Optimising HRTFs to Improve Spatial Release from Masking

Nils Marggraf-Turley, Niels Pontoppidan, Lorenzo Picinali. Optimising HRTFs to Improve Spatial Release from Masking. In: AVARIG 2026: Audio for Virtual and Augmented Reality and Immersive Games. Audio Engineering Society; 2026. Paper 464. Available from: https://aes.org/publications/elibrary-page/?id=23357

@inproceedings{NilsMarggrafTurley2026_23357,
  author = {{Nils Marggraf-Turley} and {Niels Pontoppidan} and {Lorenzo Picinali}},
  title = {{Optimising HRTFs to Improve Spatial Release from Masking}},
  booktitle = {AVARIG 2026: Audio for Virtual and Augmented Reality and Immersive Games},
  note = {Paper 464},
  year = {2026},
  month = jun,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23357}
}

TY  - CPAPER
TI  - Optimising HRTFs to Improve Spatial Release from Masking
AU  - Nils Marggraf-Turley
AU  - Niels Pontoppidan
AU  - Lorenzo Picinali
T2  - AVARIG 2026: Audio for Virtual and Augmented Reality and Immersive Games
M1  - Paper 464
PY  - 2026
DA  - 2026/06/30
UR  - https://aes.org/publications/elibrary-page/?id=23357
PB  - Audio Engineering Society
LA  - en
AB  - Binaural hearing supports speech understanding by allowing listeners to perceptually segregate spatially separated sources, a phenomenon termed spatial release from masking (SRM). The spatial cues that facilitate SRM are determined by the listeners morphology and as such are described by the head-related transfer function (HRTF). Although individual HRTFs are generally considered important for accurate localisation, previous work suggests they might not necessarily maximise performance across all aspects of spatial perception, including SRM. This motivates the concept of application-specific HRTFs. Here, we propose an HRTF augmentation method to increase model-predicted SRM in frontback cocktail-party configurations where SRM is limited. HRTF log-magnitude spectra are parameterised using principal component analysis (PCA), and PCA coefficients are optimised using a differentiable auditory-model-based objective to increase better-ear SNR while constraining broadband interaural distortions. The proposed method produces model-predicted SRM improvements of 49 dB for frontback targetmasker configurations while maintaining broadband interaural level difference (ILD) deviations below 1 dB.
ER  -