The outcome research initiative into interactive avatars resulted in Vocal Voids, which has been presented as talk, interactive digital video demonstration and published paper at Subtle Technologies, the high profile international conference exploring leading edge ideas in art and science. This invitation to present reflects the significance the international community attach to it. It is also published in the book Responsive Architectures. Vocal Voids pushes the boundaries of sonic knowledge within the fields of interactive acoustics, music composition and immersive sound/imagery. It combines the pioneering work of Thomas Funkhouser’s Interactive Acoustic Modelling (www.cs.princeton.edu/-funk) and Bose 3D architectural software with interactive sonic visual mapping as created for the games industry by Luigi Agostini of See‘n’Sound (www.seensound.aegweb.com). Whilst Agostini’s system extracts in real-time 3D movement from 2D imagery through a MIDI algorithm and Funkhouser’s work enables a walk-through acoustic simulation of a virtual building in real-time, neither system assesses the potential for combining the two theorems into a new hybrid form where the sonic possibilities in adaptive acoustics could be simultaneously explored within a ‘musical’ context. By making this connection it opens up a new world of artistic expression especially in music where the MIDI language acts as a translation medium between the visual and musical construct. Importantly, MIDI code is the key in my research process to integrate a third vital element into the equation - computational generative music or music that adapts to external stimuli. Vocal Voids uniquely uses this trinity for a ‘composer’ to conduct a fly-through singing avatar in the exploration of spatial environs that can be acoustically manipulated on-the-fly as an aesthetic element in the music composition. This is a significant breakthrough in real-time acoustic performance within the virtual world, and whilst well received in the academic forum, its application in the wider context is on-going.