Preparing this weekend for my gig this coming Saturday, I had an incredibly frustrating few hours when I absolutely could not get Soundflower to work as part of an aggregate audio device. I tried previous versions, remaking the aggregate device, all kinds of things. No sound.

What that means is that I can’t use mlr in Max and then pass the sound over to Ableton for effects, which has been my basic setup since I’ve been using a monome.

I’ve performed without effects before, but I like having some delays and other ways of mutating the sound that comes out of mlr, which is otherwise modified only by being repitched and/or reversed. So I was pretty freaked out about playing for half an hour or more with no effects.

After I got a grip on myself, I decided to try finishing a project I’d started a few weeks ago, a version of mlr with aux sends and effects within the patch. It was rough going because of my super-beginner level Max skills, but I ended up hacking something together that actually works. Three aux sends for each mlr group!

For now, I’m using effects stolen from this tutorial. They’re okay for my needs, but I’m still dreaming of something like Ableton’s granular delay, or even better the Daevl plugins. (I’m in Max 5 so can’t use them, it seems.) When this performance and a few other pressing commitments are over, I can concentrate on patching some creative effects of my own, and I’m really looking forward to it. For now, does anyone know any cool pre-made Max/MSP effects?

UPDATE: I’m going to try the vst~ object and one or more of the free Soundhack VSTs. Crossing my fingers; if this works, it’ll be awesome.

I’ve got a performance coming up next week (Saturday the 7th at 9 p.m. at The Tank), which means I’ll spend this weekend holed up in my bedroom with my trusty monome, a mic, and a collection of portable instruments (melodica, glockenspiel, harmonicas, still deciding about the bamboo flute) trying to put together something resembling a set. My playing is pretty improvisational, but for a live audience I need a structure — not a chordal or modal structure, but a structure of row assignments in mlr and effects sends in Ableton.

One of my main concerns as an electronic musician was crystallized by a concert I went to a few years ago, when The Books played the Wordless Music Series. I really like The Books and have bought at least a couple of their albums, so I was excited to see them play live.

I was bored silly during the show. Nick Zammuto and Paul de Jong got up on stage at the Ethical Culture Society, put earphones in their ears, took up their instruments, and pressed “play” on a DVD player. The DVD had a track for each song on the setlist. But it wasn’t just a projection to accompany the music — it was most of the music, all the little cut-up samples that comprise The Books’ records, minus guitar, cello, and vocals. Nick and Paul played and sang along with each track, and that was the show.

What had gone wrong? Well, Nick and Paul had faced a decision about playing live after making their name as a band with a recording. Obviously, it would be impossible to faithfully recreate a Books track completely live onstage — the tracks are the result of God knows how many hours of detailed editing and composition, and no audience wants to sit around and watch that. And the recorded tracks depend so heavily on the brilliant use of samples that a re-imagining of them for acoustic instruments alone would probably not be what a Books audience wanted, either. So Nick and Paul went with a system that reproduced the sound of their recordings. The tradeoff: It was more like karaoke than a real performance.

It’s very easy to concoct a live electronic performance that crosses into the karaoke kingdom because computers are very, very good at perfectly reproducing actions and events. The possibilities for perfect automation they provide are very powerful, but as even a cursory familiarity with the history of computer music makes clear, this same automation can very quickly lead to deadly boring results. I have nothing against an exquisitely sequenced and edited electronic composition, but as an audience member I don’t want to experience any more live performances that consist of the judicious triggering of pre-made clips.

In my own music, I’ve decided to head in the opposite direction of that Books performance. One rule I live by is No Prerecorded Samples (with the exception of an occasional field recording). I still make sample-based music, but I have to record each sample live on stage, which means the properties of each sample are new to me and have to be discovered as part of the performance.

Another consequence of the No Prerecorded Samples rule is that things don’t turn out perfectly. Lately I’ve been performing a Mississippi John Hurt song by recording a scale on a harmonica into mlr, then constructing accompaniment patterns from different degrees of the scale. Due to some timing issues between me and the software, the patterns are almost always just a little bit off, which makes it challenging to sing over them. But the off-ness builds up with each additional pattern, creating unexpected and, to me, quite pleasing complexities as the piece develops. The same thing happens even in my completely abstract improvisations — the old jazz adage that “if you play a wrong note, repeat it a couple times and it’ll be right” really gets at something important.

So I’ve adopted a second principle for my performances: Get Things Wrong as Often as You Need To. It keeps things interesting, at least.

Hi noisepages!

I’ve been making music for years, but have gotten more intensely involved after I got my first monome in March 2008. (It was a 128; I’ve upgraded to a 256 because I can always find a use for more buttons.)

I make music as Weather Ballons; album here. It’s pretty abstract, noisy stuff — minimal inputs that turn into maximal outputs. One of my central concerns is creating live computer music that includes the possibility of either failure or serendipity. I do this by using no prerecorded samples (well, a few field recordings from time to time, but nothing else), so each sample I manipulate in a performance is new and unknown. This process also contributes to an organic, human-sounding music, which is something I really like. Another way to think of it is as a way of making a virtue out of being too lazy to spend hours editing samples.

Lately I’ve been exploring vocals along with the computer soundscapes, and I did a version of Mississippi John Hurt’s “Spike Driver Blues” at the monomeet open mic using a live harmonica sample turned into accompaniment patterns in mlr. I’m also working on some simple vocal modules — texts plus tones — that can be sung over spontaneously created computer improvisation.

I’m also trying to learn Max/MSP so I can have more control over my primary instrument, the monome. I don’t have much of a programming background, but I’ll be documenting my progress (in fits and starts) here.