D-2 Mock-Up

Several months ago, it occurred to me to try using pitch shift effects to obtain notes that are out of my range on tuba. Here's my first concerted effort at it:

The "D" Series: D-2

(score here)

If you've been to a C.o.S.T. performance, you've heard this piece before; it was one of the first flexibly scored pieces I wrote with that group in mind, and so far we've played it at every show. On top of the many things that intrigue me about flexible scoring, there's the fact that by using pitch shift, I can record the entire piece all by myself (I did use Sibelius for the snare drum part, but could always have a percussionist friend overdub it later). Obviously, sound quality suffers more as you move the pitch further, but all in all, it's a far more human way to hear a piece than MIDI playback through a notation program.

I played as much of the music in the written octave as I could, and played many of the higher parts down a fourth or fifth rather than an octave so as to minimize the total amount of pitch shifting necessary. This means that most of the middle parts do not consistently use the same timbre, but in this format, it doesn't matter much (long live flexible scoring). The biggest challenge is that pitch shifting up magnifies tremendously even the slightest flaws in articulation, and I sometimes found myself trying out a few different transpositions before finding one where the right notes responded the right way.

After pitch shifting where necessary, I applied Audacity's "Noise Reduction" effect and "Columbia LP" equalization (performing these three steps out of order causes the appearance of pops and clicks in odd places). I also did more than a little bit of editing, which considering that Audacity doesn't have a built-in crossfade feature, often worked better than it reasonably well should have without it. The result is far superior to previous overdubbing projects I've undertaken with GarageBand (a program I've officially lost patience with), but could still be improved, I think, even using free software and a bare-bones set-up. More to come.

Overdubs

Most instances of overdubbing that I've run across have been effective, but there's one pitfall that I've noticed quite a few times. What makes it unusually obvious that something was overdubbed is when a player's individual quirks are apparent in all of the overdubbed parts (usually it's vibrato or articulation, but it can also just be their sound). It seems to me to be the musical equivalent of cloning (or perhaps incest). In each case, it raises the question of whether the goal really is to give the impression of different people playing together, or whether the player's artistic vision includes the intent to present copies of themselves (or perhaps the intent to not hire any additional musicians for the session). There's nothing perverse about either of these cases; the technology is now old hat. Nonetheless, I cringe a little bit each time I hear an overdub of a player with a distinctive instrumental style. This attribute is elusive for many, and is becoming ever more highly valued in today's musical environment, where many complain of not being able to tell one young saxophone player from the next. Those that possess it, however, may want to think twice about cloning themselves.