boxing the butterflies

By Robbie on February 11, 2012

Jeremy Denk has a nice piece in the February 6 New Yorker on the trying process of making a record of one's performance: the preparing, the day of, the editing, the little agonies and challenges that arise at every turn. I don't think that much in his story will be very revelatory to anyone who has made a record or two, but I'd recommend it to anyone interested in the subject on the basis of its descriptive accuracy, couched in clear and simple language, e.g.:

"When the day set for the recording arrives, I lurch awake, thinking, Today I must play this piece as well as I ever have or ever will."

"The microphones and the piano face each other like enemies."

"[The engineer] nods and makes adjustments, which may or may not be adjustments. He's certainly going to humor me this early in the process."

"Already, before I've begun to play the first note, I'm tangled in compromise, in the homely realm of 'good enough.'"

"The most maddening paradox of recording is that what you hear in the playback does not resemble what you're sure you played. You hear two tracks at once: what you desire and what you have produced."

"It's supposed to be improvised and unpredictable, and we do it perhaps ten or fifteen times. How do you wring spontaneity from its opposite?"

"I want the most rapturous takes....but the most rapturous takes don't necessarily fit together...Unfortunately, in a fair number of the takes, a fair number of the notes went whirling off...we find ourselves trying to cobble together a note-perfect devil...There is the vanity of editing yourself to be a perfect version of yourself, but there's the vanity of loving your imperfections, too."

"How terrible it is that more and more performances aim to sound like recordings rather than the other way around."

I found this a refreshing change from the New Yorker's usual (to be fair, most national outlets' usual) arts coverage, in which a decent sentence is turned seemingly at the expense of very much correspondence with reality. Naturally, I prize a phrase tuned to vanishing exquisiteness as much as the next fellow, but as one who also enjoys the proverbial wit of the common folk, and is normally resigned to doing twice the legwork to satisfy both tastes, I appreciated the above-quoted matter in the way of an unbridled belly laugh during a Father Dowling mystery. There is something of Heisenbergian hopelessness in striving to deliberately contain the magic of musical performance on a hard drive or a reel of tape; is a menace cancerously housed in the microphonic ear; is a madness that begins a-rapping on the skull in the fifteenth take or the fourth hour of editing; and it's nice to hear someone put it out there so simply and sensibly. 

The typical listener, I imagine, doesn't put on a record or play an MP3 thinking of the music as an effort made under limiting circumstances, like watching a child sing in a talent contest. I'm not sure that he or she even thinks of it as a "performance" -- someone lurched awake that day, bought many bottles of water, drove to the studio, squeezed his eyes shut hard, and gave it his best shot before the microphone and God and everybody. I suppose in certain big-budget situations recording is a bit unlike that, but I think recorded music of all budgets and styles, electronica to baroque to blues, faces at some point these challenges of simulated spontaneity, unknowable aesthetic trade-offs between multiple versions, recording equipment altering outcomes, and so on. These challenges don't shrivel away under a hailstorm of resources (including the resource of intelligence). 

Nora O'Connor once remarked, "You know that thing where you never listen to your record after it comes out?" It's interesting that, for the world, a record's life begins when it goes on sale, but for the person that made it, its life ended much earlier, right after it came back from mastering. (So there's a longer, harder-to-remember way to say what Nora said, if anyone needs one.) You don't listen afterward for two reasons. One, it's too late to fix what's wrong with it. Very frustrating. Two, you've listened so much previously that you're utterly sick of it and couldn't hear it with impartial ears if you wanted to. There may be a nostalgic idea that it wasn't ever thus, that recorded music was once turned out with less effort; that, some generations back in the misty longago, for instance, carfuls of happy Negroes from Kansas City would stop in Chicago for a quick recording date before dinner, that sage folklorists with handheld devices roamed about catching unwitting and unspoiled hillmen in acts of heedless ecstasy, but I doubt all this very much. I think that musicians have long been skilled at making themselves suffer and fret (with some outside help). Sessions fraught with tough choices between lousy alternatives aren't exactly a product of the digital age. But it is inductively and somewhat empirically clear that novel challenges arose immediately with the advent of reproduced sound; and empirically it's very clear that technological advances have only made these problems more aggravating, many-sided, and labor-intensive.

I'm not suggesting that the listener pity the beleaguered artiste. I would suggest, though, that the listener who wants to approach music as something more than a passive child or an input-dependent robot consider that a dynamite record isn't the end-result of a process. It's more like the end-result of a particular episode in psychopathic self-immersion, in which some nutbag endeavored to find an ideal balance between technical prowess and his own crooked nature, and in doing so banged his head on many walls and almost certainly shortened his lifespan. If it were a process, it would have evolved in one-hundred-and-some years, and we'd all be enjoying better and better music, recorded faster and more cheaply. With that in the open, dear listener, feel free to resume regarding the music artist with the classic combo of envy and scorn.

Mr. Denk found himself a member of a creative duo in the studio, himself plus able and concerned producer/engineer. This alliance is not uncommon, and potentially very artistically effective. The artist knows the music most thoroughly and intimately, is emotionally entwined with it, has a faith-based view of his strengths (that is, his idea of his personal best as a performer is fundamentally insecure and fearful yet fundamentally immune to contrary evidence), and is additionally motivated by a strong financial stake in the results. The producer and/or engineer, whose stake is more modestly defined, maintains a similarly thorough but more clinical view of the music being made, so that as takes mount, as well as frustrations and distractions, emotion and faith are not allowed to trump sober judgment and the need to Get Something Done. With these two, Messrs. Right and Left Brain, as it were, working diligently toward the finished piece, each respecting the other and neither overdominating, good music has a fair shot against long odds. Undeniably, one hears a lot of music that evinces way too much sober judgment. Searching through it for faith and emotion takes you on a very short walk to nowhere. Methodism, maybe. But let's not get carried away with cheap cynicism: there's so much good music out there you couldn't listen in twenty lifetimes. This morning I heard records by Nat King Cole, the Carpenters, Kris Davis, Don Stover, James Taylor, and Blossom Dearie, a broad array of styles and eras. None of this shows signs of having been produced either programatically or by someone in the grip of a mystic inspiration. Not having been present at its creation, I obviously can't say for sure. But it's a good guess that the tension between the wild soul and the gimlet eye is essential to any recording that succeeds on this level.

Currently I'm going over about 12 hours of recorded takes from a session I did last November. Thus I found Mr. Denk's piece especially timely -- these problems of getting your best onto record are as impossible to get around as they are to solve. To a previous post where I talked about a hard in-studio choice between an ambitiously convoluted passage performed with audible difficulty and a simpler alternative performed cleanly, some commenters advised me against the "perfection" preference. Well, perfection isn't exactly ever the choice, and if it were, it wouldn't be as clear a call as suggested. Denk's formulation of the competing fallacies of editing-to-"perfection" versus loving your own flaws is spot-on. Because the artist resists any aesthetic dogmatism that would lead him to positions like "flawed passion over smooth technique," and because it's his or her own ego and personality he's presiding over, each of these decisions is a fresh and a troublesome one. Complicating things, my work pile consists of multiple versions of live, off-grid, group performances. The versions, though many are closely similar, evolve and mutate and shift in mood (and other deliberate attributes too, like vocal phrasing and instrumentation). I love this happy accident on take 3, but it's followed quickly by an unhappy accident. I like take 6 all the way through except for the first chorus, and the coda is a little off. Take 2 has crazy-great touches but an equal number of little disasters; take 7 is mistake-free and really pretty good (here we are in the "homely realm"!) but lacks the thing that makes you jump and smile. Now, to edit. See the problem? The problem is inherent in messing with past events like they were game pieces.

I'm noticing something -- small, not-too-surprising, but noteworthy -- in reviewing multiple versions chronologically. I mentioned that we performed with little-to-no playback. To use Denk's words, we had access only to what we were "sure we played." This saved much time and probably some rambling or possibly counterproductive discussion, but it opened the possibility that we might consider a tune finished before it really was. For some insurance against this, I generally enouraged a take beyond the one where the collective felt it well and truly nailed. (That happens under typical, playback-saturated sessions as well, though.) 

At the end of one day, one of the players said to me, "Do you always do this many takes on your records? Jesus Christ!" It hadn't occurred to me -- and I still don't think -- that 7 or 8 takes is very extreme. In the moment, I didn't feel it to be tiresome work. But it may have been: listening back to it is pretty tiresome, and at about the time my interest as a listener is starting to strain, I hear the performances starting to slacken, too. Well, the thing I thought was striking is this. Apart from a few outlier tunes -- stuff that felt, and was, magical on the first take, or that never got off the ground no matter how many iterations -- the performances follow the same progression. The first two or three passes, the song is finding itself, creakily. The next two or three are the meatiest performances, those with the highest highs and lowest lows and clearest intention. And the last two or three are the homely ones, the most clam-free but compromised by flagging spirits and, perhaps, a sense of absurdity at having to pretend you just thought of something terrific for the eighth time in the last hour.

Interesting -- the end-of-hour blahs are of course understandable, but why doesn't more magic happen the first time, as is spoken of in the training manuals? If everyone knows the song, its structure and emotion and trouble spots, and all are properly focused, why no higher incidence of first-take payoffs, or at least second-take? Instead, the work seems rather suspiciously to get done in the time alloted to do it. And interesting that, although our sense of when a take we had just played was good or bad was not very accurate, our broad sense of when all the takes together were sufficient to make a song was accurate -- pretty sharply accurate. With one exception, we never moved on until there was enough -- sometimes just enough -- good work across the takes. So in this limited way I think my experiment bore fruit: without any impartial overseer advising them, and without hearing what they've done played back, a good group can do its work effectively, guide a song into an evolved-to-completed state in a given time, by listening to and critiquing itself on the spot. How limited the worth of that is, I won't know until trying to do some editing.