The bias incurred during non-symmetric rounding occurs because rounding decisions at the midpoints always go in the same direction. In symmetric rounding, the decision on which direction to round is based on the sign of the number. For rounding towards highest magnitude, a binary value corresponding to 0.499 is added to the accumulator result, and the inverse of the accumulator sign bit is added as a carry-in before removal of the m LSBs. As is generally the case, there are as many positive as negative numbers; the result should not be biased in either direction. This rounding mode is commonly used in general applications, mainly because it is equivalent to the MATLAB software function round(x) .