It is well recognized that attention enhances perception. But can sustained attention be too much of a good thing? A study by Sam Ling and Marisa Carrasco in Nature Neuroscience charts the time course of visual perceptual effects of sustained attention, and finds that it can actually impair perception.

Attention is thought to enhance the representation of a stimulus in a manner similar to increasing its physical contrast: more contrasting stimuli are easier to see, but focusing attention on a particular stimulus means that even a relatively low-contrast stimulus can be easily discriminated, almost as if attention 'turns up' the physical contrast of the stimulus. Neurophysiological studies also show that sustained attention amplifies the contrast gain of neural responses in the early visual cortex, providing a neural account for these behavioural enhancements.

However, attention aside, viewing a stimulus for a long time results in contrast adaptation, which in turn means that greater contrasts are required to discriminate stimuli. The magnitude of this adaptation increases with the intensity of the stimulus, so that adapting to a higher contrast stimulus results in a requirement for greater contrasts to achieve the same level of performance, and a longer period to recover from the adaptation.

Attention therefore enhances perception by increasing contrast gain, but higher contrast stimuli result in lowered contrast sensitivity after prolonged exposure. In their behavioural study, Ling and Carrasco pit these two opposing effects against each other. Four observers viewed vertical adapter 'gratings' (alternating light and dark stripes) at four locations on a screen. When they were asked to focus their attention on one particular grating, they later needed less contrast between the light and dark stripes of a test grating to discriminate its orientation. This effect only occurred when the test grating was presented at the same location as the adapter grating to which the observer had been paying attention.

Interestingly, this attentional enhancement occurred only when observers paid attention to the adapter gratings for a short time (∼300–500 ms). When they paid attention to the adapter gratings for a longer period of 4–8 seconds, their performance actually worsened: they needed more contrast in the test gratings in order to successfully discriminate its orientation. However, discriminating test gratings presented at locations the observers had not been attending to was easier: participants needed less contrast to discriminate these test gratings. So, although attention 'turns up' the perceived contrast of a visual stimulus, prolonged adaptation to this enhanced signal can worsen performance. Further studies are needed to identify the loci of these effects.