01.11.2018
For the recently developed circumaural audiometric RadioEar DD65v2 headphone, reference hearing thresholds were determined for pure-tone audiometry and free-field equalization data for use in speech audiometry. Reference hearing thresholds are specified in the ISO 389 series of standards "Acoustics – Reference zero for the calibration of audiometric equipment". For tone audiometry, the reference hearing thresholds of 25 otologically normal male and female test subjects between the ages of 18 and 25 were determined by means of sinusoidal signals of fixed frequencies between 125 Hz and 8 kHz. For this purpose, the individual hearing threshold of each of the test subjects was determined with an automated measuring arrangement by means of the so-called "bracketing method" according to ISO 8253-1.
As a background sound pressure level below the normal hearing threshold of the young test subjects is necessary, the measurements were carried out in PTB's anechoic audiometry room, which is isolated from the building and therefore has very little background noise. Due to the automated procedure, the influence of the investigator was reduced to a minimum. Prior to the hearing tests, the suitability of the test subjects was examined by an audiologist according to the requirements of the ISO 389 series of standards.
To measure the equivalent threshold sound pressure levels (ETSPLs), the headphone was operated with the signals that had been determined in the hearing tests. These signals corresponded to the mean hearing threshold of the group of test subjects. Furthermore, the sound pressure level generated according to IEC 60318-1 on an ear simulator for circumaural headphones was determined. This sound pressure level represents the mean sound pressure level generated at the eardrum of the test subjects and serves to calibrate pure-tone audiometers which are equipped with this headphone.
From the individual data, the quartiles and the mean values of the hearing thresholds were determined. For the first time, a comprehensive measurement uncertainty budget was established for such reference data determined in hearing tests and a measurement uncertainty according to GUM was stated. Figure 1 shows the equivalent reference hearing thresholds for the DD65v2 headphone as mean values with a 95% confidence interval. The frequency dependence of the data results from the combination of the frequency characteristics of the ear and of the headphone: the highest sensitivity is at 2-3 kHz; here, the sound pressure levels are the lowest, and they increase for lower and higher frequencies.
Figure 1: Mean values of equivalent reference hearing threshold for pure tones for the RadioEar DD65v2 headphone. The error bars represent the 95% confidence interval.
The free-field equalization data sets were determined based on the IEC 60268-7 standard "Sound system equipment – Part 7: Headphones and earphones". They were determined for third-octave band signals with center frequencies of between 125 Hz and 8 kHz on 17 otologically normal test subjects. These data sets are required for calibrating speech audiometers which are equipped with the RadioEar DD65v2 headphone. In the experiment, the test subjects compared – also by means of an automated process – the loudness level of the signal generated by the headphone with the loudness level of the same signal in the free field. The free-field signals are generated by a loudspeaker located at a defined distance to the test subject in the anechoic room. At the same loudness levels, the difference between the sound pressure level in the free field and the sound pressure level of the headphone (measured with an ear simulator according to IEC 60318-1) was calculated. A comprehensive measurement uncertainty budget and a measurement uncertainty according to GUM were stated as well. Figure 2 shows the free-field sensitivity levels determined with this method as mean values with a 95% confidence interval. In general, the signals which were played via the headphone were perceived as being softer than in the free field. Hence, these values are negative. Especially in the frequency range around 4 kHz, much greater sound pressure levels are necessary with the headphone in order to perceive the same loudness level as in the free field.
Figure 2: Mean values of the free-field sensitivity levels for the RadioEar DD65v2 headphone. The error bars represent the 95% confidence interval.
Contact:
Dr. Marion Bug, FB 1.6, AG 1.61, E-Mail: 
marion.bug(at)ptb.de