The Visibility Level (VL) is the ratio between the target contrast with its near background and the targetcontrast at detection threshold. The psychophysical data which contributed to Adrian's model of the VL were collected in laboratory conditions, with a number of differences compared to the reference driving scenario. We discuss some of these differences, and their impact on visual performance, through a series of 3 experiments. The first experiment shows that the VL is overestimated in a simple task conditions, compared to a double task conditions. The second experiment shows that the background complexity and apparent motion both lower the detection performance. Both of these experiments show an effect of target eccentricity. Finally, in a field experiment, it was found that even a low demanding driving task has a negative impact on target detection. The implications of these findings are discussed in terms of road lighting design.
