Skip to main content
. 2007 Oct 24;27(43):11687–11699. doi: 10.1523/JNEUROSCI.1486-07.2007

Figure 8.

Figure 8.

Ratings of perceived intensity, averaged across subjects, for a subset of sinusoidal (left), diharmonic (center), and noise (right) stimuli. For the diharmonic stimuli, ratings are plotted as a function of the amplitude of the low-frequency component (the relative amplitudes of the high-frequency components depend on their frequency; see Materials and Methods). We only included groups of stimuli that comprised at least two stimuli sharing the same frequency content and differing only in amplitude. The sinusoidal and diharmonic stimuli are fit with power functions. For sinusoids, exponents are 0.75, 0.61, and 0.54 at 10, 50, and 100 Hz, respectively; for diharmonic stimuli, the exponents are 0.76, 0.64, 0.56, and 0.50 for (5 Hz, 25 Hz), (10 Hz, 50 Hz), (25 Hz, 75 Hz), and (10 Hz, 100 Hz), respectively. The mean correlation between predicted and observed perceived intensity was 0.99 for both sinusoidal and diharmonic stimuli. No exponent was fit to the data obtained from noise stimuli because there were too few data points per stimulus group (each group sharing bandpass frequencies).