• Introduction

• Papers

  • Keynote addresses
  • Contributed papers
  • Forum on computer music laboratories
     
    

• Installations and performances

  • Installations
  • Concerts
  • The Conference Club
     
    

• Conclusion

Introduction

This is a report on the 2005 Australasian Computer Music Conference (ACMC05), held on 12-14 July at the Creative Industries Precinct, Queensland University of Technology; the conference theme was “Generate and Test”. The conference was ably organised by Greg Jenkins, assisted by Andrew Brown and Tim Opie (responsible for the conference proceedings), Jason Zadkovich (resident technician), and a team of about 20 student volunteers. Usual disclaimer: what follows is a personal view of a complex event.

In the three days we heard about 30 presentations and 30 pieces of music; in addition there were installations, a poster session, a forum about computer labs, and the after-hours Conference Club. After the conference finished, there was a day of workshops, which I did not stay on for. The conference was well attended, and there were quite a number of new faces as well as the usual regular attendees.

Papers

Keynote addresses

There were three keynote addresses. The first was by Paul Doornbusch, outgoing ACMA President. He started by describing his work on reconstructing the music of the Australian CSIRAC computer, probably the first in the world to play music. Paul went on to discuss the work of various composers, including Hiller, Brün and Xenakis, from the point of view of “generate and test”. He made a plea for artists not to get completely wrapped up in the algorithms (“generate”), but to develop their aesthetic judgement as well (“test”).

The second keynote address was by Katharine Neil, who has spent the last seven years as a computer game sound designer and programmer. She pointed out that avant-garde compositional practices such as aleatoric music and music that responds to the environment are common in game music, and reach a huge audience. The artistic and technical challenges are considerable, and sometimes bizarre: one game music composer spends his time creating bridge passages linking any one theme from the “Star Wars” movies to any other, to allow for fluid gameplay. Katharine's entertaining talk was the first of a small group with a commercial focus, something fairly new for ACMA. In this context it occurred to me that the Sydney Conservatorium of Music, where I am currently a student, has what amount to strong commercial links, as it is closely connected with the orchestras and the opera.

The third keynote address, from Ross Bencina, came right at the end of the conference. Ross started by giving a short performance using a glove and his well-known AudioMulch software. He then gave a thoughtful discussion of the process of developing software for creative purposes. Ross's software was written to support his own artistic practice, but it contains close to 100,000 lines of code; a project of this size requires a disciplined “engineering” approach, and much of the code is either implements the user interface or provides more or less general “infrastructure”. The DSP code is only a small part of AudioMulch. Any piece of music software imposes some stylistic bias and constraints (despite aspirations to the contrary); Ross said that Miller Puckette has recently called for the models and abstractions used in music software to be made explicit and documented. Incidentally, the tension between creativity and disciplined production is not specific to music software, but surely has been around at least as long as the craft workshop.

Contributed papers

These comments do not attempt to cover all the papers — order the conference proceedings for your library! The conference covered a wide range of topics; the papers did not fall neatly into subject groups, though many had some connection with performance. Those papers with a potential commercial connection were concerned either with computer games or with ways to help non-musicians or beginners to make music. There was one survey paper, Jim Barbour's timely account of the current state of delivery methods for multi-channel audio.

Steven Livingstone, Ralf Muhlberger and Andrew Brown presented an ambitious scheme to mark up music with “emotional metadata”, used to influence MIDI playback. The ultimate aim is for the emotional state of the listener to influence the emotional tone of the playback, with the main intended application being computer games, though Steven hoped the project might be useful for educational purposes as well.

Garth Paine (in joint work with Ian Stevenson) talked about mappings for a new performance controller, the “Thummer”, being developed by Jim Plamondon. Jim was at the conference and ran a post-conference workshop on the Thummer. Unfortunately Garth could not be specific about the Thummer, as Jim wants to protect its details for the time being (I assume that the workshop attendees had to sign a non-disclosure agreement). So Garth's talk was rather too abstract. We do know that the Thummer has a large number of buttons, concertina-style, laid out so that the fingering is the same in any key. It also has 10 or so continuous controllers, but what they are remains a mystery, and therefore so does the potential of these controllers for expressive performance. Although the Thummer is a controller, it is not restricted to MIDI; I understand that the controller sends raw data to an attached computer, from whence it can be sent on in various formats, of which MIDI is only one. The Thummer is also claimed to be well adapted to microtonal music. Unfortunately Ian Stevenson wasn't at the conference to present his more general paper on “Design issues for new performance systems”; it is in the conference proceedings.

A paper by Greg Schiemer and Mark Havryliv described a system for using mobile phones as musical performance instruments, via Bluetooth and Java applications running on the phones. Performers can modify the sounds coming from their own phones of from other participants' phones, including changing the tuning system. Greg and Mark have developed a PD to Java conversion tool, allowing PD patches to run on the phones. In this context Warren Burt commented on the perverse tendency of electronic musicians to continually revert to using crappy-sounding devices.

Rene Wooller and collaborators described their work on a “participatory electronic dance music environment”. built on Rene's LEMu program for generating electronic dance music in real time. Rene's group have built various controllers, including robust floor mats and the “Seuss Stick”, an instrument with a telescopic action like a bicycle pump, the ability to rotate one part relative to the other, and three pressure-sensitive pads. Rene discussed the problems involved in having a group of people collaboratively controlling a dance music engine.

Some other papers concerning performance are mentioned below in connection with the concerts. I would also like to mention briefly Alistair Riddell's inspiring talk about a performing ensemble he was involved in, using glove-like sensors, and Steven Campbell's talk about PlaY+SPaCE, a system similar to Rolf Gehlhaar's long-established SOUND=SPACE in that it uses multiple ultrasonic sensors, but having the advantages of being much cheaper and of being controlled from Max/MSP.

A small group of papers concerned algorithmic or generative composition. Luke Harrald presented his work on using the Iterated Prisoner's Dilemma as a compositional engine. In the Prisoner's Dilemma, two (guilty) prisoners are interrogated separately. If both remain silent they each receive 1 year in jail. If one confesses and dobs his mate in, the dobber gets a pardon, and the other prisoner gets 5 years. If they dob each other in, they both get 3 years. In the Iterated Prisoner's Dilemma, this is repeated many times, and the “prisoners” can make choices based on what happened to them in previous rounds. Luke has made several pieces using the actions of simulated prisoners to control musical material in various ways.

Other papers involving algorithms were Dave Burraston's paper presenting his ongoing work with cellular automata as compositional generators, and the paper by Francis Soddell and Jacques Soddell, who are respectively a computer scientist and a microbiologist, on using L-systems to make music for an installation. L-systems produce tree-like structures, and in this case an L-system was constructed to make branching shapes like those generated by a particular fungus. Traversing the structure produced a MIDI file; the choice of timbres was made freely for aesthetic reasons. In this context Warren Burt commented that in general sonifications succeed or fail as artworks according to the choice of timbres, and this work was no exception: an example with a piano timbre was unimpressive, but the final result sounded good, and well-suited to the environment of the installation (a botanical conservatory).

Peter McIlwain presented a paper (joint work with Jon McCormack) on a deceptively simple network traversal algorithm for generating melodies. An “activation point” moves around the network according to some simple rules, triggering a note at each node. The time interval between two notes depends on the the length of the line joining the nodes. The idea has some possibilities for analysis as well as melody generation, which Peter demonstrated with a discussion of “Three Blind Mice”. A couple of the papers discussed below in connection with the concerts were also related to generative composition.

There were three papers on analysis. Tim Opie presented a paper on aspects of sound amplitude as part of an ambitious study of the various attributes of natural sounds, with a goal of compositional applications. Barry Hill compared and contrasted two contemporary electronic genres, Trance and Drum n Bass, backed up by transcriptions. Unfortunately David Hirst was unable to attend the conference to present his work on computerised study scores for electro-acoustic music. This work represents a practical outcome of David's earlier more theoretical work on the analysis of electro-acoustic pieces. David's study score for a piece (Denis Smalley's “Wind Chimes”) consists of two spectrograms at different levels of detail together with a meticulously marked diagram showing all the significant sonic events. The whole thing is interactive, being realised in Flash, so one can listen to a section of the piece at the same time as scrolling through the corresponding visual representation.

Forum on computer music laboratories

There was a forum on computer music laboratories, with Alistair Riddell, Garth Paine, Tim Kreger and Andrew Brown, and contributions from the audience. The question of the role computer music labs play in an era when most students have their own machines has been around for some time. A partial consensus did seem to be reached by the forum, along the following lines. (Apologies: most of the time I didn't note who said what.)

A computer music lab will be needed for non-specialists who may only take one or two courses. Students who are specialising in computer music can be expected to buy their own machines. Garth Paine indicated that in the UK, Apple have a leasing scheme for students, and there are some moves towards setting up such a scheme here. Garth also said that the University of Western Sydney is looking at licensing deals so that students can be provided with a CD-ROM of appropriate software. This would answer a comment of Alistair's, that students needed to be weaned off simple-minded software, and fits with a general view that a common environment makes for a more cohesive student group. (Though someone defended having a multiplicity of software.)

Alistair also commented that students undertaking animation or other digital arts projects may still need more powerful computers than they can be expected to buy; in general there may be a need for some special-purpose dedicated computers. The convergence of sound and vision will need to be addressed, if that is the way our community is moving.

Given that students can do most of their work at home, why would they come in to Uni? There was a strong consensus that they should come in, as they do learn from one another. The consensus was that the money saved by having fewer lab machines should be used for facilities which the students cannot easily replicate at home, such as good recording facilities and a high-quality multi-channel playback space. It was also suggested that the typical computer lab is a fairly unpleasant environment, and that the shared spaces for students should be made as pleasant as possible. Someone also suggested that students should be given team projects, which means that they have to collaborate.

The participants said that getting centralised IT support to understand the needs of computer music was very difficult; for example IT is likely to object to computers with CD burners because of piracy concerns. Good tech support is very important, and it is not clear what the responsibilities are if students are required to provide their own machines, as opposed to choosing to provide them. The legal issues also need to be well sorted out.

Installations and Performances

Installations

There were several installations, but I only engaged with two, Amanda Cole's “Sine of the Tones”, a Max/MSP audio-and-video piece with changing coloured stripes, and Michael Yuen's “Atrium” in a foyer, where small speakers caused the glass walls to emit soft hissing and scratching sounds. At least one of the others wasn't working when I was free to see it.

Concerts

By my count there were thirty pieces in the six concerts; 11 were live performances, two were abstract videos and the rest were tape pieces. Most of these were multi-channel works (for between four and eight channels), but there were several stereo pieces, showing that the genre is not as extinct as it appeared to be a couple of years ago. Only one piece was “diffused” live, Camilla Hannan's soundscape of apparently unpeopled factories.

The highlight among the tape pieces was Robin Fox's “The Third Space” in three movements. The second movement consisted of a pulsing sound which was swept around the ring of speakers, at first slowly and then faster and faster. Some strange phasing effects resulted. The outer movements consisted of small fragments of sound carefully arranged and spatialised. Navin Dolosawa's short tape piece “Elevator Music” was made using a gadget, designed for paragliders and the like, which gives out beeps which indicate whether it is rising or falling. Navin gave a talk about this, in which he explained that the aural cues from such a device are essential for paragliding.

Warren Burt's piece “Beneath the Slopes of Mt Corrimal” was a recording of a live performance by several computer-controlled microtonal instruments located at the Logos Foundation in Ghent, Belgium, and designed by Godfried-Willem Raes. Warren composed this piece by remote control, sending MIDI files to Belgium and receiving recordings back, The effect was of many music-box-like sounds arranged in a dense texture, with occasional clangs from a computer-controlled piano. The strange, almost wooden sounds in the piece came from a quarter-tone pipe organ with one bellows per pipe, which only gives out a short puff of air per note.

Warren talked about this piece, and indicated that although the remote composition process was successful for the instruments he used, he felt unable to use the instruments in the collection that have freely variable pitch, namely sirens and a computer-controlled musical saw. Warren considered that to learn how to use these instruments successfully it would be necessary to actually go to Ghent.

Of the live performance pieces, the highlight for me in terms of artistic quality was Brigid Burke's “Grit”. Brigid is a composer, performer and visual artist, and she combined all of these talents in this piece, playing her clarinet through an effects unit along with a DVD with recorded processed clarinet sounds and visuals consisting of largely of semi-abstract collages, together with a video of a wave ebbing and flowing over sand.

Several of the pieces raised questions about performance practice. The piece “Invisible Territory” by Mark Pedersen, with Ralf Mulhberger on shakuhachi, was a sonification of a historic game of Go. Mark and Ralf gave a talk explaining the principles of the sonification, which used spatialised shakuhachi samples to indicate the distribution of the pieces, and other information. The live shakuhachi controlled the tempo of the playback of the game, but the game was not projected during the performance. Several people asked Mark why not; the reply was that this was an experiment to see if the piece had a strong enough structure without such a projection. The answer would appear to be “no”.

Greg Schiemer (who unfortunately couldn't be at the conference) had his piece involving mobile phones (mentioned above) performed, the performers swinging the phones around their heads. For me this piece didn't work: it lacked the elegance of its distant predecessor, Greg's “UFOs”, small sinewave oscillators also swung around the performers' heads. There all the interest came from the effects generated by the swinging, while with the phones it was difficult to tell what was happening. The start of the performance was delayed because it turned out there was another Bluetooth phone active in the room (belonging to Greg Jenkins).

New to ACMA was the practice of “live coding” or “live programming”, the creation of the algorithm for the piece during the actual performance. This was bravely demonstrated during the last concert by the “Andrews Sisters” (Andrew Sorensen and Andrew Brown), using an environment developed by Andrew Sorensen. We saw Andrew Sorensen's Scheme code projected up as he typed it in, complete with syntax errors. Only basic constructs were used (no high-level predefined functions). Nonetheless the two Andrews constructed a complex ostinato, with some evolving structure. Opinions differed as to the point of the exercise; at least with this audience a substantial proportion of it could more or less follow the computer code.

Also new to ACMC (I think) is the practice of on-the-fly score generation, where the players (of conventional instruments) are sight-reading the material as the computer creates it. Two such pieces were performed. The first was by Andrew Brown, for five players. Each performer was reading standard notation from a computer screen, with the computer generating material a few bars ahead. Andrew acted as “page turner”, periodically pressing a button to display the next phrase; necessary, as the tempo was set by the performers. The actual piece was boring, being produced by a process with essentially no memory. The piece should perhaps have been held over until a more convincing compositional algorithm is developed.

The second piece with a score generated in real time was “Community Aid: Resonant Energy”, by American David Birchfield. It was brought to us by David's collaborator, percussionist Robert Esler. The composition process was much more sophisticated, involving a multi-level genetic algorithm generating a complex-looking graphic score. As well as Robert's live playing, the computer was playing back percussion samples. Robert was seated with a large cymbal on his lap and a collection of smaller instruments (shakers, bowls, etc.) at each side. This particular performance was fairly relentless, with little obvious variation. Robert, who gave an informative talk about this piece, said that the algorithm can produce a much more varied result. He also indicated that he had quite a degree of interpretative freedom, in particular regarding choice of instruments. I thought that the performance looked like someone trying to solve the Towers of Hanoi puzzle, involving moving objects among three piles, and perform a ballet at the same time.

The Conference Club

This year there was an off-campus Conference Club. Held in a small inner-city space, with wine and beer available, it provided a venue for laptop improvisation, starting at 8 pm or so, and kicking on to late. On the first night people did more or less listen, but as the conference went on the crowd and, in reaction, the playing got successively louder. My own laptop performance, which was intended to be mostly quiet, took place straight after the free drinks that marked the end of the conference proceedings. The crowd was on an end-of-conference high, the noise from the room poured into my mike, and my performance ended up completely derailed. Fittingly, the night ended with an ear-splitting feedback performance by Cat Hope from Perth. It actually sounded good from my position, outside the door wearing earplugs.

So this year ACMC offered us two different performance contexts, a concert-hall sit-quiet-and-listen and a noisy club. Some of the laptop gigs I have been to in Sydney offered an intermediate environment, moderately quiet, but with some freedom to move around and talk softly. Several of the ACMC “club” performances would have been suited to such an environment, my performance and that of Simulus (Steve Adam and Tim Kreger on this occasion) for two. On the other hand two of the “concert” performances didn't really fit in the concert-hall environment, namely the laptop impro by Aaron Hull and the performance by Botborg (Scott Sinclair and Joe Musgrove) involving audio and video feedback, treating audio signals as video and vice versa.

These comments about venues are not intended as a criticism of the conference organisers. The Conference Club generated its raucous dynamic spontaneously, and the provision of more than one performance environment reflects the broadening of the ACMC in recent years to include more than just traditional “academic” computer music. (This notion of “academic” needs revision; now, the more adventurous popular genres are comfortably accommodated in at least some academic environments.)

Certainly, we were presented with a quite a range of performance practices at the conference, from playback of acousmatic tape music to through to Cat Hope's feedback performance. Some of these performance practices have more or less established social conventions surrounding them. Two certainly do not, the “live coding” of Andrew Sorensen and Andrew Brown, and the two performances where the score was generated in real time. Since in these last two pieces the performer(s) had no influence over the score, it is a fair question to ask what the difference would have been if the score had been generated in advance; in other words, what the point of this performance practice is. The answer, if there is one, must come from the social relations surrounding the performance. Maybe the natural home for this practice is the computer game, not the concert hall.

Conclusion

Greg Jenkins and his team are to be congratulated on a very successful conference, the third ACMC to be held at Queensland University of Technology. QUT's support for ACMC, and for computer music in general, is something to be very grateful for.

© Gordon Monro 2005-6.       Last modified: June 1, 2006.
Home page:
Email: