Donald Hobson comments on Convolution as smoothing