Comparison of perception and objective measurements of the human hearing at low frequencies

Project Details


Standards on hearing like threshold and equal-loudness-level contours show the normal hearing sensitivity at low frequencies as following smooth slopes. However, recent non-invasive measurements of the forward middle-ear transfer function (FMETF) reveal a "resonance feature" seen as dip and a peak in the FMETF where the slope changes approx. 6 dB/octave (around 40-65 Hz depending on person). The change in slope is attributed to the shunting effect of the helicotrema. A preliminary study has been initiated in order to see if this resonance feature measured objectively is also found in perceptual data. The FMETF was measured for five subjects and an equal-loudness contour (ELC) was measured in a fine frequency grid from 20 Hz to 100 Hz in order to see if the same irregularities are found in the perception of low-frequency sound. Preliminary results indicates that this is the case for some people meaning that these people have a narrow frequency range where they are more sensitive and a narrow range where they are less sensitive compared to the standards. Since the frequency range is subject dependent this could explain why some people are annoyed by a low-frequency sound that is not audible to other people (see project on "Explanation of annoyance from low-frequency noise at very low levels"). The study was initiated in the summer 2008, and the last ELC measurements for the five subjects were finished in the beginning of 2009 and the results were presented at two conferences during 2009. The results show significant individual differences and a larger study is needed in order to obtain sufficient data for understanding how the hearing function works at low frequencies. The study is made as collaboration between Aalborg University and Torsten Marquardt from UCL, Ear Institute, London.
Effective start/end date01/08/200731/08/2010


  • <ingen navn>


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.