Technological and Computer-based Projects
Sevenless/Bride of Sevenless, Dog of Glass, Ladybird
Early/Ladybird Late, and Bang Senseless/Bang Sensitive— all genes in
the fruit fly Drosophila melanogaster, and all titles of pieces in the genome series.
This series, started in 2001 at the Aspen Music Festival,
began simply as a series of pieces with live electronic processing and some
playback elements, requiring a performer (usually me) on some sort of
electronic gear. The first problem I ran into was one of rehearsal: because I
was basically performing an accompanying role, making electronic chamber music,
the other musician needed me around to practice. That is, me, my gear,
microphones, PA/amplifier, cables, and setup time.
I loved rehearsing these pieces, but the amount of time it
took to put them together was out of proportion to their relative lack of
technical difficulty for the instrumentalist.
The second problem was that musicians enjoyed these pieces,
and wanted to play them more. But, that would require me to travel to each
performance and schedule rehearsal time in the venue. I was lucky to travel
several times to perform Sevenless/Bride
of Sevenless (in its first version) but by then I was starting to think
about interchangeable parts for future pieces. After Sevenless I gave up external hardware and wrote for computer
processing only.
Dog of Glass and Ladybird Early/Ladybird Late were
transitional pieces: the electronic processing is based in Logic Pro, allowing
a performer to download a session file, open it in that program, and rehearse
alone. This requires the performer to own Apple's Logic Pro software, which is
one drawback. But the biggest drawback was that Logic is inherently a
timeline-based program.
Music, of course, is timeline-based, too, but unless the
performer is to robotically maintain the same tempo in every performance
situation, the electronics represent a superfluous and problematic timeline
that will inevitably clash with the performer's expressivity. The performer
starts the piece and plays, and must stay in sync with the playhead in order
for the changing effects to fall at the right place in the score.
I got around this problem in a related piece of sound-design
work I did for [Grammy-award-winning new music sextet] eighth blackbird. For their staged puppet-theater performance of Pierrot Lunaire and
Jacob Druckman's Reflections
on the Nature of Water with the Blair Thomas Puppet Theater at Chicago's
Museum of Contemporary Art, I was asked to create live, evolving multi-channel
sound processing for the movements of the Druckman
piece. This was done in a similar manner to Dog
of Glass and Ladybird, and I used
Logic Pro as a surround mixer and a series of busses to manage several
microphones placed around the marimba. Prewritten automation on the bus tracks
controlled myriad parameters of delay, reverb, gate, balance/width and
level— and I simply built in a 'safety cushion' at the end of each
movement, where the automation would continue long enough to account for an
extra-slow performance, or my accidentally starting the playhead a little bit
too early, or percussionist Matthew Duvall waiting a little longer than usual
after his signal before starting. (Or even, though this would have shocked me,
a memory slip on the part of the performer.) This required someone to 'babysit'
the electronics at the performance, but didn't ask any special skills beyond
concentration, score-following and remembering a simple keyboard shortcut.
Still, though, I knew: This could be better. And it was this
thought that first led me to buy and learn Max/MSP, which has two giant
advantages over Logic Pro[1]
for this particular application: Firstly, it is not timeline-based, or rather,
while it supports timelines, as the composer/designer/programmer you have the
option to use zero, one, or multiple timelines, and to change their behavior in
different parts of the piece. Secondly, it allows the creator of a piece to
export it in a form that can be opened by the free Runtime version of Max/MSP,
which is available as a direct-link download from Cycling74's website. Thus a
download of the score can include a link to download the software needed to
play the piece, and all the performer needs to bring to the table is a laptop,
a microphone, and their instrument.
Conceptually, the rewrite of Sevenless/Bride of Sevenless was a great jumping-off point, an
opportunity to rethink what it meant for me to write pieces for live
electronics. Inspired by the series' biological roots, as well as more
technically by some of what I'd learned about self-governing feedback systems
in the meantime, I set out to make these pieces as "aware" as possible about
their state, and able to manage effect parameters and levels as a human would,
but without any input besides what the performer gives naturally, through
playing the piece. Completely achieving this goal would mean that a performer
would play chamber music with the computer as though playing with another
human— leading, following, and most of all trusting, as we do when
we know we're playing with a true fellow musician, that what we hear back will make sense in the context of the music
we're making together.
With that in mind, I wrote tempo-sensing algorithms,
feedback-detectors, delay-feedback-cycling loops, and many interrelated
pseudo-randomizer functions (and found them, and borrowed ideas for them, as is
the spirit of coding). The performer sets a gain slider in the dress rehearsal,
to account for that particular hall, amplification and microphone setup, and
taps a foot pedal to start the piece. Each time a major change occurs in the
music, there is an instruction to tap the pedal again, and the processing
shifts. (Because this piece is for solo snare, making those changes automatic
would be very difficult. Even with a pitched instrument, I would be hesitant to
simply allow a detected state to change sections— either a certain
amount of time should be spent in the previous section, or the detector should
be 'armed' with a tap on the foot pedal anywhere in the 8 bars or so prior to
the change.) Whether auto-detection would ultimately be a benefit in terms of
navigating a piece's large-scale form is an empirical question, which will only
be answered by writing more pieces with intelligent sound environments, and
talking with the musicians that play them.)

The Genome Series:
works for solo instrument and live electronics
1. Logic is a powerful and deep program, despite a few flaws, and I use and recommend it for
writing and scoring to picture. It has a powerful Environment feature, as well,
offering some non-timeline-based functionality; but for those inclined to learn
its distinct workflow, the power of Max/MSP is best suited for the projects I
describe here. [jump back up]