Intensity resolution and loudness in high-pass noise

Intensity resolution and loudness growth for a 1000-Hz tone were studied in the presence of high-pass noise (cutoff: 1800 Hz). Intensity resolution was measured for gated and continuous standards using a three-interval forced-choice (3-IFC) adaptive procedure. Loudness matches were obtained using an adaptive, alternate binaural loudness balance procedure. Three subjects listened in three conditions (1) quiet; (2) high-pass noise with a spectrum level of 32 dB SPL; and (3) high-pass noise with a spectrum level of 42 dB SPL. Noise levels were selected so that detection thresholds were minimally affected at the test frequency; however, for frequencies in the noise passband, thresholds were shifted to either 50 or 60 dB SPL, depending on the spectrum level of the noise. On average, loudness growth and intensity resolution were unaltered by the presence of the noise for tonal levels below 40 dB SPL; above 40 dB SPL the following was generally true: (1) intensity resolution for gated standards was well described by Weber's law except at the highest levels where the Weber fraction decreased; (2) intensity resolution for continuous standards showed a “near-miss” to Weber's law, but just-noticeable differences (jnd's) were slightly larger than those in quiet for the same SPL; (3) loudness was reduced. A comparison of jnd's for equally loud tones showed that loudness is less dependent on excitation spread than the jnd. That is, jnd's in the threshold-shifted ear were larger than the ones in quiet when the comparison was made for tones judged to be equally loud. This finding is qualitatively consistent with excitation-pattern models of these phenomena. The results were also evaluated in terms of the proportional-jnd theory, which yielded more accurate predictions of loudness growth with gated-standard resolution than with continuous-standard resolution for most comparisons.