My Colleagues‘ News

Wolfgang Muench 09-04-2025
Bubbles: The Art of Simplicity

Bubbles constituted a rather small project that turned out to be enormously popular with both audience and museum curators. Bubbles was first exhibited in 2000 in the beautiful but rather unaspiring exhibition venue at Schloss Wahn, a small idyllic baroque palace that housed the Theatre Archive of the University of Cologne. I had conceived the initial concept while driving on a dark highway in the winter of 1998, and the first drawing outlining the technical set-up was sketched on a paper coaster mat in a bar, in a rather amusing tribute to the legacy of fine art traditions in media arts. Without the persistence of my Cologne theatre friends, the project would probably never have been realised. A wider audience was reached a few months later when Peter Weibel, Director of the ZKM, accidentally witnessed a prototype testing on the corridors of his institute and decided that the installation should be shown prominently in the ZKM Media Theatre for an open-house night jointly organised by all museums in Karlsruhe.

Bubbles is a multi-user computer-based interactive art environment that enables participants to interact with a real-time simulation of floating bubbles. By entering the light beam of the data projector, the participant casts a shadow onto the projection screen (1). A live video camera captures the display area, and a standard Apple PowerBook G3 evaluates the video signal through a custom-built machine-vision system. The system can detect any intersection of one of the computer-generated bubbles with any part of the shadows and computes a corresponding response for the behaviour of the bubbles. The bubbles bounce off the shadow’s silhouette and create a corresponding sound according to a sound-action design defined by composer Kiyoshi Furukawa. Defined as autonomous objects inspired by Braitenberg’s definition of a complex system of communicating objects with very simple internal structures (Braitenberg, 1984), the bubbles‘ behaviour and their response to any user interaction is based on a set of simulated physical laws following simple artificial live rules. Both the overall state of the complex system and the shadows’ interaction with virtual bubbles create nonlinear musical structures generated in real-time utilising a midi interface and midi synthesiser (Muench, 2000).

This unassuming and casual beginning of the surprisingly long record of more than 50 international exhibition venues that presented Bubbles since 2000 corresponds well to the unpretentious impulse that led to its development in the first place. The initial idea was less to create an artwork, although this might arguably always be in the back of an artist’s mind. Instead, the idea was to define a human-computer interface that uses the human body without further technological extensions as an input device. The work should run on an affordable and low-tech computer system, and should not require the studying of any ‘how-to-do’ manual, such as most interfaces at that time required the user. In essence, the technical solution that came into my mind during that night on the highway was to reverse the internal mechanisms of evaluating a video signal.

Standard motion tracking software at the time was essentially operating using the approach of evaluating each frame of a video signal in its entirety. The programme would analyse the colour value of pixels as the core element that composes a video signal, and then attempt to recognise pattern among these signals. Such a pattern, in the simplest mode clusters of dark pixels against a bright background, is then represented as a simple shape circumscribing the cluster area. Providing that the colour values would not change too much or too quickly, the computer can then process a position of this pixel cluster on each video frame, and compute information about the change in location, or intersection with other clusters. It is important to emphasise that the computer does not have any idea about what real world object is represented by coloured pixels in the video signal.

At a computational level, the evaluation of such a pattern of colours necessitates a complete scan of all pixels. The standard PAL video resolution of 720 pixels horizontally and 576 pixels vertically, a common video resolution used in interactive artworks of the late 1990s, results in the assessment of more than 400,000 pixels for each frame. At a regular rate of 25 frames per second, it required not less than 10,000,000-pixel operations for the colour value assessment of one second of a video signal. The considerable number of operations necessitated the deployment of a very fast and very expensive computer for the calculation of much information related to the video signal, although most of the laboriously derived information was entirely pointless for the software program. In most cases, the artistic concept demanded not for access to data of the entire representation of space on video, but only to small details where relevant activity is happening. Unless, of course, the concept was about the utilisation of an ‘intelligent’ vision system similar to human cognitive perception, in which case the whole concept was in danger of severely exceeding the scope of existing computer technology.

The artistic-technological concept of Bubbles was based on the idea that there is no point in computing data just to throw it away. All that was needed was simple information about when and where one of the virtual bubbles intersected with the silhouette of a human body, to calculate a corresponding, and physically plausible, rebound of the bubble. Therefore, it appeared to be much more reasonable only to assess colour values around each bubble, and not to concern the computer with anything that happened outside the clearly defined event of one bubble hitting the human silhouette. This concept allowed for a massive reduction in data computation. Only eight points around each bubble were identified, for a maximum of five bubbles simultaneously on display, for which the system evaluated their colour values as to whether they are above or below a certain threshold. The latter gave an unambiguous indication that there was an intersection with a silhouette. This simplification of the evaluation process decreased the required pixel operations to 40 per frame or a maximum of 1000 per second. Even a rather powerless, as to present standards, 1998 Apple PowerBook could handle this. Bubbles thus abandoned the artificial intelligence concept of ‘knowing’ the captured spatial environment in favour of a machine-vision system that was utterly oblivious to the overall situation in real space, except for a few precisely defined and easy to detect events of further relevance for the program. This ‘inverted’ concept for the gathering of spatial information also meant that the processing of the video signal and its display on the computer screen was not necessary, which again reduced the required processor time significantly.

The technical concept capitalised on the properties of the shadow, a familiar cultural icon in its own right, for the embodiment of the user within the hybrid environment. This approach eliminated intangible alienation of users in the process of interacting with the artwork, which often is the case when the representation of users in the virtual environment is a result of digital processing. Even a small delay in the users display in live video camera footage provokes a sentiment of incongruity between real and virtual space. In Bubbles, the fundamental physical laws of optics accomplished the displaying of the user’s depiction on the screen, a process that by definition can only happen in real-time. The interplay of the data projector, the user’s body, and the screen itself served as an 'analogue computer' that computes the size of the shadow on the screen. The inclusion of the familiar physical qualities related to spatial relationships between the shadow and its environment crucially contributed to the overall experience of the work. In a subtle manner, Bubbles addressed the aesthetics of embodiment, a topic that presents an on-going challenge for virtual reality environments, as Penny remarked (Penny, 2010, p. 179). The human body, an element often ignored in human-computer interaction, is central in this artwork for the facilitation of interaction with the software application. It presents an elegant solution for the representation of the human interactor in a digital environment without actually ‘digitising’ them. The main component in the human-computer interface remains firmly attached to the ‘analogue’, real world, although it forms part of the digital machine-vision system since it would not be recognised by it otherwise. The two-dimensional space of the projection screen constitutes the stage for a playful celebration of an analogue-digital encounter. Computer simulated bubbles and real-world shadows share the property of a certain unreality at the crossroads of two deficient realities. In this liminal space, technology remains invisible for the audience; their experience is all about the playful activity with floating bubbles.

Specific to Bubbles, the audience's embodiment within the artwork implied a scenographical solution that separates the active area for user interaction from the visible stage. The vertical display of the stage area and its split from the interactive actor-audience area marks a distinct feature that interactive virtual environments added to the field of performance design. This conceptual separation, typical to many art installations at the time, usually involved the display of live video footage of the horizontal active area on the vertical screen. In such a technical set-up, the screen functions as a 'mirror', in which the audience gazes at themselves, a metaphorical realm exploited in many early media works. Bubbles avoids the semantic connotations of the mirror, as well as the cognitive complications in controlling body movement that often appears in front of a mirror. It plays with aesthetic references to Plato’s allegory of the cave in the theatrical-cinematic setting of a screen that displays shadows of real objects. However, it deliberately evades conceptual overloading by neither encouraging nor requiring the user to embark on a philosophical-theoretical reflection on the qualities of reality and their embodiment trough digital technology. Moreover, neither the user's shadow nor the bubbles’ less than perfect image quality point towards any overarching hidden grand narrative. Apart from augmenting the bubbles' real-world properties to assure that they do not burst easily upon touch, which would render the process of interaction rather short, Bubbles is just like that: playing with bubbles.

The abundance of playfulness and the absence of explicit or candid references to theoretical discourses might have contributed to the success of Bubbles with a global audience. However, it might have also been a reason why it never received formal accolades from the media art community. It was not awarded with any prize at media art festivals, although it made it into four permanent exhibitions (2) and is most likely among the most often exhibited interactive art pieces since the early 2000s. I always suspected that this was partial because Bubbles did not provide the reviewers with a projection screen for the debate of serious avant-gardist theories related to digital technologies and art, but I might be wrong here. However, Bubbles has always been, and still is after all those years, well appreciated by the audience, and in 2006 it was honoured as ‘most popular showpiece’ by visitor’s vote at the Art Lives exhibition in Stuttgart, Germany (Boulevard Baden, 2006). Bubbles is still on display at the ZKM Centre for Art and Media, without interruption since 2001.

The project certainly perfectly achieved the main objectives of its initial concept: The realisation of a user interface that was appealing to a broad and diverse audience, uncomplicated and easy-to-use without prior skills or requirement to learning a manual, and equally low-tech and low in expense. The immense popularity of the work, and its eventual shift from an experiment in interface technology to a work of art that deploys interactive media technology presents an engaging example of the importance of a close alignment of technical and artistic concepts.

Endnotes:
1. Parts of the following description of Bubbles draws on the author's 2000 introduction (Muench, 2000), as well as on Christoph Pingel’s 2001 introduction to Bubbles, commissioned by the author on the occasion of the 2001 exhibition The Interaction 01. Dialogue with Expanded Images at IAMAS (Pingel, 2001). The author's as well as Pingel’s essay was distributed openly and reproduced on many occasions for promoting the artwork, including the author’s website. The thesis’ text reuses expressions and phrases of the original essays.

2. The permanent exhibition venues presented a rather peculiar mix: The ZKM Media Museum in Karlsruhe, the Phaeno Science Centre in Wolfsburg, the Junior Museum in Troy New York, and the LOISIUM, a museum concerned with the history of wine making in Langenlois, Austria. More than 15 years after its first presentation, at least at the ZKM and the Phaeno Science Centre the artwork is still permanently exhibited.

Bibliography
Boulevard Baden. (2006). Karlsruher Museen sind die Nummer Eins im Land, Boulevard Baden, 10. September
Braitenberg, V. (1984). Vehicles: Experiments in Synthetic Psychology. Cambridge, Mass.: MIT Press.
Muench, W. (2000). Bubbles: Introduction. Retrieved 19.11, 2016, from http://hosting.zkm.de/wmuench/bubbles
Penny, S. (2010). Desire for Virtual Space: The Technological Imaginary in 1990s Media Art. In T. Brejzek, W. Greissenegger & L. Wallen (Eds.), Space and Desire: Monitoring Scenography 03 (pp. 168-184). Zurich: ZHdK.
Pingel, C. (2001). Bubbles. In I. Sakane (Ed.), The Interaction 01. Dialogue with Expanded Images. Japan, Ogaki: IAMAS Institute for Advanced Art and Science, CMC Center for Media Culture, World Forum for Media Art and Culture Commitee.

© Wolfgang Muench.