Posts from February 2010

Yesterday evening I spent a lot of time hammering out some new lead lines for various sections of my jazz arranging project. This is probably the longest piece of music I've ever written, so I'm finding it a bit difficult to keep my eyes on the big picture; instead, I'm discovering I have a tendency to "chunk" things too much, such that each successive section of the piece follows right after the one before it with very little transitional material.

Actually, in Wednesday's class we talked about this issue quite a bit; the lecture was given by guest artist Lyle Mays, and he spent most of his time discussing how each section of his piece proceeded naturally from the next. It was a bit eye-opening for me; up until now we haven't really spent much class time on how transitions are supposed to work.

That said, the topic has come up from time to time; here are some of the techniques I've learned so far, some from my lab instructor and some from the Lyle Mays session on Wednesday:

1. Overlap. Before switching to a new texture, introduce it as a background, while the original texture is still going on. For instance, if you're about to hand the melody off from the brass to the saxes, bring the saxes in with background figures a couple of bars beforehand. This one has been suggested quite often, and works quite nicely.
2. Pedal point. Kind of an obvious trick, but it's one of the few I know at this point: when transitioning between choruses, adding a vamp of several bars over a pedal bass note can help build tension. (I think I overuse this.)
3. Sequencing (and other development strategies). Again, towards the end of a section, it's kind of nice to take a piece of the last melody phrase and pass it around to different instruments in different keys/registers. This seems to help build tension and volume; I'm using this technique to introduce my shout chorus. (Seems like this is often combined with pedal point.)

If you need melodic material for a transitional section (or background figures), take a look at the themes you've used already, including the head melody. There's probably material there you can adapt (or just flat reuse) to make your transitions work. You might think you're cheating, or that the audience will be annoyed at having to hear the same material again, but if you don't overdo it, it actually lends a great deal of cohesion to the piece.

So that's what I've learned so far. Anybody else have any ideas?

This semester in jazz arranging, I'm learning the basics of big band writing; in fact, the primary project for the whole semester is a single big band chart. The project is broken down logically into smaller assignments by chorus: for example, one week I might have to bring in the lead line for the head chorus, and the next week I'd follow it up with the melody for my sax soli. I like the approach; I think my brain is hard-wired to think of complex wholes in terms of small building blocks.

Sometimes, however, I find myself using too many different kinds of blocks. Each week my lab instructor has given me the same advice: "don't write everything you know all at once." Case in point: my original head chorus started with a simple unison small group melody, but expanded to harmony and then counterpoint within two or three bars. By the end of the first 32 bars, it had built itself up into one hot mess.

My lab instructor gave me the usual advice: "don't write everything you know;" in this case, he meant that using too many different textures in a single 32-bar chorus could be a bit overwhelming, taking away from the cohesion of the section and leaving you nowhere to go for the rest of the chart. Instead, he suggested, why not start the small group in three-part harmony and then have one instrument wander off into counterpoint occasionally throughout the chorus? So that's what I did, and the end result is a lot better.

I think this is sort of a common problem among artists in new mediums; we get so excited about the new expressive tools at our disposal that we forget simplicity. We also forget that the observer of the art will not pay it nearly as much attention as we did when we were making it; and so something that seems repetitive or over-simplified from the artist's standpoint may actually be just right in the eye (ear?) of the beholder.

This is not, of course, to say that complexity is to be avoided at all costs …it's just that it might be a good idea to build up to it instead of introducing it all at once.

Over the last few months I've been pretty dissatisfied with the performance of this blog. Not only were page load times sometimes upwards of 10 seconds, but occasionally my swap usage would max out and crash the server, requiring a hard reboot. And it's a blog, for crying out loud—nothing this simple should ever flat-out crash a server, even if it's only got 256MB of RAM.

Well, this past weekend my employer closed down for a couple of days due to our own little Dallas Snowpocalypse, and I had the chance to implement a single, simple fix I'd been planning for some time. Here are the results in terms of home page ping time:

So what did I do? Simple: I switched my web server software from Apache to Nginx. The hardest part was setting up the PHP FastCGI process; although there are lots of instructions online as to how to do this, most of them seem a bit outdated. I ended up using an init script from the Nginx wiki; once that was taken care of, it was a simple matter of converting my Apache confs to Nginx's syntax, switching the ports over, and watching my site's performance improve fantastically.

So there you have it—my blog is now practically readable again, and it turns out the performance problems had nothing to do with my programming! Good news on all fronts today.

Those who know me well know that I really, really hate to throw away data. I have all kinds of stuff sitting around on my home server, some of it dating all the way back to middle school, and most of it of very little interest to anyone today (even me). Well, the other day I stumbled across something that was sort of interesting: trombone practice recordings I'd made early on in college.

For as long as I've been a musician, my teachers have told me that one of the best ways to discover where you need to improve is to record yourself playing; for some reason, however, I've rarely bothered.

It's partly a discipline problem—I've never been as consistent a practicer as I should be—but I think it may also have something to do with my data obsession. When I make a recording, I don't just listen to it a few times, note the things I need to fix, and then throw it away. No, I think to myself, "what if I want to listen back to this five years from now and hear if I've improved?" And so I keep it, and not just as an MP3…no, I keep the original, huge, lossless WAV file. For-ev-er.

Now, this used to take up one heck of a lot of space, and a lot of manual backup effort too. As a result, I would rarely do it …too much effort to archive a daily audio practice session when I've got other things I need to store in that precious space.

These days I don't worry about that, for two reasons:

1. Storage is cheap. I have a 500GB RAID-1 network attached storage device in my living room. If it fills up (which won't be happening soon), I'll just get bigger drives; they don't cost that much in the long run.
2. Not all compression is lossy. If I re-encode my original WAVs to lossless FLACs via some automated process, I can store them in half the space and still play them back without uncompressing …no data loss, and very little loss of convenience.

As a result, I've started recording my practice sessions again. This has had a few important benefits:

1. I can listen to my playing after the fact, discovering issues I didn't notice the first time around.
2. I actually practice regularly, since I don't want my archives to be missing a day when (if?) I look back into them five years from now.
3. Also, I've found it a lot easier to notice the negative effects of skipping a day of practicing.

But at the end of the day, the best benefit is that I'm starting to recover some of my old improv chops. I've still got a long ways to go, but it's really encouraging to feel somewhat skilled at the trombone again.

Last November my coworker Adrian Rollett and I got the opportunity to present a talk at DrupalCamp Austin; I've blogged about the content previously, but now that the video is up, I thought I'd post it here as well.

Read more...