Head-Related Transfer Functions (HRTFs) govern how sound waves are filtered by the head, pinna, and torso before reaching the eardrum. These acoustic cues form the foundation of spatial hearing, allowing listeners to determine the elevation, azimuth, and distance of sound sources. As the human body ages, both the physical structures responsible for sound filtering and the neural pathways that decode these cues undergo substantial and often progressive alterations. Understanding these changes is critical for designing effective spatial audio systems that remain usable across the lifespan.

Structural Alterations of the Outer Ear

The pinna, or outer ear, performs essential high-frequency spectral filtering by introducing characteristic notches and peaks that provide vertical localization cues. With advancing age, the cartilage of the pinna loses elasticity, and the overall shape tends to become elongated or flattened. Specific changes include decreased concha depth, altered helix curvature, and a reduction in the prominence of the tragus and anti-tragus. These morphological shifts directly modify the spectral patterns of the HRTF. Research demonstrates that even minor changes in pinna shape can shift the perceived elevation of a sound source by several degrees—enough to cause front–back or up–down confusion in virtual audio environments. Additionally, the ear canal undergoes involution: the diameter often decreases and the length shortens, which alters the quarter-wave resonance that normally amplifies frequencies in the 2–4 kHz region. This resonance change can reduce the natural amplification of important speech frequencies, further complicating spatial perception.

Presbycusis and Auditory Sensitivity

Age-related hearing loss, known as presbycusis, predominantly impairs high-frequency hearing. Because HRTF localization cues, especially vertical and distance cues, rely heavily on spectral information above 4 kHz, the loss of sensitivity in this range degrades the ability to resolve fine spatial details. Older adults consistently demonstrate elevated detection thresholds for interaural level differences (ILDs) at high frequencies, making it harder to distinguish sources at different azimuths, particularly for sounds with broad frequency content. Even when hearing aids compensate for the loss, their compression and amplification algorithms can distort the natural HRTF cues—for example, by reducing the dynamic range of spectral notches. The result is a systematic degradation of spatial awareness that is not fully restored by simply amplifying sound.

Neural Processing Degradation

The central auditory system also undergoes age-related degeneration. Key nuclei—the cochlear nucleus, superior olivary complex, and inferior colliculus—exhibit neuronal loss and reduced synaptic density. This decline slows neural conduction velocity, impairing the processing of sub-millisecond interaural time differences (ITDs) that are critical for low-frequency localization. Furthermore, the aging brain shows diminished cortical representation of spatial cues in areas such as the auditory cortex and the planum temporale. As a result, older listeners often require longer sound exposures to reliably identify a source position and become more susceptible to disruption of the precedence effect, a mechanism that normally suppresses echoes to preserve directional clarity. The combination of peripheral and central changes leads to a measurable decline in both ITD and ILD discrimination accuracy.

Impact on Binaural and Spectral Cues

Both ITDs and ILDs are affected by age. For steady-state low-frequency sounds, older adults exhibit increased jitter in ITD discrimination thresholds, often exceeding the 20–30 microsecond difference that younger listeners can detect. For broadband transient sounds, ILD thresholds may increase by 2–3 dB. Spectral notch cues become less reliable as listeners lose high-frequency sensitivity and as pinna shape changes. Consequently, the overall “headphone transfer function” that each individual experiences is altered, leading to systematic up–down and front–back confusions. These deficits are especially pronounced in reverberant environments where temporal and spectral cues are already degraded.

Implications for Spatial Audio Applications

Virtual Reality and Gaming

Immersive virtual reality (VR) depends on accurate HRTF-based binaural rendering to create a convincing sense of presence. For older users, a generic HRTF profile—such as the widely used KEMAR mannequin—can produce severe localization errors. A sound intended to originate from above may be perceived as behind, or a voice coming from the left front might be indistinguishable from a sound to the right rear. In complex acoustic scenes like multiplayer games with multiple footsteps, gunshots, and environmental sounds, older players report greater mental effort and reduced spatial awareness. This can decrease enjoyment, increase simulator sickness, or even exclude older individuals from certain VR experiences. Sound designers should consider offering age-specific HRTF profiles or easy calibration tools to mitigate these issues.

Communication in Multi-Talker Environments

Age-related HRTF perception degradation directly impacts speech understanding in noisy, crowded spaces—the so-called cocktail party problem. Binaural unmasking, which normally allows listeners to filter a target speech stream from competing talkers based on spatial cues, becomes less effective with age. Older adults require a higher signal-to-noise ratio to achieve the same identification accuracy as younger listeners. When teleconferencing systems or hearing aids apply spatial audio processing with mismatched HRTFs, the problem worsens. For example, a hearing aid that uses a generic head-related impulse response (HRIR) may create conflicting directional cues that interfere with the user’s residual natural spatial hearing. Designers of assistive listening devices must account for these age-related changes to improve real-world communication outcomes.

Hearing Aid and Cochlear Implant Performance

Modern hearing aids and cochlear implant processors increasingly incorporate directional microphones and binaural beamforming algorithms that mimic HRTF filtering. However, if the device’s HRTF model does not account for the user’s individual ear geometry and hearing loss profile, the artificial spatial cues may conflict with residual natural cues. This can cause inconsistent localization, especially in reverberant environments. For bilateral cochlear implant users, the lack of fine-structure temporal information further complicates ITD processing, and age-related neural degeneration multiplies the challenge. Emerging research suggests that personalized HRTF filters, updated as the user ages or changes hearing aid settings, can significantly improve localization accuracy and reduce listening effort.

Automotive and Aviation Audio

In vehicle infotainment systems, spatial audio can provide navigation prompts that seem to originate from the direction of the turn. For older drivers, inaccurate HRTF rendering could cause confusion or delayed reaction times. Similarly, in aviation, pilot audio alerts must be precisely localized to avoid misinterpretation during critical maneuvers. Age-related changes in the HRTF must therefore be considered in the design of safety-critical auditory interfaces. Some manufacturers are beginning to include age-adjustable spatial audio settings in their systems, allowing users to calibrate the perceived direction of alerts based on their individual hearing profile.

Strategies for HRTF Customization and Compensation

Individualized HRTF Measurement

The most accurate approach to personalization is direct measurement of a listener’s HRTFs using in-ear microphones or a binaural recording headset while playing test signals from a spherical speaker array. This process yields a personal HRTF set covering many directions, capturing the unique spectral patterns produced by the individual’s anatomy. However, traditional measurement is time-consuming, requires specialized equipment, and may not be feasible for large-scale deployment. Newer solutions use low-cost microphone arrays embedded in earphones and prerecorded pseudorandom noise sequences that can be performed at home. For older listeners, the measurement should be repeated if hearing aid fittings change or after significant weight loss or gain that alters neck or torso geometry.

Model-Based Customization Using Anthropometric Parameters

When direct measurement is impractical, HRTFs can be estimated from easily measurable anatomical features. Researchers have developed databases linking 40–60 anthropometric measurements—pinna height, width, concha depth, head width, torso height—to corresponding HRTF filter coefficients. By applying a regression or neural network model, a personalized HRTF can be generated without acoustic recording. For older adults, the model must account for age-related changes in ear elasticity and cartilage stiffness, factors not always captured in standard anthropometric sets. Recent work uses 3D morphable models of the ear derived from photographs, enabling greater precision. These models can also incorporate age as a parameter, adjusting predicted HRTFs based on typical anatomical shifts observed in older populations.

Adaptive and Perceptual Calibration

Another approach adjusts HRTFs through interactive user feedback. In a listening test, the user indicates the perceived direction of a test sound, and an algorithm—such as a least-mean-squares optimizer—tweaks HRTF parameters to minimize localization error. This compensates for both physiological changes and individual preference. Adaptive calibration works well for real-time applications like gaming or VR, where the user can quickly run a short setup procedure. For older listeners, the calibration should use sounds at frequencies they can still hear comfortably, and it should be repeated periodically as hearing and anatomical features continue to change. Some commercial VR systems already include such calibration routines, and they are increasingly being adapted for hearing-impaired users.

Enhancing Spectral Cues Through Equalization

To counteract high-frequency hearing loss, spectral notches and peaks in the HRTF can be artificially boosted in the 4–8 kHz range—or within the listener’s residual hearing region. This equalization must be applied carefully to avoid creating unnatural timbre. A “spatial enhancement” filter that selectively amplifies frequencies carrying important localization cues, while preserving overall level, can improve localization accuracy without causing listener fatigue. Some commercial spatial audio plugins now include optional “senior” modes that apply such EQ based on the user’s audiogram. For example, a plugin might apply a gentle boost around 6 kHz to restore the spectral notches that guide vertical localization. The key is to balance enhancement with naturalness, as over boosting can lead to harshness and reduce listening comfort.

Multimodal Integration

When auditory spatial cues alone are insufficient, other sensory modalities can supplement them. Visual augmentation—such as a subtle glowing marker on a VR controller that moves with sound source—can guide the listener’s attention. Haptic feedback, such as a vibration on the side of the head corresponding to the sound direction, can reinforce localization. For hearing aid users, combining directional microphones with visual cues from a head-mounted display has shown promise for improving speech perception in noise. While multimodal integration does not directly address HRTF inaccuracies, it provides a compensatory mechanism that can make spatial audio more accessible to older users. Developers should design these cues to be intuitive and non-intrusive, ensuring they enhance rather than distract from the experience.

Current Research and Technological Development

Databases of Aging Ears

Historically, most HRTF databases were limited to young adults (20–35 years). Newer efforts, such as the “AHRTF” project, have collected HRTF measurements and anthropometric data from individuals aged 55–85. These datasets reveal systematic variations: older ears often have a shallower concha and a smaller cymba, leading to a reduced spectral notch depth. Machine learning models trained on these datasets can now predict age-appropriate HRTFs with smaller gain errors compared to generic models. The increasing availability of such databases enables more inclusive design of spatial audio systems. Researchers are also collecting longitudinal data to track how individual HRTFs change over time, which will inform adaptive algorithms.

Machine Learning for Real-Time Adaptation

Researchers are developing convolutional neural networks (CNNs) that can infer an optimized HRTF from a single photograph of the user’s ear and a brief hearing test. These models run on mobile hardware and can update the filter in less than a second. For older users, the model can incorporate hearing thresholds as input features, automatically boosting or cutting specific frequency bands in the spatial filter. This represents a major step toward truly universal personalized spatial audio. Future systems may combine such inference with continuous monitoring of ear canal acoustics via in-ear microphones, allowing the HRTF to adapt in real time as the user’s anatomy or hearing changes.

Integration with Hearing Aids

Hearing aid manufacturers are beginning to embed HRTF filters into wireless streaming of audio from TVs, phones, and VR headsets. The Oticon More™ and Starkey Evolv AI™, for example, use binaural processing that attempts to preserve spatial cues. However, they still rely on generic head-related impulse responses. Future developments will allow hearing aids to load a custom HRTF set measured or modeled for the individual user, and to dynamically switch between modes for different soundscapes (quiet, noisy, reverberant). This will require close collaboration between hearing aid manufacturers, audio software developers, and researchers to ensure compatibility and ease of use. Open standards like the SOFA (Spatially Oriented Format for Acoustics) convention facilitate such integration.

Practical Recommendations for Content Creators and Developers

  • Provide customization options: Allow users to calibrate spatial audio with a simple guided procedure (e.g., locate a test sound from several directions). Store the calibration profile in the cloud so it transfers between devices.
  • Use age-appropriate generic HRTFs: If personalization is not possible, offer a selection of HRTF templates based on age group, head size, and ear shape. Avoid a one-size-fits-all approach.
  • Enable hearing loss compensation: Include a “hearing-compensated” mode that applies frequency shaping to the spatial audio based on the user’s audiogram (if available). This is critical for users over 60.
  • Test with older participants: When developing VR or audio game experiences, recruit a sample of older users during user testing. Their performance in localization tasks will reveal mismatches in the HRTF modeling.
  • Educate users: Explain that spatial audio perception may require periodic recalibration as they age, and provide clear instructions for resetting the profile.
  • Design for hearing aid compatibility: Ensure that streams support HRTF-enabled processing and that latency remains low enough to avoid desynchronization with visual cues.

Future Directions

The next frontier is continuous adaptation. Wearable sensors—such as in-ear headphones with integrated microphones—could monitor the user’s ear canal acoustics over time and update the HRTF filter automatically. For example, if the user gains weight or the ear canal lining thickens, spectral cues shift; an algorithm could detect the mismatch by comparing predicted and actual otoacoustic emissions or by analyzing the feedback from a short localization test. Similarly, hearing aids that measure real-time otoacoustic emissions could adjust spatial cues on the fly, maintaining optimal localization accuracy despite anatomical changes.

Another promising avenue is auditory perceptual training. Older listeners can improve localization accuracy through gamified training exercises where they identify sound directions and receive corrective feedback. While this does not change the underlying HRTF, it can strengthen the neural decoding of spatial cues, partially compensating for sensory degradation. Combining training sessions with personalized HRTFs may yield synergistic benefits that improve real-world spatial hearing.

Finally, open-source platforms such as the SOFA (Spatially Oriented Format for Acoustics) and the LISTEN HRTF database are expanding to include diverse demographics, including older adults. Developers should adopt these standards to ensure cross-platform compatibility and to facilitate research. The ultimate goal is to make high-quality spatial audio accessible and effective for every listener, regardless of age or hearing ability. For further reading on age-related changes in binaural hearing, see Eddins and Hall (2016) in JASA. An overview of HRTF customization techniques can be found in Gupta et al., IEEE Access 2020. For practical guidance on hearing aid integration, consult ASHA’s guidelines on hearing aid selection.