Fictional Soundscapes
Imaginary cities through sound and machine learning
An audiovisual installation blending sounds from cities worldwide to create fabricated environmental soundscapes of places that do not exist.
Fictional Soundscapes is a collaborative project with sound artist and urban composer Kenta Tanaka that creates fabricated environmental sounds by blending audio from cities around the world.
Concept
“Cities are pictured in people’s minds, they are an imaginary, conceptual phenomenon.”
The project explores the relationship between human beings and environmental sounds in the city through phenomenological listening. By applying urban theory to sound art production, the work establishes an “Urban Composition” that weaves sounds around the theme of the city.
The Installation
Fictional Soundscapes blends sounds from places around the world to create environmental sounds from cities that do not actually exist. The sounds of crowds, trams, the whispering of the sea and forests create a blended, fabricated soundscape that feels familiar yet belongs to no real place.
Audio Generation
The sounds are generated by a system which automatically mixes soundscapes of different cities from all over the world. Field recordings from Tokyo, Brussels, and other urban centers are algorithmically combined to produce continuously evolving sonic environments.
Visual Generation
As an element for seeing, the computer generates fictional landscapes by means of machine learning techniques. The visuals complement the audio, creating imaginary cityscapes that match the fabricated soundscapes.
About Kenta Tanaka
Kenta Tanaka (b. 1993, Tokyo) is a guitarist, sound artist, and urban composer based between Brussels and Tokyo. His work unravels the relationship between human beings and environmental sounds in the city through phenomenological listening.
Notable works include:
- Urban Reminiscence - Sound, Object, and Rhythm (Sta., Tokyo, 2020)
- Urban Rhythmability (TOKYO culture research, Roppongi, 2020)
- Algorithmic Urban Composition (Stanford University, 2019)
Technical Role
As Technical Director, I developed the real-time audio mixing system and machine learning pipeline for generating the visual landscapes, enabling the seamless blending of field recordings from multiple cities into cohesive fictional environments.
Gallery
More Images
