The preregistration for this experiment is available [here.][1] In Experiments [1][2] and [2][3], noticing rates varied little regardless of how long the unexpected object was visible onscreen, dropping less than 10% even when exposure time was cut from 5 seconds to just 1.5 seconds. Based on these results, and the results from the location reports in which people tend to report the unexpected object near fixation when they detect it, it seems that if an unexpected object is going to be detected, it will happen soon after onset. Adding additional exposure time, even doubling it, has only a minimal impact on the likelihood of noticing. Does this mean that the additional exposure time has *no* effect whatsoever? In this experiment, we explore whether additional exposure time has an effect on the quality of the representation of the unexpected object amongst those who notice it, even if the probability of noticing it in the first place is relatively unaffected. The unexpected object can appear in one of several hues, and rather than selecting its color from a drop-down menu, subjects report its color using a color slider. In this way, we can examine both accuracy of the color reports and precision of the reports as a function of exposure time. [1]: https://osf.io/mgd7y [2]: https://osf.io/yekzc/ [3]: https://osf.io/jx9vs/