HRTFs

Head-related transfer functions (HRTFs) are functions that characterize how an individual's ears receive sound from a point in space. They encode the filtering effects of the listener's anatomy, including the head, torso, and especially the outer ear (pinna), on the sound's amplitude and phase across frequency.

HRTFs are usually represented in the frequency domain as HRTFs, with the corresponding time-domain impulse responses called head-related impulse responses (HRIRs). Measurements are typically taken with microphones placed at the entrance of the ear canals of a subject or a dummy head in an anechoic chamber, for many directions around the listener.

The HRTF varies with azimuth and elevation, and also with distance. They are highly individualized, reflecting anatomy, though generic or averaged sets are used in many applications.

In audio rendering, HRTFs are used to synthesize binaural signals for headphones, enabling localisation cues such as interaural time difference (ITD), interaural level difference (ILD), and spectral cues produced by the pinna. They are essential in virtual reality, gaming, and 3D audio.

Applications often use parametric approximations or decompositions (e.g., spherical harmonics) to reduce data size, and may use non-individual HRTFs or personalized measurements. Limitations include variability between individuals, changes with head position, and distance effects; real-time rendering must account for head tracking for accurate perception.