Pressure Matching Using Data-Driven Estimation for Sound Fields and Transfer Functions

Sound field reproduction is a technology that uses multiple loudspeakers to reproduce a realistic sound field. The pressure matching, well-known sound field reproduction method, is not easily implemented because it requires a large number of measurements to obtain the transfer functions of loudspeakers and the desired sound pressures at the control points. In this study, we propose a method to derive the driving functions of loudspeakers based on the pressure matching using data-driven estimation of sound field. The proposed method estimates the desired sound pressures and the transfer functions from loudspeakers at the control points from the limited number of microphones. From simulation experiments, the proposed method significantly improves the reproduction accuracy and reduces the number of microphones by approximately 70 percent.