Abstract
Following the lead of electronic computers, optical computers must adopt floating point calculation to allow for both high accuracy and high dynamic range. Given here is a method for using spatial encoding for that purpose.