Hi, did anyone ever test "deblocking" in DCT domain? Or maybe in image domain, but with direct effect on the resulting bitstream (supposed to mean: not only in postpressing). gruel