Park H K, Bradley J S
Chonnam National University, Gwangju 500-757, Korea. [corrected]
J Acoust Soc Am. 2009 Sep;126(3):1219-30. doi: 10.1121/1.3192347.
Subjective ratings of the audibility, annoyance, and loudness of music and speech sounds transmitted through 20 different simulated walls were used to identify better single number ratings of airborne sound insulation. The first part of this research considered standard measures such as the sound transmission class the weighted sound reduction index (R(w)) and variations of these measures [H. K. Park and J. S. Bradley, J. Acoust. Soc. Am. 126, 208-219 (2009)]. This paper considers a number of other measures including signal-to-noise ratios related to the intelligibility of speech and measures related to the loudness of sounds. An exploration of the importance of the included frequencies showed that the optimum ranges of included frequencies were different for speech and music sounds. Measures related to speech intelligibility were useful indicators of responses to speech sounds but were not as successful for music sounds. A-weighted level differences, signal-to-noise ratios and an A-weighted sound transmission loss measure were good predictors of responses when the included frequencies were optimized for each type of sound. The addition of new spectrum adaptation terms to R(w) values were found to be the most practical approach for achieving more accurate predictions of subjective ratings of transmitted speech and music sounds.
通过20种不同模拟墙体传输的音乐和语音声音的可听度、烦扰度和响度的主观评分，被用于确定更好的空气声隔绝单一数值评分。本研究的第一部分考虑了诸如传声等级、加权隔音指数（R(w)）等标准测量方法以及这些测量方法的变体[H. K. 帕克和J. S. 布拉德利，《美国声学学会杂志》126, 208 - 219 (2009)]。本文考虑了许多其他测量方法，包括与语音清晰度相关的信噪比以及与声音响度相关的测量方法。对所包含频率重要性的探索表明，所包含频率的最佳范围对于语音和音乐声音是不同的。与语音清晰度相关的测量方法是对语音声音反应的有用指标，但对音乐声音则不太成功。当针对每种声音类型优化所包含频率时，A加权电平差、信噪比和A加权传声损失测量方法是反应的良好预测指标。发现将新的频谱适应项添加到R(w)值中是实现对传输的语音和音乐声音主观评分更准确预测的最实用方法。
