Quantitative Measurement of Signal-to-Noise Ratio with the Percentile Analysis according to IEC 60118-15
* Presenting author
One way to analyze the effect of a noise reduction in hearing aids is measuring the improvement of signal-to-noise ratio after signal processing. However, at the output of the hearing aid, speech and noise are mixed together so that special methods are necessary to determine the SNR, e.g. the method of Hagerman and Olofsson. Another possibility is using the percentile analysis according to IEC 60118-15. This analysis has already been used as qualitative evaluation of the SNR. However, a quantitative interpretation of the results has not been shown yet. Therefore, in this work, it is investigated how the percentile analysis can be used to determine the SNR at the output of hearing aids. To this end, the results of the percentile analysis for a superposition of the International Speech Test Signal (ISTS) and International Female Noise (IFnoise) are investigated. These results show that a linear relation between the percentile analysis and the SNR can be found. Next, this relation is used to determine the SNR of data recorded at the output of a hearing aid. Finally, the same data is analyzed with the method of Hagerman and Olofsson and both results are compared to each other.