A few weeks ago, I posted a big blog entry about how I constructed the second movement of my new percussion concerto, Drum Music.  (If you want to skip to the new demo recording now, without reading all of this blog post, you can click this.  But I’ll be very sad.) The piece premiered on October 11 at Tennessee Tech University, with Eric Willie playing the solo part, and Joseph Hermann conducting.

The first movement, “Infiltrate,” worked fine “in real life,” and I don’t think it needs to be heavily revised.  There are some small corrections, and I’ve decided to ask for a sizzle cymbal rather than a standard suspended cymbal, but that’s about it. (For future reference: Moving a 5-octave marimba is a pain in the ass.)

In the second movement, “Incubate,” I added one new very crunchy note to the chorale at the end of the movement – a chorale which had been straight-up all-white-note harmony, even though there had been a lot of G#’s earlier in the movement.  (You may have seen the blog post about the writing process for that movement.) The revision came about through an email exchange with composer David Rakowski.  (Rakowski – I’ve said it before, and I’ll say it again – has the best composer blog on the InterTubes.)  After hearing the original MIDI, Dr. Davy said:

“As to the G-sharps — they seem sometimes strangely tragic and very expressive. They feel sad, even. Funny that musically there gets to be so much weight on what an analyst would simply gloss over as a detail…  If I were to analyze your G-sharp, I’d either call it a Neapolitan to the implied tonic G that is tragically unresolved, or an appoggiatura to an A minorish chord that is a vagrant chord. But really what it is is a sad note.”

This got me thinking two things: First, why did I always resolve the G# to an A?  What if it resolved downward – to that G-natural?  And second, if the G# (or Ab) is a “sad” note, why did I abandon it before the climax of the movement?  What would happen if it appeared one more time, right before the end of the movement, but this time, I resolved it down to G, instead of up to A?

So I tried it, saving this resolution for the huge chorale at the end of the movement, and — holy crapolli, it’s fantastic.  See measure 55 of movement 2, “Incubate.”)  It’s amazing what a big change a single note can make – and what a difference it is when it’s treated as an Ab instead of a G#.

As a reward, I had some banana pudding from Bobby Q’s in Cookeville, Tennessee.

There’s an unintentionally cool moment in the second movement when the soloist has to bow the vibraphone while also playing it with mallets. What I didn’t consider was the fact that if I wasn’t careful, the soloist would have to bow between the mallets. I, um, wasn’t careful. Awfully cool visual, though – and Eric nailed it.

The only other change to the second movement was the doubling of what was originally a single trombone line, but is now also in low trumpets (rehearsal “N”). Okay, and I also revoiced the sax parts. And the clarinet parts. In the original, I was trying to be a little too clever. In this case, simple is better. (I also stuck a bunch of tempo tweaks in the new MIDI demo to make it more free in time — and more musical — but that doesn’t count.)

So, through two movements, only a few tiny revisions! That’s not so bad!

Then there’s the third movement, “Incinerate.”  Damn it. I kind of blew this one.

First, there were the balance problems.  It turns out that if you stick a percussion soloist in front of an ensemble and ask said soloist to play fortissimo tom-toms and cymbals, you can not hear the band.  This was a problem in my first percussion concerto, but that one had an orchestral accompaniment, and I figured “band is louder than orchestra, so balance shouldn’t be a problem.”  Well, I was wrong.

Eric Willie and Joe Hermann were very patient and accommodating with me as I tried to figure out how to make the accompaniment audible.  We took the bottom heads off the toms, making them punchier and giving them less ring.  This helped a little (I definitely liked the drum set-like punch), but not enough to fix it.  I asked to hear the band play on its own, thinking maybe they were playing “concerto dynamics” – as in quieter than marked, so they wouldn’t cover the soloist.  Again, I was wrong.  Without the soloist, the band was clearly very, very loud.  The drums were just louder.

So we moved Eric to the back of the stage and put him on a riser.  This was a great solution, as it allowed him to play the first movement and two-thirds of the second movement from the front of the stage (where he played primarily marimba and vibes), then he worked his way to the riser for the end of the second movement, where he wails on a bass drum.  The music is designed to be pretty dramatic anyway (going from quiet vibraphone to fortississimo bass drum is going to be dramatic), but the visual of watching the soloist walk to the back of the stage and ascend a 2′-tall platform before hitting the bass drum was awfully effective.  He stayed on that riser for the whole last movement, which visually was much cooler. He totally looked like a rock drummer.  The balance issues were largely fixed.  Largely.

But not entirely, because, again, I had screwed up the scoring of the last movement.  Some low register parts, originally written just for low saxes, bassoons, and bass clarinets, were completely covered.  Eric really couldn’t play much quieter in these parts and still get the ultra-punchy rock drummer sound I was after.  In other words, the problem was not Eric, or the ensemble.  The problem was my orchestration.  I’d been tricked by the balance I heard in the MIDI, where I can make a middle C on a flute sound louder than a trombone.

The other day, Eric sent me a video of his performance of the concerto.  (You can find said video on YouTube.) Now I found an entirely new problem: my trumpet writing.  In order to get a big, bright sound in a few places, I put multiple trumpets in unison – sometimes as many as three of them. Not a good idea in general, but especially bad when they’re unison above the staff.  Playing fortissimo.

In the hall, it wasn’t so bad, but when you hear the recording, which takes away a lot of the “safe cover” provided by a very reverberant hall, you can hear what I accomplished.  Not just volume.  On the recording, much of the last movement, because of the upper trumpets in unison, sounds like… marching band.

I have nothing against marching band, as I’ve made very clear in past blog posts.  But you don’t score something for indoor performance the way you have to score it for outdoor performance.  I made the classic orchestration error: I thought “more players = bigger sound.”  No, “more players” is almost never the best solution.  “More players” in general makes band music sound like “band music,” and “more players of the same instrument in unison,” especially if those instruments are clarinets or trumpets, makes an indoor ensemble sound like a marching band.

I had no choice but to revise the last movement – and revise it considerably.  Material that was originally scored just low winds and marimba (rehearsal letter “O”) now has low-register straight-muted bass trombone with a few hits of harmon-muted trumpet, stopped horn, and harmon-muted upper trombones.  When the “tune” comes in (rehearsal letter “P”), the entire bass line was lost in the hall, so now the solo part is marked down a bit dynamically, the trombones are marked louder, and now they play with accents where they once has staccatos.  (Staccatos on power chords?  What the hell was I thinking?) My weird little countermelody is now doubled in harmon-muted trombone, which should add a bright, odd color to the texture.  But a good odd.

I’d always hated measures 42-43.  Totally dorky, I thought, but I couldn’t figure out a better solution. They’re now less awful, thanks to a better (in this case, simpler) solo part, and the addition of a nasty trombone section glissando into the next bar.  You can get away with a lot if you have a ballsy trombone section gliss to distract from other problems.

There’s now a loud, atonal thing in the xylophone part starting at measure 75.  I dig the atonal obligato against the pretty straightforward tonal chord progression in the rest of the ensemble. The crunchier harmonic language probably again owes a bit to Rakowski’s influence.

At rehearsal letter “R,” when the tune comes in a second time, I used to have trumpets in unison in one octave (going up to D above the staff — in unison!), and the horns in unison an octave lower.

No.  Bad.  Very, very bad.  Horns in unison : a GREAT sound.  But the trumpets — ouch, at least when they’re high. In the revision, there are four trumpets — two in unison on the top octave, two in unison on the bottom octave (with the horns), but for each pair, one player is open, and the other plays with a harmon mute.  (That’s going to be a bitch to get in tune, but if it ever is in tune, it’ll be a nice color.) I also lowered some ridiculously high parts that, at this tempo, sounded a bit too “jazz lead trumpet,” which was never the goal.  I love Earth, Wind, and Fire, but this piece is not that.  Unfortunately.)

I took a similar approach in the big chorale (which is a quote and expansion of the chorale from the second movement – only with “rock drums” underneath it here, transforming it to arena rock — bust out your cigarette lighters if you’ve traveled from 1982; bust out your cell phone if you’re from the present).  No trumpets in unison above the staff.  Now, just a single player, while the other three play the chord progression.  Lordy, it’s better.

I also made considerable revisions to the solo part.  Many of my original fills sounded too similar to one another, as if I was rushing when I was writing them. When I listened to the piece after I finished it (or originally thought I’d finished it), the solo part didn’t do what my ear wanted it to do, or what I tried to “air drum.”  (Yes, I wrote the solo part by air-drumming it.) Now, though, the solo part is “correct.”  Measures 65-69 are now much cooler.  (I’m pretty excited about measure 68 in particular.)  There’s a lot more emphasis on the lowest tom (or mounted kick drum) throughout the movement (especially measures 81, 176, 184), now more reminiscent of some of the dual-kick bass drum pedal in classic Metallica.  The part used to feel entirely Danny Carey (on purpose – he’s a huge influence on my writing – without Lateralus, there’d probably be no Asphalt Cocktail), but now there’s some Lars Ulrich in there, too. The whole solo part is also considerably harder now (is measure 99-102 even possible?), which was an accident, but I don’t think many percussionists will complain that “this percussion concerto is too flashy.”

(Here’s one of the bass drum multi-purpose mallets that Eric made for the solo part)

Other minor tweaks include the addition of a lot more minor seconds where there were previously boring ol’ unisons. The makes the outer sections of the last movement a lot nastier (the good kind of nasty).

The full demo recording is now live, including all of these revisions.  The revised part is also online – and fully printable.  You can see the revised full score, too. This is the first time I’ve posted the full recording of the last movement, and I’m glad I waited, ’cause the original version was simply not right.  Eric and the TTU ensemble played the hell out of what they were given, and thanks to them, I figured how to make the piece better.

The lesson?  You can write a lot of music and still get things wrong.

I think I need more practice.

(thank you to Preston Bennett for that)