Logo of the Physikalisch-Technische Bundesanstalt
Into the Future with Metrology - The Challenges of Medical Technology

Evaluation of response times of test subjects in the automated determination of reference hearing thresholds


Audiometers are calibrated based on reference hearing thresholds. These thresholds are determined by standardized procedures and based on a representative, normal hearing group of test subjects. Often, also at PTB, these data are measured by means of automated routines, which evaluate response times of the test subjects to a stimulus. The evaluation of the response times by the applied algorithms were investigated, focusing on their influence on measured hearing thresholds and the measurement duration.

An important parameter for the assessment of hearing is the pure-tone audiogram. The audiometers and procedures used for this purpose are described in ISO 8253-1 (2010). For each type of audiometric headphone, dedicated reference hearing thresholds are required. These must be collected for 25 young adults with healthy hearing under defined conditions and with defined procedures according to the ISO 389-9 (2009) standard. Often, automated procedures are applied in which the measurement of response times is used to evaluate whether a pure-tone stimulus was heard. Typical response times are the time interval RTON (ON response time) between the stimulus onset and pressing a response button and the time interval RTOFF (OFF response time) between the stimulus end and the release of the response button.  If the response times are within defined limits RTMAX, the stimulus is considered "heard". At PTB, the determination of the reference hearing thresholds is carried out automatically using the bracketing method (Richter 1979). Here, a stimulus (e.g. a pure tone with a frequency of 1000 Hz) is presented several times, whereby the response of the test subject leads to a change in the sound pressure level with each presentation. The sound pressure level will oscillate around the subject's threshold of hearing according to the bracketing method, so that it is sometimes audible and sometimes inaudible. The typical time profile of a single stimulus presentation and of a subject's response are outlined in Figure 1.


Figure 1. Time profile of the stimulus followed by a pause (orange curve at the top) and time profile of a test subject's response rated as "heard" in the form of pressing a button (blue curve at the bottom).


A recent study investigated the influence of different response-time limits on measured hearing thresholds and measurement duration (Bug et al. 2023). The hearing thresholds of 25 subjects were determined with two different response time limits (RTMAX with 600 ms and 1000 ms) for the audiometric frequencies 125, 250, 500, 1000, 2000, 4000 and 8000 Hz. The measurement of the two RTMAX took place in random order. It was found that a response-time limit of 1000 ms results in hearing thresholds that are, on average, 2.6 dB lower in sound pressure level than when a limit RTMAX of 600 ms is applied (Figure 2). This shift is similar for all audiometric frequencies and does not change for even longer RTMAX of 1200 ms. The latter was found in a preliminary study. The measurement duration per hearing threshold determination, on the other hand, showed no dependence on RTMAX and was 104 s ± 18 s.


Figure 2: Difference in equivalent threshold sound pressure levels (ETSPLs) for each subject between the 1000 ms RTMAX measurement and the 600 ms measurement at different audiometric frequencies. The plot is a boxplot with the shaded box indicating 50 % of all subjects' readings, the line inside the box indicating the median value and the whiskers indicating the lowest and highest values. Data points statistically evaluated as outliers are marked by circles.


Based on these results, the measurement of reference hearing thresholds at PTB will in future be performed with a limit value of 1000 ms for the response times and it is recommended to adopt this value in the corresponding ISO 389-9 (2009) standard. Nevertheless, it should be noted that the influence of 2.6 dB on the hearing thresholds measured with different response-time limits is in the order of empirically determined standard deviations for repeat measurements in audiometry (approx. 2.5 dB).




M. U. Bug, M. Strüp, S. Vollbort, T. Fedtke. 2023. Influence of response-time limits on automated hearing threshold determination. Int. J. Audiol. Opens external link in new windowhttps://doi.org/10.1080/14992027.2023.2250554

ISO 389-9. 2009. Acoustics – Reference Zero for the Calibration of Audiometric Equipment - Part 9: Preferred Test Conditions for the Determination of Reference Hearing Threshold Levels. Geneva: International Organization for Standardization.

ISO 8253-1. 2010. Acoustics – Audiometric Test Methods – Part 1: Pure-Tone Air and Bone Conduction Audiometry. Geneva: International Organization for Standardization.

U. Richter. 1979. Ein automatisierter Meßplatz für subjektiv-akustische Messungen. PTB-Bericht PTB-Ak-18, ISSN 0340–8639.



Marion Bug, FB 1.6, E-Mail: Opens local program for sending emailmarion.bug@ptb.de