High resolution transmission electron microscopy (HREM) is an excellent experimental method to image grain boundary structures with atomic resolution. The advantage of the method is the short exposure time of only about one second that is needed to record an image. Other methods like Z-contrast imaging require much longer exposure times and are therefore much more prone to specimen drift during recording. However there is the remaining difficulty to HREM that the evaluation of experimental images is not straightforward and a thorough analysis of the images is necessary in order to deduce quantitative information with small error bars of only a few pm (10-15m). A second inherent difficulty common to all atomic resolution imaging techniques is that the information is retrieved from a very small area of a specimen. The question arising from that is: can we nevertheless be sure to obtain a representative answer to a “real world” material science problem? A positive answer to this question is given by the investigations presented here.