"Figure-ground" deals with the mental and cognitive processes that oscillate between reflection and dissociation. As part of the performance, an audiovisual collage is created in real time. Both the visual and sonic layers aim to create a kind of pareidolia effect. Using a real-time cany-edge detector, the outlines of a performer's fragmented self-portrait feed a Gray-Scott model of reaction diffusion. Its values control the parameters of a sonic feedback system.