[Paper Review] HRTF Individualization: A Survey

Corentin Guézénoc, Renaud Séguier|arXiv (Cornell University)|Mar 13, 2020

Multisensory perception and integration17 citations

TL;DR

This survey reviews four approaches to HRTF individualization—acoustic measurement, numerical simulation, morphology-based indirect methods, and perceptual feedback-based methods—evaluating their technical trade-offs and, critically, the perceptual validity of their results. It finds that while acoustic measurement yields the most perceptually accurate HRTFs, it is impractical for end-users; perceptual and morphology-based methods show promise but lack sufficient large-scale validation, highlighting a critical gap in robust, user-friendly individualization solutions.

ABSTRACT

The individuality of head-related transfer functions (HRTFs) is a key issue for binaural synthesis. While, over the years, a lot of work has been accomplished to propose end-user-friendly solutions to HRTF personalization, it remains a challenge. In this article we establish a state-of-the-art of that work. We classify the various proposed methods, review their respective advantages and disadvantages, and, above all, methodically check if and how the perceptual validity of the resulting HRTFs was assessed.

Motivation & Objective

• To provide a comprehensive state-of-the-art review of HRTF individualization techniques for end-users.
• To classify and compare the advantages and disadvantages of four main HRTF individualization approaches: acoustic measurement, numerical simulation, morphology-based, and perceptual feedback-based methods.
• To systematically evaluate the perceptual validity of each method, focusing on whether and how perceptual studies were conducted.
• To identify research gaps, particularly the lack of large-scale, perceptually validated studies in morphology- and statistics-based methods.
• To guide future research toward more practical, perceptually effective, and user-friendly HRTF individualization solutions.

Proposed method

• The paper conducts a systematic literature review of HRTF individualization methods published up to 2023, focusing on four categories: acoustic measurement, numerical simulation, indirect individualization using morphological data, and indirect individualization using perceptual feedback.
• For each method, the authors analyze technical implementation, equipment requirements, measurement time, and user comfort, drawing from studies published between 1992 and 2017.
• The study emphasizes perceptual validation by reviewing whether each method was tested in localization or preference experiments, and evaluates the number of participants and experimental design quality.
• A key analytical tool is Table VI, which synthesizes the perceptual results of 34 studies, assessing the number of participants, directionality tested, and perceptual outcomes.
• The authors use a structured classification to compare methods based on accuracy, practicality, and reproducibility, with special attention to the reliability of perceptual data.
• The methodology includes critical assessment of limitations such as subject movement during measurement, geometric inaccuracies in simulations, and the lack of large databases in statistical modeling.

Experimental results

Research questions

• RQ1 What are the main technical and perceptual trade-offs among acoustic measurement, numerical simulation, morphology-based, and perceptual feedback-based HRTF individualization methods?
• RQ2 To what extent have perceptual validation studies been conducted for each HRTF individualization method, and what are their methodological strengths and weaknesses?
• RQ3 Why do some indirect methods—especially those based on anthropometric measurements or statistical models—show mixed or inconclusive perceptual results despite theoretical promise?
• RQ4 How does the size and quality of HRTF databases affect the performance of statistical modeling approaches in HRTF individualization?
• RQ5 What are the key barriers to developing a practical, end-user-friendly, and perceptually accurate HRTF individualization solution?

Key findings

• Acoustic measurement remains the gold standard for perceptual accuracy, with studies showing individualized HRTFs achieve localization performance equivalent to real free-field sources.
• Despite its accuracy, acoustic measurement is impractical for end-users due to high equipment cost, long measurement times (up to 1h45), and the need for anechoic environments.
• Recent methods using continuous rotation and adaptive filtering have reduced measurement time to as little as 4 minutes, significantly improving usability.
• Subject movement during measurement introduces directional imprecision, but visual feedback systems have reduced this error to imperceptible levels in recent studies.
• Numerical simulation offers high repeatability and user comfort but suffers from inaccuracies due to geometric acquisition errors and acoustic modeling limitations.
• Morphology-based methods using smartphone photos show promise for low-cost individualization, but perceptual validation is limited, especially for complex regression models, and databases remain small (43–50 subjects).