Select to sum the frames in a 16-bit buffer and then linearly scale the result to 8-bits (1-254). If Fix Scale is checked, scaling is controlled by Min and Max otherwise it is based on the actual minimum and maximum 16-bit values.