The absence of a measurable “critical band” at low suprathreshold contrasts

Subjects adjusted the contrast of a computer-generated test grating until it appeared to have the same overall contrast as a 6.25-c/deg grating of fixed contrast. Results show that the relative phase of the components of a compound test grating does not affect the perceived contrast of the grating. This suggests that independent mechanisms may process the components, as is the case at contrast threshold. The component contrasts, however, appear to be combined by a square-law summation mechanism below 20% contrast, which results in the absence of a measurable “critical band” in grating perception at such contrast levels. Thus, these results are also consistent with a simple single-channel model in which perceived contrast is directly related to root-mean-square contrast. It is not possible to reject either the single- or multiple-channel hypothesis on the basis of these experiments.