Research by David Eagleman shows that it takes about 80 milliseconds for the brain to process all of the sensory information and construct our model of reality, effectively generating consciousness. Our interactions occur with that afterglow, the 80ms delayed model that our brains conjure.
It is because of this 80ms threshold that sounds can appear out of time if far enough away. A sound which occurs over a distance of 30m from us will take longer than 80ms to reach our ears, and as such the brain cannot include it within the model, stitched together along with the visual input.