Greenwood D D
School of Audiology and Speech Sciences, University of British Columbia, Vancouver, Canada.
Hear Res. 1993 Feb;65(1-2):1-39. doi: 10.1016/0378-5955(93)90198-a.
In Greenwood [J. Acoust. Soc. Am. 33, 484-502 (1961a)] the ratio of masked signal threshold to masker level (S/M) decreased about 4 dB at a masker level of about 50 dB SL, the 'transition' level, when noise bands were subcritical but not when supercritical. Schlauch et al. [J. Acoust. Soc. Am. 71, S73 (1982)] report a related result. A pilot study [Greenwood, Harvard Psychoacoustic Lab. Status Report 37, 8-9 (1961)] in which pure tones masked identical tones in-phase showed a larger change in S/M. Detailed tone-tone growth-of-masking curves from over a dozen subjects in 1967-69, and in 1960, are reported here. A transition in slope, of variable abruptness, often begins to occur at about 50 dB SL, dropping S/M ratio by 6 to 8 dB or more [Rabinowitz et al., J. Acoust. Soc. Am. 35, 1053 (1976)]; the curves sometimes possess two segments, sometimes are simply convex. All have overall slopes less than 1.0, known also as the 'near miss'. Consistent with other results [Zwicker, Acustica 6, 365-396 (1956); Viemeister, J. Acoust. Soc. Am. 51, 1265-1296 (1972); Moore and Raab, J. Acoust. Soc. Am. 55, 1049-1060 (1974)], addition of low-level wide-band and high-pass noise was found to counteract the change in S/M, i.e., to raise the high-level section of the growth-of-masking curve. However, the ability of narrow 'band-pass' noise to exert this effect was greatest when added at a frequency ratio (band/masking-tone) of 1.3 to 1.5, which seems more closely to link the effects of added noise to the effects of increasing a masking band from sub- to supercritical width (above). Interpretation of the decrease in DL with level begins by noting that the 'transition' level correlates approximately with the level at which a primary unit population excited by a given pure tone begins rapidly to expand basally. Underlying this, the basalward shift of a tone's displacement envelope peak accelerates at about the same level [Rhode, J. Acoust. Soc. Am. 49, 1218-1231 (1971); Sellick et al., J. Acoust. Soc. Am. 72, 131-141 (1982)].(ABSTRACT TRUNCATED AT 400 WORDS)
在格林伍德的研究中[《美国声学学会杂志》33, 484 - 502 (1961a)]，当噪声频段为亚临界但非超临界时，在约50 dB SL的掩蔽声级（“转折”声级）下，被掩蔽信号阈值与掩蔽声级之比（S/M）下降了约4 dB。施劳赫等人[《美国声学学会杂志》71, S73 (1982)]报告了一个相关结果。一项初步研究[格林伍德，哈佛心理声学实验室状态报告37, 8 - 9 (1961)]中，纯音对同相的相同纯音进行掩蔽，结果显示S/M有更大变化。这里报告了1967 - 1969年以及1960年来自十几名受试者的详细的音 - 音掩蔽增长曲线。斜率的转变，其突然程度各异，通常在约50 dB SL时开始出现，使S/M比值下降6至8 dB或更多[拉比诺维茨等人，《美国声学学会杂志》35, 1053 (1976)]；这些曲线有时有两段，有时只是凸形。所有曲线的总体斜率都小于1.0，这也被称为“接近失配”。与其他结果一致[茨维克，《声学学报》6, 365 - 396 (1956)；维梅斯特，《美国声学学会杂志》51, 1265 - 1296 (1972)；穆尔和拉布，《美国声学学会杂志》55, 1049 - 1060 (1974)]，发现添加低电平宽带和高通噪声可抵消S/M的变化，即提高掩蔽增长曲线的高电平部分。然而，如果以1.3至1.5的频率比（频段/掩蔽音）添加窄“带通”噪声，其产生这种效果的能力最强，这似乎更紧密地将添加噪声的效果与将掩蔽频段从亚临界宽度增加到超临界宽度（见上文）的效果联系起来。对随着声级增加辨别阈降低的解释首先要注意到，“转折”声级大致与由给定纯音激发的初级单元群体开始迅速向基底扩展的声级相关。在此基础上，音调位移包络峰值的向基底方向移动在大约相同声级时加速[罗德，《美国声学学会杂志》49, 1218 - 1231 (1971)；塞利克等人，《美国声学学会杂志》72, 131 - 141 (1982)]。（摘要截取自400字）
