Out-of-head sound localization using adaptive inverse filter

It is well known that the transfer functions for out-of-head sound localization differ with the listener. Our goal is to find a method that easily realizes excellent sound localization without measuring the individual transfer functions. This paper proposes an out-of-head sound localization system that uses adaptive inverse filtering without any measuring preprocesses. This system estimates the inverse Ear Canal Transfer Function (ECTF), and can adaptively obtain the transfer function to fit the listener in real-time. This paper also proposes an adaptive inverse filtering method for outof-head sound localization that differs from the filtered-x method. In addition, the relationship between convergence time and initial value is assessed. It is clarified that the proposed method is more effective, in terms of convergence, if the initial value is the average of many listeners’ impulse responses.