Suppr超能文献

作为频率和强度函数的时间窗形状。

Temporal window shape as a function of frequency and level.

作者信息

Plack C J, Moore B C

机构信息

Department of Experimental Psychology, University of Cambridge, England.

出版信息

J Acoust Soc Am. 1990 May;87(5):2178-87. doi: 10.1121/1.399185.

Abstract

In an earlier article [Moore et al., J. Acoust. Soc. Am. 83, 1102-1116 (1988)], preliminary work on the temporal-window model of temporal resolution in the auditory system was described. The temporal window is conceived of as a temporal integrator that slides in time and that is implemented as an intensity-weighting function. The shape of the temporal window was estimated by measuring the threshold for a brief sinusoidal signal presented in a temporal gap between two bursts of noise as a function of the duration of the gap and the position of the signal within the gap. In this paper, a much more thorough examination of the effects of level and frequency on the shape of the window is presented, using the same basic technique. Temporal window shapes were measured at four different frequencies (300, 900, 2700, and 8100 Hz) and at three different masker levels covering a 20-dB range at each frequency. The shape of the temporal window was well described by modeling each side as the sum of two rounded-exponential (roex) functions. The equivalent rectangular duration (ERD) of the window decreased from about 13 to 9 ms as the center frequency increased from 300 to 900 Hz, but decreased only slightly, to 7 ms, as the center frequency increased to 8100. The greater ERD at 300 Hz does not seem to be explicable in terms of "ringing" in the auditory filter. The ERD decreased somewhat with increasing level, for example, having a value of about 10 ms at 2700 Hz with a 20-dB masker spectrum level and about 7 ms with a 40-dB masker spectrum level.

摘要

在之前的一篇文章中[摩尔等人,《美国声学学会杂志》83, 1102 - 1116 (1988)],描述了关于听觉系统中时间分辨率的时间窗口模型的初步工作。时间窗口被设想为一个在时间上滑动的时间积分器,并被实现为一个强度加权函数。通过测量在两个噪声脉冲之间的时间间隙中呈现的短暂正弦信号的阈值,作为间隙持续时间和信号在间隙内位置的函数,来估计时间窗口的形状。在本文中,使用相同的基本技术,对声级和频率对窗口形状的影响进行了更全面的研究。在四个不同频率(300、900、2700和8100赫兹)以及每个频率覆盖20分贝范围的三个不同掩蔽声级下测量了时间窗口形状。通过将时间窗口的每一侧建模为两个圆形指数(roex)函数的和,可以很好地描述时间窗口的形状。随着中心频率从300赫兹增加到900赫兹,窗口的等效矩形持续时间(ERD)从约13毫秒减少到9毫秒,但当中心频率增加到8100赫兹时,仅略微减少到7毫秒。300赫兹时较大的ERD似乎无法用听觉滤波器中的“振铃”来解释。ERD随声级增加而有所降低,例如,在2700赫兹时,掩蔽声声谱级为20分贝时ERD值约为10毫秒,掩蔽声声谱级为40分贝时约为7毫秒。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验