Binaural Capture and Synthesis of Ambient Soundscapes

Phase I work is proposed to design, to begin implementation of, and to evaluate a high-quality, portable soundscape capture and synthesis system. The aim of such a system is to record all relevant acoustic information in a particular environment so that it can be re-synthesized for subsequent presentation to a human listener. The current effort will yield a system where the synthesized soundscape preserves both the source levels and perceived source locations of the original soundscape. Source levels will be captured and synthesized accurately over a dynamic range that spans from the threshold of hearing to 110 dB SPL and over a frequency range spanning 100-12000 Hz.. Realistic source localization cues will be provided in the form of (1) binaural (ITD and ILD) cues resulting from microphone mounting on a cylindrical model head and (2) spectral cues resulting from directional microphone-cluster processing. Additionally, the synthesized soundscape will permit listener head-movement in order to increase the realism and sense of immersion. The resulting system will be evaluated to verify level accuracy over the desired dynamic range and the equivalence of source localization and masking between the original and synthesized soundscapes.