RADIOSANDS, Fahrbereitschaft (Haubrok Foundation), Berlin 2019, Image © Klaus Fritze


We doubtlessly encounter a growing variety of media information on a daily base and with increasing speed and quantity. The rapid selection of small chunks of media deriving from inconsistent sources has become a valid consumption habit. How does this affect our perception? Do the fragmented bits and parts of mediated speech, music and inexplicable noises that we expose ourselves to form their own language – a unitary message made from incoherent noises? A transcendent imperative that we unconsciously follow?

Radiosands looks into these questions by analyzing and synthesizing simultaneous media flows within a walkthrough installation space. The work stands in the tradition of the sound collage. Nevertheless it utilizes the speed and algorithmic potency of the digital realm to generate a genuinely new experience: A real-time collage of mediastreams that is distributed throughout multiple spots within an expansive sculpture. The choreography of the audio cuts and their spatial alignment is perceived as an acoustic continuum –a new entity– due to its psychoacoustic nature.

Listening, RADIOSANDS, Fahrbereitschaft (Haubrok Foundation), Berlin 2019, Image © Jie Liang Lin
Walkthrough, RADIOSANDS, Fahrbereitschaft (Haubrok Foundation), Berlin 2019, Image © Jie Liang Lin


Radiosands explores the emotional and aesthetic impacts created by an incisive manipulation of mediastreams. The installation is based on an old acquaintance, the radio, that stands as archetype for various kinds of information streams.

Sixteen custom made radio receivers are placed in a room. Via a specifically designed software each radio can remotely be turned on and off and its audio dynamics can be controlled. Thus a choreo- graphy of sonic cutouts permeates the space, applying chunks and bits of audio that are drawn in real-time from the various public radio streams.

Particles of spoken language build up sonic cascades that spread throughout the space. Snippets of pop and classical music become rhythmically arranged amongst the diverse radio sets. Granular pieces of broadcasts cross the exhibition site like an acoustic swarm… Visitors can move freely through the room experiencing the sonic architecture that is stretched by the multiple radio objects.

Composition / Radiosands Software

The electronic circuits of the radio sets are modified such that they can be remotely controlled through a composition software program. They can individually be switched on and off, their volume can be faded in and out, and further manipulations as e.g. frequency filtering can be applied. Thus the current radio streams become persistingly processed to form a self-renewing sonic collage.

The composition stands in the tradition of the sonic collage as introduce with the Musique Concrète of the late forties and followed by the development of a vast variety of digital audio sampling techniques. In contrary to the known practices, Radiosands utilizes life audio-streams as source material and allocates the sound particles towards various positions within the physical room. Thus creating the experience of being submerged within the sonic field.
Radiosands induces sonic continua – the impression of listening to a coherent sound– by increasing the distribution speed of the audio chunks throughout the room. The choreography of the sound particles is controlled by an array of random parameters that are correlated with a semantic network.

Semantic Network

On a deeper layer of the composition software a speech analysis program scans each individual audio stream for terms and expressions that refer to the semantic field of reality. A predefined keyword list will include terms from the realm of social and political reality building as disinformation, alternative reality, media manipulation and the like, but also comprises the denominations of human perception as touch, hear, feel, sense, etc. The list represents a poetic preselection of search terms that will continuously be augmented.
A Machine Learning Algorithm analyzes the semantic context to identify structures within the constant information feed. The machine creates meaning from the data according to its own standards of “world building“ and feeds its substantial findings immediately into the compositional structure of the software. Thus the data interpretation directly affects the sonic performance.

The speech analysis can be operated multi-lingual (English, German, French, Spanish, Portuguese).

Drawing, Landscape RADIOSANDS, Berlin 2018, Image ©Thom Kubli
Drawing, Immersion, Loops and Holes, RADIOSANDS, Berlin 2018, Image ©Thom Kubli

Who do you trust?

The radio has become an acquainted object that stands synonymous for information, communication, civilization. It is inscribed with the ability to wirelessly receive meaning at any random point in a room and to play it back to the listener. The various forms and shapes of radio-sets physically filter the sound according to their size and material. And the radio has undergone an technosocial aging process, though it never seized to exist aside from the spotlight of current high-tech innovations. This adds to an subtle anthropomorphic quality, making it kind of a reliable companion -a trustworthy item. Thus it seems perfectly suited as object for a tweak, an unexpected expansion of its cerebral capacities.

Radio as Media Archetype

The radio represents an archetype of broadcast media streams: It was declared dead as many times as it was re-invented. It served as source of entertainment and information, as integral voice for modern audio compositions and as object within contemporary art installations. But foremost it was an instrument of political agenda since it was introduced to the public more then a century ago. So called broadcast psychological operations became relevant in the early 1930s and were still applied in the struggles in Afghanistan and Iraq in the late 2000s. They meant to air drop radios and radio leaflets into eras of conflict and combat. Through this propaganda and educational programs were conveyed to audiences in the cities and in hard to reach mountainous areas in order to influence emotions and objective reasoning through radio broadcasts.

Composition Software, RADIOSANDS, Fahrbereitschaft (Haubrok Foundation), Berlin 2019, Image © Jie Liang Lin
Receiverfarm, RADIOSANDS, Fahrbereitschaft (Haubrok Foundation), Berlin 2019, Image © Jie Liang Lin

The project is a collaboration between Thom Kubli and scientists from the Zurich University of Applied Sciences ZHAW, Institute for Applied Simulation.
Artistic idea, project management, composition: Thom Kubli | Scientific Director ZHAW: Prof. Dr. Sven Hirsch
Computational Linguistics: Dr. Manuel Gil, Dr. Martin Schüle | Audio analysis: Prof. Dr. med. Thilo Stadelmann, Daniel Wassmer, Tobias Schlatter | Programming: Lydia Ickler | Technology and Mechatronics: Florian Guist, Marek Olkusz, David Jaschik, Michael Barocca | Assistant: Christian Maximilian Blasius

Special thanks to Speechmatics for providing the Speech Recognition Engine