Concert in Cyberspace

Have you ever thought about what possibly could happen, if every single one of your co-TV-watchers at your home had his own remote-control zapping just as you do? This was the initial idea of a project, which – after more than a year of preparation – I could present last January in collaboration with the Freiburg Experimentalstudio der Heinrich-Strobel-Stiftung, the Berlin company Chemie.DE Information Service and SWR (South West Broadcast German National Radio): INTErnetGEneratedRadio. The TV set was substituted by the radio-receiver and the remote control was situtated on the web.

There were 16 pieces of classical music from Bach to Bartok to choose from. The following rules (among others) applied:

      1. Is there more than one program selection (triggered by a left mouse button click) during an interval of 10 seconds, not the actual selected but the overall most selected (the “favorite”) piece of all participants will sound.

      2. Several clicks during an interval of 4 seconds will trigger the sound-processing-modus, which – dependent on the behavior of the participants – will re-synthesize the original material.

      3. Aren’t there any clicks for 4 seconds, INTErnetGEneratedRadio will switch back to the original-modus

Consequently, without any participation from the internet, the performance would consist of the pure performance of the introducing Beethoven String Quartet op. 132.

There were 2 aspects to be considered for the choice of the material: 1. traditional classical music is easy to be recognized. 2. I intended to overcome frozen barriers of musical styles and wanted to join together the classical music listener, the one from the academic contemporary music field, as well as the Experimental/Techno/HipHop/Ambient-programmer (both softwares I used to program INTErnetGEneratedRadio: MAX/MSP as well as REAKTOR are used in this scene).

The major task composing INTErnetGEneratedRadio was to allow the participant enough freedom to realize the consequences of his actions on one hand, and on the other hand, to create a framework, which guarantees a “senseful” musical result. I programmed 14 very characteristic effect-routes, which switched in intervalls of 30 or 60 seconds and could not be influenced by the participants. There were 3 categories of effects: 1. Effects modulating the original sound in real time like filter (the very steep digital filter of the Experimantalstudio), comb filter (created by a Lexicon device), envelopes, ringmodulator, pitchshifter, bitcrasher and vocoder (by Creamware’s Pulsar DSP System). 2. Effects using memory buffers (live-sampling) like several samplers and granular synthesis (I used Max/MSP for this). And 3. a very complex ensemble of the software REAKTOR by Native Instruments, which generated new sounds by the information of the amplitude of a single sample.

In contrast to the predefined effect routes all effect parameters, even for the recording and playing back of samples resulted from the click behavior of the participants. For this, not only the selected program button, but also the duration of the click (holding down the mouse button) was used as a source of information. The longer the duration of the click was, the higher the “expression” value was set. Additionally, there were special effect routes, which were triggered either by a specified frequency of incoming clicks or by certain numerical calculations. They could only last for up to 2 seconds and were inserted to loosen the hard cut structure which was imposed on the piece. To monitor one’s own behavior and the one of the other participants, the user was provided the information about the most frequently selected piece, the amount of clicks per seconds and minutes as well as the value of his last click duration (”expression value”). This was all presented and constantly updated in the browser window.

Technically, this was realized by JavaScript and CGI-Scripts, which provided also the data for the self-programmed webserver (by Hans Benedict and Barbara Wess from the sponsoring Berlin company Chemie.DE). The Berlin situated webserver then sent UDP packages to the studio in Freiburg, where  these were transferred into MIDI-data, which controlled the computer running the INTErnetGEneratedRadio –program. (For the complete setup, please have a look at the introducing picture.)

The problems arising with a project of such bandwidth are not only technical and aesthetic, but also consist of the implementation in present infrastructure, or even the creation of new infrastructures: INTErnetGEneratedRadio requires feedback by an audience, it even makes the feedback of an audience to a theme. During an usual broadcast transmission there is – opposite to a concert – no option for a direct response, and especially not one that could be realized by a performer in real time. On the other hand, INTErnetGEneratedRadio isn’t a concert, where there is a specially designed place for, like a concert hall with a subscribed audience. Even if the broadcast-station SWR (South West State Broadcast) is represented on cable in some major German cities, the transmission area is limited (although you could receive it world wide via satellite) and active participation only makes sense for houses with a radio receiver AND a computer with internet access in ONE room. There was a major PR campaign by SWR, sending out thousands of postcards, official press releases and emails to compensate with the issues above.

The performance of INTErnetGEneratedRadio (special credits to Reinhold Braig from Experimentalstudio and Hans Benedict from Chemie.DE) was highly successful: There were 35.000 clicks by estimably 500 participants during the 22-minute-piece, which equaled to a click-machine-gun-attack. Although INTErnetGEneratedRadio was prepared for such an amount of incoming information (in tests we generated up to 76.000 over this period of time on the Berlin server, whereby astonishingly enough not the inter- or SWR-intra-net proved to be the weakest point but the highly professional midi-interface to the MAX/MSP-computer) this behavior of the participants didn’t go along with the given rule, to respond to what is being heard. Some of the participants apparently developed some ambitions to approach the favorite piece, just like the highscore in a computer game. There was nothing of a crowd behavior, which would respond to musical developments with sometimes more, sometimes less and sometimes even no clicks at all.

By its characterization as a live-concert INTErnetGEneratedRadio opens a new field of artistic and social investigations, that was prepared e.g. by the Polish-American multimedia-artist Multimedia-Künstler Miroslaw Rogala in his PhD-paper (Strategies for Interactive Public Art) for interactive artworks in general. Rogala compares the behavior of single persons and groups of people in interactive spaces. It also is a continuation of the studies on crowd behavior by Elias Canetti. Is there – or can there be an isolated crowd or a social environment like an audience on the internet? Is it possible to define a space, which – analog to a concert – assembles an audience with generally known socially recognized behaviors and accepted rituals at a given time?  INTErnetGEneratedRadio managed to assemble an audience and it doesn’t make sense to complain about missing rituals in a first performance of its kind. But: a concert is more than the sum of all components, and its audience is more than the sum of its participants. Here shows the limitation of INTErnetGEneratedRadio: A real space allows the simultaneous perception of many events and creates a feeling of community. Virtual space appears not to be appropriate to build an audience in that sense. The information provided to the participants on their screens could not replace the crowd experience of a real room. There was no community. Democracy on the internet consists more of the tremendous possibilities of individualization than of a social enviroment.

Alternatively one could perform INTErnetGEneratedRadio over a period of 3, 4, or even more hours: then, there wouldn’t be a self-contained group of participants anymore, but more or less incidentally passing-by-visitors, which would stay for a while and leave. Everyone would contribute his part to the whole, without knowing the complete course of the piece. So it wouldn’t be a live-concert anymore but a work-in-progress.