Grantham D W
J Acoust Soc Am. 1982 Oct;72(4):1178-84. doi: 10.1121/1.388326.
In a two-interval, forced-choice task, observers discriminated a binaural noise whose interaural correlation r varied according to the function r(t) = m cos 2 pi fmt from an interaurally uncorrelated noise [NU; r(t) = 0.0]. The former stimulus produces a perceptual binaural "flutter," where the flutter rate is equal to the modulation frequency fm, and the amount of flutter corresponds to m. The stimuli were 0.4-octave-wide Gaussian noises with center frequencies of 500, 1000, 2000, or 4000 Hz. Presentation levels were 70-75 dB SPL; duration was 1.0 s. For a given modulation frequency, the peak interaural correlation m was varied in a blocked procedure, and thresholds were estimated as that value of m required for 75% discriminability. Plotting thresholds as a function of fm yields a modulation function, which can be interpreted as an "attenuation characteristic" of the binaural system. For the 500-Hz stimulus, thresholds increased from m - 0.28 at fm = 1 Hz to m - 0.80 at fm - 50 Hz. For higher frequency stimuli, discrimination was generally poorer and the modulation functions were flatter: thus, for 2000 Hz, m = 0.54 at fm = 1 Hz and m = 0.71 at fm = 20 Hz; for 4000 Hz, m = 0.71 at fm = 1 Hz and m = 0.81 at fm = 10 Hz. The modulation function from an "ideal observer" (at 500 Hz) differed in both form and absolute level from those of real observers. The data are discussed in relation to previous work on dynamic binaural processing.
在一个双间隔的迫选任务中，观察者要从双耳不相关噪声[NU；r(t)=0.0]中辨别出一种双耳噪声，该噪声的双耳相关性r根据函数r(t)=m cos(2πfmt)变化。前一种刺激会产生一种听觉上的双耳“颤动”，其中颤动速率等于调制频率fm，颤动量对应于m。刺激声是中心频率为500、1000、2000或4000 Hz的0.4倍频程宽的高斯噪声。呈现水平为70 - 75 dB SPL；持续时间为1.0 s。对于给定的调制频率，峰值双耳相关性m在一个分组程序中变化，阈值被估计为达到75%可辨别性所需的m值。将阈值绘制成fm的函数可得到一个调制函数，它可以被解释为双耳系统的“衰减特性”。对于500 Hz的刺激声，阈值从fm = 1 Hz时的m = 0.28增加到fm = 50 Hz时的m = 0.80。对于更高频率的刺激声，辨别通常更差，调制函数更平缓：因此，对于2000 Hz，fm = 1 Hz时m = 0.54，fm = 20 Hz时m = 0.71；对于4000 Hz，fm = 1 Hz时m = 0.71，fm = 10 Hz时m = 0.81。“理想观察者”（在500 Hz时）的调制函数在形式和绝对水平上都与真实观察者的不同。结合之前关于动态双耳处理的工作对这些数据进行了讨论。
