I have a question about the Normalisation process. Is the spatial information stored in the header or the image itself? Furthermore, I have noticed that the normalised image (when using the "Normalise(Estimate)" function in SPM8) is half the size of the original image, and am wondering why it is so.