The validation of a sound field simulator for measurement of the real world performance of directional microphones for hearing aids

This investigation compared the effectiveness of four test environments in assessing real-world directional microphone performance: Two traditional sound booth environments, a real restaurant, and a newly developed sound-field simulation system consisting of an octagonal array of eight loudspeakers.;Tested with KEMAR recordings of three hearing aid microphones having radically different directional characteristics, normal hearing subjects required nearly identical signal-to-noise ratios (SNRs) for 50% correct performance in either the real or the simulated environment and scores improved as directivity increased. This did not occur in the traditional settings. The rank ordering of hearing aid microphone performance in the test booth with a single noise source overhead (90°) was similar to the diffuse environments, with the exception of significantly better performance for the array microphones (which had pickup nulls overhead). Compared to the other three conditions, performance in the test booth with a single source behind (180°) was slightly worse for the omnidirectional microphones, similar for the super-cardioid microphones (D-MIC), and significantly worse for the array microphones. Performance was attributed to the fact that the omnidirectional microphones had slightly better pickup from behind than from overhead and the array microphones had a large pickup lobe located directly behind.;It was concluded that, if one's goal is to assess hearing aid microphone performance under realistic conditions where speech is directly in front (0°) of the listener and ambient noise is more or less diffuse and surrounds the listener (e.g. in a restaurant or other social situation), then the simulator technique comes closer to imitating such a situation than do the other conditions---and does so using real, unaltered restaurant noise. Further, the efficacy of assessing hearing aid microphone performance by placing the target speech signal directly in front of the listener and using a single competing noise source either behind or directly overhead is unrealistic and can provide a contrived performance advantage or deficit, depending on the polar pattern of the microphone. The major strength of the simulator system is that a realistic estimate of the in-noise performance of any hearing aid microphone can be assessed without having detailed information about its polar pattern.