RLPF.ar

Last bit about SC Symposium 2009; general news

As promised (weeks ago), just a few remarks about the Saturday night concert. Unfortunately time fades memory and the pieces are no longer fresh in my mind. In general, I was drawn most to the pieces that involved a live performance component. Mark Ballora's Singularity was handsome if relatively conventional (not really a criticism, as my own work steers clear of the far-far-out avant-garde); Andrew Greenwald's Blocks.flt(remix) disassembles the instrument into separate components of fingering, breath, embouchure and vocal effects, recombining them in carefully-deployed extended techniques.

Nick Collins, though, stole the show with a piano "concerto" overflowing with bombastic, ironically-intended classical music gestures, replacing Emerson, Lake and Palmer's acid with crystal meth. In the follow-up work, Kinesics, machine-listening algorithms select from various activities ranging from broad-brush gestures on the piano to diverse tics of insanity in which the performer is instructed to touch himself on the head or torso in various places between notes, or play some registers of the instrument randomly with his heels. Played with an absolute straight face, it was as close as I've seen to the Ministry of Silly Walks in any concert. (Like Monty Python, Nick is English and obsessively fond of acronyms, unleashing these works under the nom de plume of Lan Klekturin. Other work has been released as Sick Lincoln or Click Nilson.)

Since then: up until a few days ago, I was working on an introductory SuperCollider workshop for NIME 2009 (New Interfaces for Musical Expression). Unfortunately they didn't get the number of registrations they hoped for, and the workshop has been canceled. That allows me to turn my attention back to the violin piece full-time, or as full time as I can manage outside of work. This weekend I assembled all the existing segments into a continuously-running sequence -- my first chance to hear the flow of the electronic parts from the beginning up to... as far as I've written. Very exciting to see it coming together! But there is still much to do.

Symposium, Saturday PM, continued

The final session Saturday afternoon was on some compositions and performances using supercollider. Sam Pluta kicked off with an explanation of data structures/monoliths ii (for chion). Of particular interest were the challenges of keeping the video (Quartz composer) in time with the audio from supercollider, and layout of the performance interface, designed (as Sam put it) for someone who doesn't play any instruments at all.

Dan St. Clair's Call Notes had been amusing me the whole weekend. It's an ingenious installation, using solar powered speakers driven by microcontrollers to mimic the sounds of birds singing pop songs ("Like a Virgin," "Der Kommissar" and such). Some of my work got a mention in the talk, from a couple of years back when I did some work on analyzing FFT data to identify spectral peaks. He used it on bird song recordings to figure out the spectrum to morph the pop songs onto.

The installation, by the way, is effective. I didn't see it for myself, but others told me they saw unsuspecting students walking near some of the speakers and stopping short, unsure of what they were hearing.

Last up was Vincent Rioux, who's been involved in some inventive multimedia, dance and street theater performances in France. Discussions were brief, but the video clips were spectacular. The most memorable was the last, of a street theater piece where the performers strapped on sensor suits and proceeded to clean the streets and nearby buildings in an unexpected and very funny ways. I've said it before, and I'll say it again -- God bless the French!

I'll take up the Saturday concert next time --

Saturday talks, part 1

Continuing on with Saturday...

Tom Tlalim was unfortunately not able to make it at the last minute, so we didn't get to hear about his intriguing-sounding mapping library. Instead we had an extensive demo of Marije Baalman's sense world data network. This library identifies a real problem working with networked machines and sensor data coming from multiple sources. The problem is that it's tempting to hack together a quick and dirty communication protocol for each piece, but it's ultimately harder to get things right that way, especially if all the personnel come together for only a few days before the performance. Instead, her library sets up a central hub to which other clients can subscribe, simplifying testing and rehearsal.

Next, Ron Kuivila and I took the stage to go over topics relating to patterns. I spent some time over the winter writing "A Practical Guide to Patterns," a set of help files describing theory and practical uses of patterns for sequencing data and musical events. Ron followed up with more sequencing tricks using Pspawner and demonstrated the Conductor library for handling preset data and easy to build graphic interfaces.

After lunch -- I missed part of Scott Wilson's talk on BEASTmulch for managing huge multichannel pieces going up to 100 speakers or more. What I saw was some very slick work building a virtual model of the speaker system including positioning, directionality and speaker characteristics (complete with a Quartz Composer animated 3-D visualization!).

Jan Trutzschler presented a multitrack sequencer application built totally in SuperCollider, called TeaTracks. Although I like to keep my own use of SC's GUI capabilities fairly simple*, I've enjoyed seeing those capabilities grow and then be put to good use making really polished interfaces.

*My performing interface looks complicated, but it's all fairly simple stuff -- just a lot of them!

The remaining talks were on some compositions written in SC -- I'll take them up in the next post.

Symposium, Friday night concert

Highlights:

Witty, clever theater from Ben Klein in Tubabox the Doppelgänger. He put a box with a primitivist face on it and a cheap piezoelectric transducer to get some grungy lo-fi source audio into the computer.



Angry Sparrow by Chikashi Miyama. Spectacular to watch -- he uses a custom box of seven infrared sensors and choreographs control of sonic parameters with his hands. My one issue with the piece is that the sound design is on the one-dimensional side and could benefit from more variety. But, it's astonishing how much the sound did change just from different motions over the controls -- and don't underestimate the audacity of the visual + sonic image. The box easily solves the problem of boring-to-watch electronic performances, and the in-your-face downtown edge is refreshing.

Sam Pluta's data structures/monoliths ii (for chion) -- a screamer of a mashup of some of the best sounding moments in Hollywood. I liked the vertiginous quality of "anything could happen next" and the irreverence of mixing art with drivel (which reminded me of Anthony Lane's comment about preferring "trash and classics" over the dread, pretentious middlebrow).

Marije Baalman's Livecoding session was cut short by a technical glitch, but was very promising up to that point, especially in the use of typing signals from the keyboard as audio triggers -- the act of performance (typing) becomes part of the sonic result of the performance.

I'd rather not dwell on the lowlights. Most of the other works were enjoyable. <RANT APPROACHING> Only one piece in the second half really stood out as a self-indulgent avant-garde exercise which made its point in two minutes (but went on for another 20). On one season of Project Runway, one of the designers gave an articulate defense of her design, invoking cultural-criticism mumbo-jumbo and suggesting some profound meaning to the garment -- which Jay McCarroll deflated easily by saying, "That's a big talk you've got there, but the talk doesn't match the dress." (Michael Kors opined, "Come on, it's just a tube dress with some tape on it, get over it.") I feel the same when I read a long, jargon-y program note but then the music is aggressively inarticulate. But that's the way of things at computer music conferences. </RANT>

And my performance? Well... almost. The mic levels in my software were not calibrated properly so the sound was not consistent. But I think it was enjoyable nonetheless.

Afterward, Brian Parks (Wesleyan U student) played a one-hour minimalist organ piece by hand in the Chapel. I didn't think I would stay for the whole thing, but then I started up some walking meditation and by the time I was finished, so was the piece! It settled my nerves after the concert. Some of the harmonic changes reminded me of my Duke colleague Chris Adler's work (but Chris would have thrown in more surprises).

Symposium, Friday (day)

Continuing with the symposium's Friday events -- neural networks and machine learning in the morning. Most of the machine learning talks were about self organizing maps. This work is entirely relevant to my algorithmic composition approaches, but it will take some time for me to digest the material.

Marije Baalman spoke about cross-platform issues and mentioned some important things not to do. One slide with general advice on writing for multiple platforms included the clever advice to "try to write cross-platform code"! It doesn't have to be circular logic, if interpreted to mean that one should keep cross-platform matters in mind while writing code -- at least then one is less likely to assume that if it works on my machine, it must work on everybody's.

After lunch, some theory: methods for working with serial techniques, a game after Plato, and a spectacular demonstration of the virtual gamelan project out of Graz. The virtual gamelan encompasses not only sonic analysis and resynthesis of the instruments, but also stylistic analysis and algorithmic "recomposition" of the behaviors of the players, including tempo awareness and signaling between instruments. Very sophisticated.

I missed the Timbral Analysis panel because I had to fetch my performance kit from the hotel. That's a shame because I could really have used the material for the piece I'm working on now.

More about the concert later...

Symposium notes: Thursday

I just got back Sunday night from the third annual SuperCollider symposium, held this year at Wesleyan University in Middletown, CT. A very fine time indeed.

Some scattershot reflections on the events... more to follow in subsequent posts.

Thursday: Nothing in particular planned for the day, plenty of time for informal chats and working on issues (Ron and I knocked one out during this time). Very pleasant to see many people whom I hadn't seen since the first symposium in Birmingham, UK.

The evening's keynote address by James McCartney discussed an older prototype of SC3 that was never completed. This version would have permitted something that's difficult in the current release, namely "single-sample feedback." The concept is brilliant: have unit generator objects in the SC language write C code to calculate the signal processing graph. Unfortunately it wasn't practical because running an external C compiler to produce the object code to load into the synthesis server took too long, killing interactivity. Also, (paraphrase) "Single-sample will always be slower than block calculation as long as computer memory is hierarchical [that is, CPU registers < L1, L2... Ln cache < RAM < disk], and memory is becoming more hierarchical over time, not less." So block calculation is here to stay -- fine with me, I like how scary-efficient SC is today.

A few people have posted photos and video online:

http://www.flickr.com/photos/djensenius/
http://www.flickr.com/photos/danstowell/sets/72157616657624801/
http://www.flickr.com/photos/hecanjog/
http://vimeo.com/4093195

SuperCollider Symposium 2009: Pattern Guide materials

For the convenience of the attendees of the SuperCollider Symposium 2009, my lecture materials are online:

http://www.dewdrop-world.net/sc3/sym09

SuperCollider exposé (hush hush!)

I decided to put up the PDFs of the so-called "scores" for the two supercollider pieces I wrote up recently on my website. They are pretty technical documents (and not enough in themselves to reproduce the pieces) but I think they're interesting as documentation of some of my compositional thought processes.

Sangha flower / audio excerpt (from a different performance)
Got an itch to scratch / audio excerpt

For the musical notation, I discovered that I could export snippets from Finale into Encapsulated PostScript files, and then OpenOffice Writer could import them just like any other graphic. So I got the notes in the middle of the text, and everything is vector graphics so it looks great no matter how much you zoom in.

Getting back in touch with older pieces

Busy as always... just haven't had time to post! That situation is not likely to change anytime soon.

Recently I had the opportunity to go over some past pieces: a couple of supercollider tracks and also my Ph.D. dissertation. (Perhaps later I'll explain why.) A most interesting exercise to lay out the results of several years' work in an admittedly highly technical form, but one that I hope has a chance of being intelligible to a human being other than myself. I wouldn't have that expectation of the supercollider code! It's also a pleasure to note that, years later, my dissertation strikes me as stronger work than I remembered. After finishing it, I think I just wanted to put it behind me and move onto the next phase of development, but now that seems to be a mistake. Too much of the piece is emotionally gripping to let it go to waste. Perhaps it isn't too late to find players.

The violin piece is moving out of the concept stage -- notes are coming together and I feel like I'm picking up some momentum.

Recent listening: I've been looking for a long time for a decent recording on period instruments of Bach's flute sonatas. I became deeply intimate with the main six sonatas (BWV 1030-1035) and the solo partita in a minor (BWV 1013) as an undergrad, though I didn't get to know the fabulous G major trio sonata (BWV 1039) until graduate school and somehow missed the g minor (BWV 1020) sonata altogether. Then, not long ago while listening to an online Baroque music radio station, I heard recordings of some of the sonatas in the hands of Lisa Beznosiuk and admired both the technical excellence and strong interpretations. Lo and behold, the iTunes music store carries the entire set, so I bought it online and have been enjoying it ever since! The pieces are every bit as exciting as I felt they were when I was learning to play them, and I've really enjoyed getting back in touch with some repertoire that meant a great deal to me in college.

Although one Amazon reviewer panned the recording engineering, I find the sound quality entirely pleasant, and any quibbles here are easily overshadowed by the exquisite playing. Beznosiuk's tone is ravishing and her interpretive decisions are impeccable, well thought through and daring in places. I'd recommend this recording without hesitation.

A Grand Day Out

Today was my show in town, and all and all, it was a considerable success. The software remain stable for three performances of about one hour each, which is encouraging considering one longtime electronic musician's opinion that in performance, "whenever there's a computer involved, crashes become much more likely" (can't find the reference right now). There were a few flubs, but I can attribute all of them to performance errors or weird corner cases in my code. (That's one of the things I like about SuperCollider -- if it's my fault, I can fix it, and if it's the program's fault, the source is available to get fixed.) I also got some feedback from the microphone, which was most unwelcome but I was able to stop it pretty quickly.

Here's one photo, where I'm playing the xiao flute I got in Beijing back in 2005. The park itself is to the left of the frame, with a cafe stand, some small tables, an algae-laden pond hosting about a dozen hungry ducks. In the back of the frame is one of the tour buses that crawls DC daily -- the air brakes were a bit of an impediment to concentration, but I kept my wits about me nonetheless.



While I can't say the performances were perfect, I got some great feedback -- very encouraging to see others get enthusiastic about my music! -- and a couple of other opportunities might materialize. Overall, it was a fantastic experience and I'm very eager to play again, somewhere, soon!

EDIT: An excerpt is now online.