Quantization
	n Clearly if we encode a coefficient to less precision we need fewer bits to do so. n A suitable technique for this is to divide the coefficients by an integer and ignore the remainder on encoding and do the reverse on decoding.