Digital signal processing (3rd ed.): principles, algorithms, and applications
Digital signal processing (3rd ed.): principles, algorithms, and applications
Hi-index | 0.00 |
In order to achieve highly accurate 3D audio spatialization, an intended listener must undergo measurements in a specialized instrumentation system to obtain "individual" Head-Related Impulse Responses (HRIRs). Our goal is to create a customizable sound spatialization system that would not require complex individual measurements. Our approach is based on the reliable decomposition of measured HRIRs into damped sinusoidals. A previously developed method showed good performance in decomposing HRIRs collected by us at a sampling frequency of 96 kHz, but other HRIR databases comprise HRIRs recorded at lower rates. This paper compares the performance of our automated decomposition method with sampling rates of 96 kHz and 48 kHz.