Binaural audio uses Head-Related Transfer Functions (HRTFs) to simulate 3D spatial audio over headphones

Human binaural hearing relies on interaural time and level differences (ITD/ILD) and the spectral coloration introduced by the pinnae and head (the Head-Related Transfer Function, HRTF). The HRTF varies with elevation and azimuth. Binaural audio synthesis convolves a mono signal with the HRTF for the desired virtual position, then plays the result over headphones, creating the illusion of a sound in 3D space. A key limitation is that HRTFs are individual-specific: a generic HRTF derived from an average or another person’s measurements may degrade externalization (the cue that a sound is outside the head). Personalised HRTFs can be measured with a microphone in the listener’s ear canal.

Examples

Apple AirPods Spatial Audio uses head tracking with HRTFs to simulate 5.1 or Atmos surround over headphones. Video game audio engines use real-time HRTF convolution for 3D sound positioning.

Assessment

Why might a generic HRTF produce worse spatial audio than a personalised one? What are the two primary physical cues (ITD and ILD) that the HRTF encodes?