AVIS: Trans-Acoustic Mapping is an immersive audio-visual performance that reconfigures the architectural space of the MAM through a decentralized computational system. Grounded in the Co-Perception Model, the project treats sound not as a static acoustic object, but as a mutable substance and an autonomous spatial agent. Within this Shared Perceptual Field, the traditional boundaries between creator, system, and observer are dissolved. The performance executes an algorithmic pipeline designed in Max/MSP for real-time granular synthesis and spectral processing, operating entirely under the 'Live Reaction' paradigm without a pre-recorded click track. Low-level audio features and descriptors (such as RMS amplitude, spectral centroid, and harmonic vectors) are continuously extracted and mapped into cross-modal latent representations. Through network protocols (OSC/Spout), these data matrices are translated into real-time geometric modulations and generative visual topographies. Listening is thus transformed into a situated act and a multi-sensory navigation. The environment and the audience enter a state of Bidirectional Modulation, wherein the spatial distribution of light is directly sculpted by the emergence of sonic behaviors, manifesting an ecosystem of distributed agency.
