated from each other by thick foam baffles (or even in separate rooms) to keep instruments from “bleeding” into each others’ microphones, and the players hear each other through headphones. All of this, ironically, in the interest of creating as convincing an illusion as possible of actual performance in an acoustically friendly space – and even more ironically, I tend to be of the opinion that such an approach is usually more successful than the “purist,” straight-into-stereo-mics-in-a-concert-hall approach.

Well, I’ve done it both ways, many times, but for Bach Suites For Mandolin Volume One we didn’t do it either way…or rather, we did it somewhere between the two approaches, borrowing a little from each.


First of all, the entire project was a single instrument – suites composed for solo cello, being interpreted on solo mandolin – so there was no need for headphones, or more than two mics (one for each stereo channel). So far, not dissimilar to the purist classical approach: play normally, let the mics capture it, and Bob’s yer uncle. However: the space I chose for recording wasn’t a church, or a cathedral, or a concert hall – it was the downstairs room at Al Petteway and Amy White’s house on top of a mountain near Weaverville, North Carolina, where Al has his studio set up. Al and Amy and I have been the best of friends since forever and they’re two of my favorite musicians (and just generally all-around creative spirits), and among his many other talents Al is an expert recording engineer whom I’ve entrusted with many of my CD projects, so there was no question about where to record or with whom. Especially since a) I live only about half an hour away, b) there’s a beautiful view from up there, and c) they have the coolest dogs and cats that are big fun to play with. (Kip, their Australian shepherd, is the perfect studio dog: he was born deaf, as many Aussie shepherds are, and once he gets over the initial tail-wagging-jumping-up-and-down-oh-boy-oh-boy excitement of greeting a visitor, he settles down and goes to sleep on the studio floor – completely unfazed by take after take after take of the same piece of music.)

However, as relaxed a setting as that room is for recording, and it does have nice acoustics for an ordinary room in a house, it’s not acoustically resonant in the same way as a hall or church would be – so, we availed ourselves of the same sort of electronic sleight-of-hand that more conventional studios do: digital reverb added later to imitate the natural reverb of such a space, and equalization as necessary to make sure no frequencies were jumping out unnaturally. Once again, all in the interest of making it sound “real” and occurring in a beautiful acoustic space. The art of recording, after all, is ultimately the art of creating a convincing sonic illusion, and anything that allows us to make that illusion more convincing is fair game. (A philosophy I picked up from mastering engineer extraordinaire Bill Wolf, the other player on this project’s studio team.)

So, it’s all an illusion, yet it really is me playing, and that really is how it sounds when I play it – except it’s just right, because we can do that.


As it happened, I recorded Suite No. 1 a couple of years before Nos. 2 and 3 – other projects came up in between, and along the way I acquired several more vintage Gibson mandolins and decided to use a different one for each suite. I also learned, from the experience of recording the first suite, not to take on too much in one day or on successive days – solo Bach, even the easier movements (if there are any) is ungodly demanding on the hands and arms, especially if you’re recording many takes, and after pushing myself through long days recording the first four movements of Suite No. 1 my hands were tingling and my arms aching so badly I had to take a week off and get some massage on my arms, hands and shoulders to be able to finish the suite. (And here’s another example of using studio trickery to our advantage: by the time I was able to record the last two movements, the tone of the mandolin and the sound of the room had changed, infinitesimally but perceptibly – as they naturally do, with changes in air pressure, humidity and so forth – so we used a bit of EQ to make the last two movements tonally consistent with the first four.)

From that I learned: don’t try to do too much at once. Yet the six movements of a suite did have to be recorded as close together in time as possible, for the aforementioned reasons of sonic shift, so for each of the other two suites on Volume One, I booked six consecutive days with Al, one for each movement. Drove over each morning, a lovely scenic ride along the Blue Ridge Parkway to get me into a serene headspace, spent a couple of hours knocking out take after take of that day’s movement, called it a day, came back the next day and repeated the process. This turned out to be a wonderfully civilized approach; it allowed me to focus fully on that day’s particular movement, remain calm and relaxed, and not hurt myself physically through over-exertion.

Then I took my time listening to all the takes and deciding which were the best bits to splice together, again to create the illusion of a single, flowing performance, and Al and I reconvened when I was ready and made all the edits in another series of intense, but slightly less terrifying, sessions. (Before you fault me for cheating in such a way, just know that virtually every CD in your collection, at least after a certain point in about the 1960s, has plenty of this kind of “fixing.” The legendary pianist Glenn Gould, a pioneer of editing in classical recordings, maintained that the number of edits on a recording was of no more consequence than the number of revisions to a book or the number of stagehands backstage at a play, and I’m with him. Once more: it’s all about creating a convincing illusion when all is said and done. But can’t you hear the edits?, you ask…nope, not if it’s done right. With digital editing it’s possible to create crossfades as short as 1/100th of a second between different takes, way beyond what even the sharpest human ear can perceive.)

And there you have it. Nothing to it, really.


Tech notes, for those who care: Al used a matched pair of Sennheiser MKH40 microphones running through API preamps into a MOTU 828x thunderbolt interface, and finally into the computer using Digital Performer 8 software for recording and editing. The mics were in what’s called an “X-Y” placement (the mics crossing at a 90-degree angle, each at about a 45-degree angle to the front of the instrument), about two feet in front of the mandolin. Reverb was added during the mastering stage, and was a digital sample of the actual reverberation shape and rate of Santa Cecilia Auditorium in Rome – so in a way, what you hear on the CD is indeed the sound of me playing the suites in the gorgeous acoustics of that hall, even though I actually did it at Al and Amy’s place on the mountaintop, with Kip curled up at my feet.