Hi, I am new member over here. I wanted to know how to perform discrete wavelet transform, probably using symmlet 8 wavelet with 16 Low/High pass filter coefficients using matrix transformation method when decomposing to a certain level the approximate coefficients i.e. sample size becomes less than the number of filter coefficients (say 8 samples compared to 16 filter coefficients). Matrix method have been given in several books e.g. "Ripples in Mathematics: The Discrete Wavelet Transform" - A. Jensen and A.la Cour-Harbo. However, the book only deals with cases where the number of filter coefficients are less than sample size.