Abstract

This article documents a project that deals with the application of a generative approach for creating audio-visual narration. The project investigates the possibility of producing spatio-temporal montage, offering a kaleidoscopic view of pre-recorded events. Fragmented narratives synthesize a complex whole, which evolve in space and time according to the viewer's behaviour in space. Thus, the viewer becomes the player of a live, constantly changing puzzle. The aim is to create new experiences derived from the synthesis is being pre-recorded memories, according to kinetic commands. A kaleidoscopic narrative is being created according to users' placement, to their movement in space, and to interaction between them. The system's narratives are different if a single observer is interacting with it via his displacement, or if many observers are moving independently towards different directions or as a group. Two kinds of interaction are being developed: humanmachine interaction and humanhuman interaction. Condensings and rarefactions are perceived by the system as different stimuli, and thus they lead to different audio-visual narratives. The environment created by the system is planned to be both procedural and participatory. It is led by rules and by people at the same time. It reacts to the users' acts and is offered for multiple simultaneous personal readings. When interactive environments are offered for simultaneous readings by a number of users, they are altered in a way that depasses the sum of individual interactions, creating an emergent narrative and a collective memory. Video as a means of projecting recorded memories is being edited by the system in order to provoke new experiences and connotations, through semantic montage.