Marshall L
J Aud Res. 1980 Apr;20(2):107-12.
Young adults (N = 11) with musical training, and with previous practice in the use of direct magnitude estimation of loudness for pure tones, judged the loudness of 3 orchestral excerpts each lasting 15-25 sec, varied in 5 -db steps over a 35 -db range. Excerpt I was relatively constant in both pitch and loudness; Excerpt II varied in loudness but did not change greatly in pitch; and Excerpt III changed both in pitch and loudness. The slopes of power functions for loudness judgments as a function of SPL fell within the usual range for pure-tone and for white-noise stimuli, as did indexes of determination; but both measures differed significantly from those in the literature for continuous speech. Apparently the acoustic variability in speech is an insufficient explanation for the difference between speech and non-speech stimuli in these areas, a finding which points to the contribution of non-acoustic parameters, such as semantic meaning, in loudness judgments for speech.
11名接受过音乐训练且曾使用纯音响度直接量值估计法进行过练习的年轻人，对3段时长为15 - 25秒的管弦乐选段的响度进行了判断。这些选段在35分贝的范围内以5分贝的步长变化。选段I在音高和响度上相对稳定；选段II响度变化但音高变化不大；选段III在音高和响度上均有变化。响度判断的幂函数斜率作为声压级的函数，落在了纯音和白噪声刺激的通常范围内，决定系数也是如此；但这两种测量方法与文献中连续语音的测量结果均有显著差异。显然，语音中的声学变异性不足以解释语音和非语音刺激在这些方面的差异，这一发现表明非声学参数（如语义）在语音响度判断中起到了作用。
