Thresholds for formant‐frequency discrimination in isolated vowels
Thresholds for formant‐frequency discrimination in isolated vowels
复制标题
孤立元音中共振峰频率辨别的阈值
DOI：
--
复制
发表时间：
1990
期刊：
影响因子：
--
通讯作者：
D. Kewley
中科院分区：
文献类型：
--
作者：
D. Kewley
基于Flanagan [J. Acoust.美国社会27，613 - 617（1955）]。然而，这个范围可能并不代表听觉系统中分辨率的极限，因为对于大多数实验，受试者没有经过训练，并且在试验块内呈现了几对刺激。在本实验中，使用自适应跟踪范式，从受过良好训练的受试者在最小刺激不确定性下听元音获得阈值。对于模拟女性说话者的十个合成稳态元音，确定了F1和F2中的增量和减量的阈值。结果表明，共振峰频率的增量和减量的DL作为频率的函数线性增加，除了50%更高的阈值，其中谐波正好落在共振峰频率上。共振峰频率的ΔF/F比约有80%小于0.02。
The threshold of discrimination (or DL) of a shift in formant frequency is thought to be in the range of 3%‐6%, based on the oft‐replicated results of Flanagan [J. Acoust. Soc. Am. 27, 613‐617 (1955)]. This range may not, however, represent the limits of resolution in the auditory system because, for most experiments, subjects were not trained and several pairs of stimuli were presented within trial blocks. In present experiments, thresholds were obtained from well‐trained subjects listening to vowels under minimal stimulus uncertainty using an adaptive‐tracking paradigm. Thresholds were determined for increments and decrements in F 1 and F 2 for ten synthetic, steady‐state vowels modeling a female talker. Results indicated that the DL for both increments and decrements in frequency of the formants increased linearly as a function of frequency, with the exception of 50% higher thresholds where a harmonic fell exactly on the formant frequency. About 80% of the ΔF/F ratios were below 0.02 for formant freque...