ThoughtStorms Wiki

phil says : Tom wrote this in 1999 after mailing me about Gbloink!

Original at :

I've reproduced it here without getting explicit permission, but I hope Tom understands. It's just that I'm concerned that the original doesn't disappear leaving me without a copy to discuss. (CategoryCopyrightRisk)


Revelatory absurdism and finesse in new music software.

by Tom Rodwell, 1999.

Author's note: this piece is a little bit inconsistent, but there you go.

An anecdote?

Composer Conlon Nancarrow fought the fascists in Spain, but being refused a passport upon return to the USA he was forced to move permanently to Mexico City. He began to write densely polyrhythmic music that could only be performed by his collection of player pianos. For most of his life he remained unknown and unheard, living with his Japanese archaeologist wife, spending most nights tapping away at the hole-punching machine that created the ?instructions? on the player piano rolls. Over time his efforts produced a few music contacts in Europe and the USA, (among them Ligeti), and a massive left forearm. Very slowly, recordings of his works began to trickle out, and he passed away in 1997 having achieved a modicum of ?fame? in avant-garde music circles. It was in one such circle that he came into contact with German instrument builder Trimpin, who constructs, amongst other things, software-controlled mechanical percussion ensembles. Trimpin eventually performed Nancarrow?s work on tuned wooden shoes. This is a complex and serious level of absurdism.

The confluence of artistic procedures and technological experimentation has historically produced much intriguing culture. The intrigue currently continues in the field of music software. This is no glib cyber-industry promotion, for I would like to suggest that the most interesting results are produced by freeware and shareware programs, which exploit digital resolution as a kind of dadaism.

Absurdist culture (not only dadaism, it should be noted) was informed by the most ?avant-garde? practices, and by the political implications that culture implied (and implies). The riots caused by modern classical music (Stravinsky?s Rite of Spring etc?) focused attention on experimental music techniques as simple anarchism. The ?basics? of avant-garde music (odd-time signatures, non-tonality, complex structures and textures) continue to function as disruptions to passive listening, and absurdists continue to champion the anti-bourgeois nature of these ?basics?. This remains true, indeed it could be said that much new music freeware functions in the same way, as disruptions of passive computing, and indeed the dull conventions of ?normal? computer music environments. How?

Dada, and absurdism in general, pays attention to minutiae, (the dadaists were obsessed with print typefaces, photographs and syllables, while Albert Camus recommended a focus on the immediate, on worldly concerns). The binary nature of digital is highly appropriate for absurdist investigation, in that it affords microscopic representations and editing of audio information. Digital initiates an absurdist ultrafocus. Dadaist footballer Tristan Tzara said in 1924 "perhaps you will understand me better when I tell you that Dada is a virgin microbe that penetrates with the insistence of air into all the spaces that reason has not been able to fill with words or conventions". This is like the music that freeware programs generate and which ?fill up? the gaps in the desktop-music environment.

Professional music programs are, however, certainly impressive. One professional program boasts a MIDI resolution of 960 units per standard quarter note, (i.e. allowing for the accurate representation and editing of anything up to 3840th notes). Sound-file editing programs too have a legendary abilities: extreme precision and complexity is far easier than in analogue media. But these professional programs are fatuous by default, being necessarily trend-orientated, and produced with commercial viability in mind. The user is required to work in order to get past the preset effects or the bland strictures of the autoload screen. Not that this work aesthetic is bad, but rather that the immediate conservatism is simply too easy, and too concealing of the implications of digital ultrafocus.

Shareware and freeware programs, however, are instantly commercially useless. Their intelligent yet disruptive nature marks a strong antagonism, roughing up the bogus professionalism with which computer musicians like to coddle themselves. Independent writers focus on gaps in the market, and more often than not, these are the bizarre musical desires of bizarre musicans. Niche ? focus has created an anarchic culture of obscurity in software.

Freeware, shareware and experimental programs are often the hobbies of the writers. They are playful and musically intelligent. But they can be unstable, confusing, and obtuse. This instability is suitable, however. Tzara again: "we are often told that we are incoherent, but into the word people try to put an insult that is rather hard for me to fathom. Everything is incoherent. The gentleman who decides to take a bath but goes to the movies instead. The one who wants to be quiet but says things that haven?t even entered his head. Another who has a precise idea on some subject but succeeds only in expressing the opposite in words which for him are a poor translation. There is no logic. Only relative necessities discovered a posteriori, valid not in any exact sense but only as explanations".

Absurdist confusion and illogic parodies normal sensibilities and methods, (rather like the Dadaist anti-advertising agency). The software writers themselves are eclectic and informed, and often work outside of the normal software industry, (one I spoke to engineers synthetic DNA).

NYC?s Nick Didkovsky described the industry, and the position of small writers: "Two of the crushing influences on commercial creative software are 1) deadlines and 2) the fact that someone pays real money for it. Deadlines: If I have a program that needs to go out next month, I will spend the time bullet-proofing what's there, and removing out stuff that only-sort-of-works, rather than adding cool new ideas. Money: I am much more apt to put risky algorithms in software I am giving away for free. Then I won't be (rightfully) hassled by someone who uses it, crashes, and then wants to know what they spent their money on. These are very different activities: hacking for one's own creative growth, and programming commercial applications. Of course there's some creative overlap, but market rules change things." Quasi Fractal Composer author Paul Whalley: "Freeware writers have the distinct advantage that they can develop their works without compromising their ideals. Such creative and intellectual independence is of paramount importance in the arts. If there is an "ethos" among them, I hope that it is only to maintain the purity of the art." Philip Jones, author of Gbloink!, noted that "its always good that if you have a great idea, you aren't obliged to justify it to some manager with the imagination of a brick before you can start work on it."

From a technical viewpoint too, independence is important, even ethical. Didkovsky described the importance of "building your own tools instead of settling for the prejudices and implied aesthestics of off-the-shelf music software." But this independence can be compromising and limiting as Swedish writer Rasmus Ekman noted: "the private shareware writers may be somewhat helpful in spreading some of the tools to a wider audience, but we usually don't do much to develop the technologies themselves. Perhaps this will change, but then it will also depend on what you regard as progress."

Indeed, is there progress in this field? (And would it matter if there was?) Can small-time hobbyists and intellectuals really affect change in the home-computer arts sector, given the power of the pro big names? Perhaps a discussion of specific programs is in order. I talked to three writers about their programs: Rasmus Ekman from Sweden, Paul Whalley from the USA, and finally Philip Jones from England.

Programs? The complicated processors in home computers are too often used for a kind of ineffective replication of analogue musical technologies, (guitar effects, drum machines, analogue synthesisers). Rasmus Ekman, a software writer who himself has a long history of working with unusual analogue synthesisers, pointed out the digital replication problem. "MetaSynth might be an example of an original tool from an independent developer?but then some people might not regard it as significantly new: "just the good old massive additive synthesis, we had that in '76 etc?." MetaSynth itself is distributed by IRCAM, famous home of Pierre Boulez, who once said in an interview "We have a very important direction which is to make more accessible the interface: man-machine. This involves a higher level of language, language which is more symbolic and less numerical. For instance, if you see a curve it means more to your imagination than figures. Also ?real time? work is important, so that you can modify the field of sound during a performance?"

Ekman may have had Boulez's words in mind when he wrote Granulab and Coagula. The former is a granular synthesiser and sound-file processor, and it is as a processor that it is perhaps at its most intriguing. A file, for instance 10 seconds of someone singing, is continuously looped, whereupon the user can, in real time, force the file to randomly play back from any point, for any length of time, at any pitch / speed, and at any stereo-placement point. The new version allows for the playback of up to eight files simultaneously, (assuming ones? processor can take the strain). Ekman himself has a dry sense of humour when discussing his programs, which as ?software synthesisers? should theoretically be rather fashionable. "There are several facilities for tonal / beat-orientated work which could be said to be ?missing??(which) would have been obvious and necessary for anyone with pop music in focus? Both my programs lend themselves quite freely to pop music applications, but I?ve mainly had?modernist music in mind when designing them." The results of Granulab are unrelentingly avant-garde, since the "important?aim I?ve inherited from?modernism is to open up unexpected connections in sounds." This focus on new relations, unusual connections, based on granular (grain) synthesis makes the program forcefully experimental. It is hard to hear the results in any other context.

Coagula is itself derived from the tools used in Metasynth, and is a ?colour note organ? that generates a sound-file of synthesiser sounds based entirely on graphic images provided and manipulated by the user. One can ?draw-in? circles, squares, lines, etc? , or import a complete digitised image, and then filter or boil the image, editing it in various obscure but attractive ways. The program can then be instructed to generate a sound file from the information of the image. The results are extremely odd, both in a strictly musical manner, but also in their suggestiveness: the coagulation of image and sound almost exclusively points out "unexpected connections".

Ekman points out that "figurative pictures will usually not connect with the output sound in any interesting way _ but do whatever you like." Because interesting musical results do not come from standard visual-arts techniques (like representation) the program enforces a purposeful abstraction. Ekman?s wry humour arises again: "The attraction of the program, if any, lies partly in the tools used to edit the image. The point of the tools is of course to facilitate image editing which will be sonically meaningful. Thus, there are some features less common in other image editing programs (or never seen before). Lots of stuff that would be really useful is also missing." This use of niche tools for the program's surreal connexions is deeply radical, and perhaps it is the moment of this coagulation that is most revealing. Ekman?s programs transform the ?professional entities? of wav. sound files or bmp. picture files into absurdist insinuations. Quasi Fractal Composer is written by Paul Whalley, who also sees himself as functioning within the avant-garde tradition. The program generates a MIDI file based on mathematical decisions, influenced by user determined "seeds" (or structures), and employing stochastic procedures to manipulate / randomise the ensuing results. Whalley describes the procedure thus: "(The) structure is a string of from 3 to 16 digits. This string is the basis of the self-similar structure that directly, or indirectly, controls or influences most parts of the composition." There is also a mapping function: "This control seeds the program's random number generator (when the Compose button is pressed), thereby determining the values of all randomly set parameters inside the program." Thus the user provides the (serial) raw materials, and the program initiates the manipulations and variations of the resultant structure and its parameters. QFC also allows for various other user manipulations of the "self-similar structure", producing interesting results that combine informed structural abstraction with an improvisational subtlety. The method initiated by QFC is interesting: "the ?conceptual angle? that marks my work is the hybridisation of musical methodologies. I believe that no single approach (e.g. fractals, serialism, algorithms, stochastics) can suffice in the computer-generation of non-trivial compositions. The key is in developing a rational interplay of diverse musical tools."

This interplay is often musically astonishing, with complex polyrhythms recalling Nancarrow, structures recalling Xenakis, and serial harmonies all easily, indeed immediately produced. Whalley?s musical influences have clearly bled into his software writing: "Schoenberg's twelve-tone system and the approaches of the later serialists have had an intense influence on my compositional methodology. Stravinsky's use of linear counterpoint and Zappa's peculiar adherence to musical form have also worked their way into my musical psyche." Interestingly enough, Whalley is not actually a musician himself. QFC is his composition.

Though heavily informed by serious methodologies, QFC?s format is necessarily open to absurdist experimentation. When entering in the initial numeric sequence(s), the user is able to type in illogical non-patterns, telephone numbers, lottery numbers, the date. The musical results are however consistently intriguing as investigations of compositional processes, rather than the non sequiturs generated by strictly algorithmic programs. The music produced by QFC can be saved and edited by other (MIDI) programs, allowing for a retrospective analysis of the musical results, including those generated from the barcode numbers from a packet of crisps.

Paul Whalley's program differs from the dry academic fractal-music programs in that while it employs and is aware of strict compositional procedures, its format allows for a playful and translucent experimentation. It is uncanny in its ability to generate useful (non-trivial) material from both intellectual and absurd procedures. Whalley described his attitude to academic computer-music: "Most of the?experiments that I?ve heard have left me cold?none were particularly satisfying. Most have only fuelled my arrogant belief that I could do better". This view is relatively common among freeware writers, who often have as much antagonism towards academic environments as they do to their commercial equivalents. Both dictate a climate of seriousness and professional respectability. This climate is being slyly (albeit slowly) altered by freeware. The theme of useful software, creating useful and interesting music is an abiding concern among writers. Philip Jones is the author of Gbloink!, who has also pondered serious compositional procedures: "I've been thinking about applying genetic algorithms (artificial evolution) to music. However? I am stuck on the interface side of things. I can imagine lots of ways which are tokenistic, ?yeah let's use genetic algorithms to compose music? but I need to find a way that would really get an interesting result while allowing users to properly interact." Gbloink! is at once fascinating and useless, a potent dadaism, that learns from, extends, then derides both academic and commercial music software procedures and functions.

Jones drew on his graphic and computer-game design experience in creating Gbloink!. It functions in a similar manner to a game, the user attempting to manipulate the path of several drifting balls flying around the pitch area, (pitch not only as a playing area, but frequency). The height at which the ball strikes the wall determines the pitch of the triggered MIDI note. The user can, with one click of the mouse, enter randomly-sized blocks into the path of the ball(s). Again a MIDI note is emitted at the level in which the ball and object make contact. One can also change the MIDI instrument assigned to specific balls, and the speed at which they move, while another function allows for changes in scale. The random nature of the program is exacerbated by the fact that none of the ball-controls are labelled, resulting in much nervous tinkering on the part of the user. The balls also chip-away at the blocks they hit, eventually clearing a path into unknown regions.

Jones deliberately configures Gbloink! on the fine line between intelligent experimentalism, and absurdist play:

  • Gbloink! is not a serious composition tool. It's a very silly composition tool...
  • Gbloink! is not a game. It's a toy. Games have objectives, notions of winning and losing. Gbloink! has none of that.
  • Gbloink! is not educational. Don't trust us on major, minor, diminished and eastern scales. We just guessed 'em up. And don't think we claim Gbloink! has the power to demonstrate anti-chaos or self-organization just because so many seemingly chaotic worlds quickly fall into cyclic structure. Nor does it tell you anything about the n-body problem or the dynamics of real-world objects. :-)

Meanwhile the intro screen proudly declares: "Once Upon a Time there was a pinball machine that wanted to be Ravel, Satie, Eno, Birtwistle, and MIT Media Lab?s Hyper-Instruments project. All at the same time?"

The music too, treads a fine line. Other real-time random music programs (as differentiated from random music generators like QFC) are often ambient, gentile, constrained. Gbloink!, however is immediately musically perverse. The game-like interface initiates paranoia in the user, who is forced to constantly determine, and repair the block-environment of the balls. Too-high or too low, and the MIDI sounds are muddy and musical. Too-small gaps create annoyingly blurred repetitions. The user is forced to update, to work, (moving ever closer to Repetitive Strain Injury with every mouse click). The user-patrolled environment, while often focused and responsive, can also be annoying. It forces an awareness of the user intentions: ?is this too tonal? Is this too fast? Is this too clustered?? These attentions are consistently disrupted and modified as the balls break through the block-walls.

While the musical results are patently game-like and absurd, there is also a striking sense that the music generated is employable. This author has, in fact, recorded large amounts of free improvisation involving other musicians and the program. There was an even more intense ?user-paranoia? during these interactions, for not only did I have to improvise myself, and control the blocks and spaces on screen, but also prevent the generated music from becoming too repetitive or tonal. The interface forces improvisation, and has the same attention to "continual refinement and adjustment" as live music-performance.

This attention to use is crucial: "the interactive experience, i.e. tightness of the feedback loop between the program (generating) and the user (testing) was the main concern." Jones sees his program (and others) as useful for their influence on ideas and procedures, like the way that chance events are co-opted into the methods of composition, improvisation, DJing, recording. "It seems more sensible to try to understand this music ?ecologically? than through an analysis of its grammar."

The flexibility and humour of Gbloink! is representative of Jones? playful approach: I'm increasingly attracted to a more low key, quirky, mood rather than a dramatic or grotesque one. Gbloink! also has a lot of influence from the aesthetic of children's toys and games; something of the Nintendo cute look; and I was blown away by the Teletubbies when I first saw them. No programme I'd seen before had this aesthetic; this mixture of intense unworldly colour with rural tranquility. Being within a childish aesthetic mood does allow some freedom. You feel less constrained to have to explain yourself; or make everything consistent."

Gbloink!'s music functions at once like a baby's mobile and a free-improvising percussion ensemble. Nursery-rhyme style repetitions evolve into crunching clusters, then twinkling serial spirals, then nursery-rhyme style repetitions...

And inconclusion...

The music produced by music-software programs, (and explicitly so by freeware / shareware programs) divulges much about employment of culture. The interface-screens can be tinkered with in a child-like, inquisitive manner, while the music that results is necessarily radical and flippant. The procedures and results of freeware programs thus bear great similarities to experiments in serious (and especially live) music. Indeed, those musics are simultaneously useful and useless. If I may be rather flippant myself here, I would like to suggest that this uselessness is its prime use, for by being crucially and consistently radical, these programs and their results reveal the inadequacies of software and music that is supposedly ?useful?. Freeware programs make valuable specific contributions to the actualities of music, but they also parody the Important Programs and thus Serious Music. Freeware is a protest against the crass dryness of that which is useful, or professional, or even artistic, for it degrades those sterile environments and introduces the user to absurdist alternatives. Like Chindogu, the Japanese art of unuseless inventions, freeware is a dadaist reclamation of cultural procedures from the bogus job-kultur. The boring paranoia inherent in computing (Isn't that article due soon? Am I running out of memory? Is my shoulder freezing up?) is relieved by the music that freeware creates. I like it.

This material is to be considered copyright 01999 by Tom Rodwell. No commercial republication is allowed without specific written consent from the author and copyright holder.