Art from code - Generator.x
Generator.x is a conference and exhibition examining the current role of software and generative strategies in art and design. [Read more...]
Sound works

The current popularity of sound art and generative sound installations (whether mechanical or software-based) makes sound an exciting area to work in. New temporal and spatial approaches to sound are being explored outside the traditional venues of sound consumption.

abstrakt Abstrakt: Zimoun

Zimoun: 150 prepared dc-motors, filler wire 1.0 mm
2009/2010, 730×120cm

Zimoun creates complex kinetic sound sculptures by arranging industrially produced parts according to seemingly simple rules. Using motors, wires, ventilators etc., he creates closed systems that develop their own behavior and rules similarly to artificial creatures. Once running, they are left to themselves and go through an indeterminable process of (de)generation.

These quasi autonomous creatures exist in an absolutely synthetic sphere of lifeless matter. However, within the precise, determinist systems creative categories suddenly reappear, such as deviation, refusal and transcience out of which complex patterns of behavior evolve.

Read the rest of this entry »

abstrakt Abstrakt: Leander Herzog - Sound Object

Leander Herzog – Sound object (data.matrix, Ryoji Ikeda), 2010
Laser cut plastic

The human ear perceives sound as an intangible presence, produced by vibrations travelling through a physical medium (air, water, even solids). Recorded sound is produced by measuring these vibrations for later reproduction by mechanical means. What we experience as harmony and rhythm appears to the computer as a one-dimensional number sequence.

Leander Herzog’s data sculptures perpetuate this disregard of the emotional dimension of music, looking at the sound input simply as a frozen space of random-access data. Herzog replicates this data stream by perforating plastic ribbons at intervals matching amplitude measurements, producing ornate collections of short and long loops. The result is a curious data artifact, numerically correct but completely disconnected from its own origin as sound.

This text is taken from the NODE10 catalogue, written by Eno Henze and Marius Watz and edited by Valérie-Françoise Vogt. Please read the introductory curator text for an overview of the exhibition topic.

abstrakt Abstrakt: Leander Herzog - Sound Object

Andy Huntington: Cylinder

Andy Huntington/ Drew Allan: Cylinder (”Seahorses”, “Designed”, “Market”)

Cylinder by Andy Huntington and Drew Allan is an elegant series of data sculpture based on sound analysis. A mapping of the frequency and time domains produces cylindrical forms representing the spatial characteristics of the sound input. Physical versions of the digital 3D models are then 3D printed using stereolithography.

The idea of mapping sound to space is not unfamiliar. The Cylinder project shows similar strategies to those used in the exhibition Frozen, which showed sound represented as a continous space rather than as a one-dimensional signal. However, Cylinder is from 2003, predating Frozen and making it somewhat of an early example of the data sculpture genre.

There is a tangential similarity between Huntington’s pristine objects and Booshan & Widrig’s Binaural object. But in fact the spiky geometries of both works are a result of the numeric data underlying the form. Any data set will yield inherent patterns, and in the case of digital sound two “defaults” present themselves: The waveform (a 1D graph) and the spectral map found through FFT analysis, which represents a 2D map of spectral energies in the time domain. Any translation of these numeric representations into visual form must grapple with the fact that while they may be faithful representations of the data, they rarely give a good idea of how the sound is experienced by a human listener.

The Cylinder series show a range of different waveforms, some showing an apparent orderly structure with others suggesting a noisier sound input. Titles like “Seahorse”, “Design” and “Breath” imply the source sounds used to produce the forms. Their success as aesthetic objects derive from their complexity as well as from the clean quality given by the 3D printing process.


Skot (Frank / Gmachl): aka (audio by General Magic) from Tina Frank on Vimeo.

We have posted about the Vienna scene and the Austrian Abstracts here on previous occasions, but the video work that was central to that movement has generally not been available for viewing online. Therefore, it’s with great pleasure we see that Tina Frank has posted some early videos to Vimeo. Let’s hope other artists follow her initiative, it would be nice to have an online archive of these early experiments somewhere.

Shown above is the video AKA by Skot, produced for Gasbook 4. Skot was the name used by Tina Frank and Mathias Gmachl for a number of collaborations from 1996 to 2000. Gmachl is also one of the founders of farmersmanual, a collective that was central to the Vienna scene. “Aka” means “red” in Japanese, and the video was made with Image/ine software from Steim, one of the very first softwares to support realtime processing of video on a regular computer.

Frank created the video "iii" below by taking digital audio files of the music by Peter Rehberg (Pita) and opening them as raw pixel data in Photoshop. An oval image mask was superimposed, giving a more specific form to the resulting video. The result is classic glitch, taking a signal of a given form and deliberately misinterpreting it as something else.

More videos on Tina Frank's Vimeo stream.

Tina Frank: iii (audio by Pita) from Tina Frank on Vimeo.


CTM.08­ / Generator.x 2.0: Audio-Visual
Fri Feb 1st, 20:00 – 23:00, Ballhaus Naunynstrasse

  • Alexander Rishaug / Marius Watz
  • Keiichiro Shibuya [JP]
  • alva noto [DE]

CTM.08 and Generator.x present an evening of audiovisual concerts, consisting of three projects that use generative methods for live performance.The artists’ work is based on program code that integrates processes that develop over time autonomously. Music and image are therefore not “composed” in the usual sense of the word; the artists at most structure audio and visual output but without determining its every detail. The visual and acoustic material available to them is not pre-processed; it exists rather, only as matter to be subjected to and modified by certain development principles and rules of transformation in real time. Therefore chance and stochastic processes are major factors, while images can be mapped on audio parameters, and vice versa.

Alongside Alexander Rishaug’s audio miniatures in interaction with Marius Watz’s drawing machines, and Japanese Keiichiro Shibuya’s digital noise based on cellular automata, Carsten Nicolai aka alva noto will present his new project, “xerrox”, in which he explores the artistic potential and unpredictable results of copy processes. Nicolai demonstrates that the act of copying is itself a source of interesting, artistically valuable mistakes and mutations that permit each new generation of a copy to further liberate itself from the original and ultimately become an independent artwork with new meaning. In “xerrox” alva noto works exclusively with samples of Muzak – the wraparound sound ubiquitous in department stores, advertising, film scores and entertainment software – yet uses his own specially developed copying techniques to alter its melodic (micro-)structures beyond recognition.

Generator.x 2.0: Audio-Visual was curated by Jan Rohlf of the Club Transmediale.

[text from Club Transmediale]


The early-to-mid 1990’s were an interesting time. “Multimedia” was a hot buzzword, and people were wondering if CD-ROM and Internet was here to stay. Macromedia Director ruled the world of interactive graphics, and World Wide Web and HTML was finally transforming the Internet into a visual environment.

Early experiments using the web for art purposes quickly became iconic: Jodi hacked HTML, Form Art was briefly defined as a genre, considered ironic approaches to art production via this new channel and artists like Stanza explored Director as a tool for generative graphics.

During this (golden) period, Vienna was a hotbed of experimentation. A large group of artists pushed the boundaries of abstraction in visual art as well as music, often experimenting with code-based tools. It should be noted that the term “generative art” was not in use at the time. Nevertheless, the work produced at the time clearly articulated generative and procedural approaches to sound and image synthesis, prefiguring the current interest in such work.

Among this loosely affiliated group were artists like Farmers Manual, Tina Frank, Monoscope, Pure, Lia and Dextro. The music label MEGO and the film label Sixpackfilm provided publishing outlets. Norbert Pfaffenbichler put together an overview of the scene in the exhibition Austrian Abstracts in 2006, which expanded on the previous exhibition Abstraction Now, focusing specifically on the activities of Austrian artists.


Dextro: Turux piece / c079

Early pioneers of generative Director programming, Lia and Dextro quickly became influential both inside and outside the Director community. Their mix of crisp pixels, erratic animation and blurred surfaces was unique at the time, presenting a perfect visual counterpoint to a musical scene experimenting with glitch and sound defects.

Together, they produced Turux, a seminal web site which featured Director “soundtoys” and generative visual sketches. Thanks to the site’s intentionally cryptic interface design and the “anonymous author” fad popular with the Vienna artists (many of which used pseudonyms or group names), the authorship of Turux was unclear to outsiders. Often, visitors had no idea if Lia, Dextro or Turux were actual people or just project names. Nevertheless, Turux became an important reference for the nascent scene, its fame only heightened by its obscure origin.

When the collaboration ended some time later, Turux remained online practically unchanged. As a document of a specific time period, it became a time capsule of styles and strategies.

The original is now offline for good, having been replaced by a placeholder. But Lia and Dextro have both set up their own archives. Lia recently launched, a partial archive of her half of the project. Included are 21 works in Director, documented as stills and interactive Shockwave movies.

Dextro’s Turux experiments have been integrated into, which presents his work chronologically organized from his early period up to now. See the Turux subpage for a list of sketches. For an example of his newer work, see c079.

Seelenlose Automaten (VVVV video)

Schmidt / Groß: Seelenlose Automaten

Seelenlose Automaten is a generative music visualization with an unusal approach: Instead of analyzing the music for cues for visual change, MIDI control messages are sent simultaneously to the sound and image generators. Each mapping to a specific visual or sound effect, these messages are a vocabulary of rules giving structure to the composition. All change can be precisely predicted, and as a result the entire composition is perfectly synchronized.

Created in VVVV by Patric Schmidt and Benedikt Groß, the video uses minimalist forms to good effect. Polygons, lines and circles twist and turn in a gray void, responding to the smallest change in sound. Lacking depth cues, the images frequently read as flat 2D, only to become 3D once again upon the next movement. The total impression is of a glitch aesthetic, even though the deterministic nature of the system is antithetical to the glitch philosophy of creative breakdown.

Benedikt Groß is a student at the HFG Schwäbisch Gmünd design school in southern Germany. He recently completed an interactive installation presenting the principles of generative systems.

Ps. Three times is a charm: This is the third post in a week about a project created with VVVV… VVVV might still not be as wide-spread as Processing or Max/MSP, but there is certainly some very high quality work being produced with it. It will be well worth watching in the near future.

Bibliodyssey: Musical notation

From BibliOdyssey: George Crumb: Makrokosmos I / Barry Guy: Bird Gong Game

BibliOdyssey is a wonderful blog that deals in archive images from obscure sources (usually old books). Typically, it presents old scientific diagrams, pattern samples, anatomical studies, ancient maps or just anything that has a strong visual attraction combined with a sense of the obscure and arcane. All told, it is a delightful image resource for anyone with even a slight sense of the magical.

Today’s post on the visual context of music is of potential interest to Generator.x readers. It deals with unconventional visual forms of musical notation, from the illustrative to the conqrete, from the ancient to contemporary. It should prove intriguing and well worth the time to indulge in both the images and links provided.

Simon Elvins: Mapping sound

Simon Elvins: Silent London (detail) / Notation

Simon Elvins is concerned with sound as an ubiquitous force. Through a series of projects he has been documenting how sound is an often ignored dimension of our physical environment. Silent London plots quiet spaces in the English capitol using noise level data. An embossed print shows quiet areas raised up from the paper, bringing them to the attention of the viewer, while noise areas become blanked out valleys. noisy areas raised up from the paper while quiet areas become blank areas of peace. His FM Radio Map serves a dual purpose. On the one hand it plots the physical locations of commercial and pirate FM radio stations broadcasting in London. But circuits conductive pencil lines placed on the back of the map also turns it into a physical interface. Using a modified radio the map can be aurally “navigated” by placing metail contacts on points on the map.

These projects are poetic but ultimately functional. Taking a conceptual design approach (Elvins studied Communication Art & Design at the Royal College of Art), they present numerical data in an aesthetic context. By choosing low-tech materials (paper, electronics) Elvins creates fragile objects whose material qualities belie their sophisticated technical content.

Parallel to Elvin’s interest in sound is his fascination with mapping of physical and intangible forces. Both the aforementioned projects are classic mapping projects, while Notation is a more abstract exploration of how sound can be represented visually as marks on paper. Reminiscent of experiments with graphic notation (see Eno etc), the project consists of studies of representations of tonal patterns using pencil on paper.

The Notation project page seems to indicate that these drawings can ultimately be used to produce sound, but no details are available. If so, it would be an inversion of Elvin’s excellent Paper Record Player, where he constructed a functional record player out of paper, complete with its own conical paper amplifier.

(Thanks to TomC. See also Mount Fear.)


Vague Terrain: Jeremy Rotzstain / Paul Webb

The online digital art journal Vague Terrain launched Vague Terrain 03: Generative Art last week, featuring work by 11 artists and musicians. Edited by Greg Smith and Neil Wiernik, Vague Terrain asks contributing artists not just for works (images, videos, mp3s), but also to articulate the context those works exist in. The final result is both artistically challenging and theoretically weighty.

This issue explores generative strategies in art, an approach to image production as well as for sound and live performance. From the abstract images of Meta to the image producing machines of Jeremy Rotzstain or the random music systems of Paul Webb, the works presented show a range of possible uses of generative systems. Of special interest are the many examples of sound works. Lately the generative art field has gotten most of its attention for visual art, belying the rich history of such systems in musical composition.

The most important theoretical contribution is a new paper from Philip Galanter, in which he clears up some common misunderstandings about his by now canonical definition of generative art (see this interview). Entitled Generative art and rules-based art, it traces a genealogy of generative art, presenting a number of historical works that are generative without being computer-based. Specifically, he looks at two exhibitions (Logical Conclusions: 40 years of rule-based art at Pace Wildenstein, New York and Beyond Geometry at Los Angeles County Museum of Art). Using these as a reference point, he clarifies some often-misunderstood terms and analyzes how the works in these shows can be understood in relation to a generative tradition.


Vague Terrain: meta / Ben Bogart

Galanter’s presentation of these precedents is thorough. In focusing on the principles used to produce the works, he presents a firm platform for defining a basic terminology. The downside is that his paper does little to connect the aesthetic content of historical works to the works typical of the generative art field today. This is both a strength and weakness. In looking only at a limited formal aspect, Galanter avoids getting bogged down in art history. He can thus compare works that would otherwise be seen as representatives of different art movements. But in doing so, he risks giving the impression that a work like Alfred Jensen’s The Apex is Nothing is concerned with the same issues as Salavon’s Shoes, Domestic Production 1960-1968.

This dilemma clearly illustrates the problem of generative art as a definition that only describes a methodology. Apart from identifying an interest in systems, it says nothing about the resulting work and so constitutes a tenuous common link. Uncritical use of the term risks conflating artists from different periods and assuming that their artistic interests are the same, when in fact the contexts in which they produce their works are very different. Philip Galanter is no doubt aware of this, and likely does not intend to make this implication. But the recent historicization of generative art has tended towards ignoring differences in the desire to find similarities.

Finding historical references for generative art has the potential effect of legitimizing the current work by placing it in an art historical context. But it also risks ironing over the unique qualities of works. In the long run, the current generation of artists working with code and software as artistic materials might do well to specify their position as being separate from this historical context. This would allow them not only to state what is unique about their work, but also to be understood by the art world at large as more than an updated appendix to work done in the 1960’s.

All texts from Vague Terrain are under a Creative Commons license and are available for download as PDFs.