Abstract: Binary arithmetic coding involves recursive partitioning the range [0,1) in accordance with the relative probabilities of occurrence of the two input symbols. We describe a modification of ...