Spectral-Matched Optimization of Head-Related Impulse Response Reverberation Parameters for Binaural Room Acoustics Emulation

Hu C. Spectral-Matched Optimization of Head-Related Impulse Response Reverberation Parameters for Binaural Room Acoustics Emulation. In: Express Paper. Audio Engineering Society; 2025.pp. 359:1-. Available from: https://aes.org/publications/elibrary-page/?id=23033

@inproceedings{Hu2025_23033,
  author = {Hu, Chenghao},
  title = {{Spectral-Matched Optimization of Head-Related Impulse Response Reverberation Parameters for Binaural Room Acoustics Emulation}},
  pages = {359:1-},
  year = {2025},
  month = oct,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23033}
}

TY  - CPAPER
TI  - Spectral-Matched Optimization of Head-Related Impulse Response Reverberation Parameters for Binaural Room Acoustics Emulation
AU  - Hu, Chenghao
SP  - 1
M1  - Paper 359
PY  - 2025
DA  - 2025/10/14
UR  - https://aes.org/publications/elibrary-page/?id=23033
PB  - Audio Engineering Society
LA  - en
AB  - This study investigates parameter optimization strategies for aligning artificially reverberated Head-Related Impulse Responses (HRIRs) with the characteristics of Binaural Room Impulse Responses (BRIRs), using spectral error analysis. A systematic parameter sweep evaluates reverberation time (RT60: 0.52.0 s) and the early-to-late reflection energy ratio (ER Ratio: 0.30.7), using root mean square error (RMSE) between HRIR+reverberator and BRIR-processed spectra as the fidelity metric. The methodology employs the SADIE II databases HRIR (D2_HRIR) and BRIR (D2_BRIR) setsspatially standardized binaural measurements compliant with the SOFA formatcombined with feedback delay network (FDN) reverberator modeling using randomized delay networks. Results demonstrate a linear correlation between increases in the ER Ratio and reductions in RMSE, with RT60 minima localized in two intervals: 0.901.30 s, 1.501.70 s. A global spectral minimum occurs at RT60 = 1.10 s and ER Ratio = 0.70, validated through parametric contour mapping. This work provides actionable insights for balancing perceptual realism and computational efficiency in spatial audio rendering, particularly for virtual environments where direct BRIR measurements are impractical.
ER  -