Mental image generation and the contrast sensitivity function.

In one experiment, observers were instructed to generate small and large visual mental images in a light-adapted or a dark-adapted viewing condition, and to rate the vividness of each image. In light-adapted viewing, small images were generated faster than large images. However, the pattern of results was reversed in dark-adapted viewing. Furthermore, the dark-adapted images were more vivid than the light-adapted images. The results show that mental images can be mapped onto some regions of the contrast sensitivity function (CSF), so that the latency (and vividness) of a particular image can be predicted. Since the CSF reflects visual processes occurring mainly in V1, the results agree with functional magnetic resonance imaging data in indicating that early visual pathways participate in imagery. Nonetheless, imagery and perception may involve different processes hosted by the same neural structures as indicated by the direct relationship between image latency and contrast sensitivity.