Previous reports [D. Fishken et al., J. Acoust. Soc. Am. Suppl. 1 61, S61 (1977); B. Scharf and T. J. Horton, J. Acoust. Soc. Am. Suppl. 1 63, SI6 (1978)] of loudness adaptation based upon the method of successive magnitude estimation showed that the loudness of a steady high-frequency tone presented through earphones decreases with duration provided the tone is within 20 to 30 dB of threshold. Tones below about 1 kHz and also white noise show much less adaptation. These measurements have been extended to freefield listening. Seated in an anechoic room with his head held steady. O faced a loudspeaker 3.7 m away. Tones were each presented for 3 min at l0 dB above threshold. Every 20 s a light came on for 2 s. The O assigned a number corresponding to the loudness of the tone during the 2-s interval. Results from 26 Os showed that the loudness of a 4-kHz tone decreased after 2 min by 60% and after 3 min by 76%. These values are similar to those measured previously for earphone listening. Tentative results at 8, 10, and 12 kHz suggest somewhat less adaptation at these higher frequencies. Why loudness adapts at low levels and high frequencies and not otherwise remains a mystery. [Research supported in part by NIH.]

The level of a 1000-Hz tone burst required for localization and that required for detection were independently measured in the presence of a preceding tone burst as a function of the frequency difference between the tones. Three subjects made the judgments in an anechoic room. A 30-ms filtered burst—the signal—was delivered first from one double-cone loudspeaker 15° to the subject's left and then from another to his right. A third speaker directly in front delivered a 30-ms masker, 20 ms prior to each signal. At masker frequencies between 500–930 Hz, signal level had to be about 20 dB higher for perception of a change in direction, than for 80 percent detection. (Change in direction was usually heard as a small shift from left to right in the locus of the fused ensemble comprising masker and signal.) Evidently, the precedence effect applies not only when sounds are identical but also when they differ considerably in frequency. Moreover, even a soft lagging sound contributes to the perceived direction of the fused ensemble.