The white noise standard sample method and application for subjective noise evaluation

噪声主观评价中的白噪声标准样本法及其应用

¹ School of Marine Science and Technology, Northwestern Polytechnical University, Xi’an 710072, China
² State Key Laboratory of Power Grid Environmental Protection, Wuhan 430000, China

Received: 1 November 2021

Abstract

The anchored scoring method is one of the widely used subjective sound evaluation methods. The comparability between subjective sound evaluation data obtained from different sample sets can be enhanced by selecting the standard sound sample as reference sample in subjective sound evaluation, thus facilitating the horizontal comparison between different evaluation results. However, the applicability of the standard sound sample method is poor in the experiments on evaluating some types of noise because the mechanisms of annoyance caused by different types of noise are discrepant. Therefore, the standard sound sample method for white noise that takes the existing pure tone as standard sample is proposed. Taking the evaluation results of the common anchored scoring method as benchmark, the experimental study of the performances of the standard sound sample method and the anchored scoring method was carried out according to the evaluation of the annoyance degrees of the two types of similar noise combination or dissimilar noise combination that have obviously different spectra and perceptual characteristics. The results show that the standard sound sample method for pure tone is suitable for noise with obvious tonal perception and that the standard sound sample method for white noise is suitable for broadband noise with weak tonal perception. Finally, the method of conversion between the annoyance evaluation results on different reference samples is studied to enhance the applicability of the standard sound sample method.

摘要

参考评分法是常用的声音主观评价方法之一, 在其中选取标准声样本作为参考样本可以提高不同样本组数据之间的可比性, 使不同评价结果之间的横向比较成为可能。然而, 不同类型噪声引起烦恼感的机制不同, 在部分类型噪声评价实验中, 已有标准样本法存在适用性不强的问题。为此, 在已有标准样本为纯音的基础上, 提出了白噪声标准样本法, 针对频谱特性和感知特性明显不同的3组相似噪声组合与不相似噪声组合的烦恼度评价, 以待测样本法评价结果为基准, 开展了2类标准样本法性能的实验研究。结果表明, 纯音标准样本法更适合音调感明显的噪声, 白噪声标准样本法更适合音调感不强的宽带噪声。研究了不同参考样本法的烦恼度评价结果之间的转化方法, 扩大了标准样本法的适用范围。