In Utero

Spatial and immersive audio already support real-time control over an experience and can hold up collaborative storytelling on their own, without a screen doing the narrative work. The question this project asked was how spatial audio specifically could open up new kinds of interaction in participatory, mixed-reality performance, not as an effect layered onto a visual scene but as the scene itself.

The design goal that followed was a sonic experience built around envelopment: a listener surrounded by sound from every direction, able to move through the piece from any perspective inside it. That goal is tied to anamnesis, a term for the way a sound or a sonic context can revive a past atmosphere in a listener's consciousness, often involuntarily. The piece isn't just meant to be heard, it's meant to reawaken something already there.

John Cage's 1952 score Williams Mix organized 600 recordings into six categories, city sounds, country sounds, electronic sounds, manually produced sounds, wind-produced sounds, and small sounds requiring amplification, and let their combination, not any melody, carry the piece. Le Corbusier, Iannis Xenakis, and Edgar Varèse's Philips Pavilion at the 1958 Brussels World's Fair went further, fusing architecture, film, light, and music into what Philips itself described wanting as a platform rather than a hall. That bonding of media is often read as an early precedent for virtual reality: an immersive, multi-sensory installation built before the hardware to render it digitally existed.

Ambisonic audio is speaker-agnostic: rather than assigning a recording to a channel, it places a sound source on the surface of a spherical soundfield and encodes that sphere into a b-format signal, where each channel carries one spherical harmonic. First-order ambisonics needs four channels; each order up adds more. Binaural audio works differently and only for headphones, filtering a sound through head-related transfer functions so two ears perceive a source's exact location and elevation, the same illusion your own ears perform on real sound without any processing at all. Combining both, an ambisonic soundfield decoded for speakers and a binaural mix for headphone playback, was the hybrid approach this project used to make the piece hold up across playback systems.

Each sketch below is a timed listening exercise, translating an everyday sound environment into a gesture on paper before it ever became a mix decision. Ten minutes on a park bench, pacing around a house, walking a windy backyard at night, half a phone conversation overheard, each drawn the way it was heard, not the way it looked.