Has anybody (successfully) tried to deconvolve a seismic signal (time series in general) using R and package signal? I have the Matlab routines: [bb,aa] = zp2tf(zero,pole',gain); [bb,aa] = bilinear(bb,aa,df); transfer = freqz(bb,aa,f,df); But the (apparent) equivalents of R produce quite different results: ZPG <- Zpg(zero = zero, pole = pole, gain = gain) B <- bilinear(ZPG, T = 1/df) transfer <- freqz(h = B, Fs = 1/df, n = f) In Matlab, one complex vector (transfer) is returned, in R a different complex vector and a numeric vector are returned. Are the problems in the specification of zeros and poles (Hz vs. rad/s) or in parameter setting of freqz()? Or does anyone know someone who uses R for signal processing? Best and Thanks, Micha [[alternative HTML version deleted]]