Google has an algorithm based on the "Discontinuous seam carving", one of Ariel Shamir already on the www.slashcam.de/artikel/Editorials/Ausblick-auf-neue- imaging techniques - Content-Aware Image Sizing - Context-sensitive Groessenveraenderung-of-Bildern.html # Conte (Siggraph 2007 featured technology), developed that is capable of context the size of videos change-sensitive, ie Videos are not simply the new size scaled to a letterbox marked or, but intelligent (and for the viewer almost imperceptibly adjusted). This is achieved by distinguishing between important (such as faces, objects in the foreground, actors) image content that are largely unchanged stay and less salient image content (such as background, sky), which can actually be cut without changing the image visible. Practically, the method is used whenever a building is to be scaled with a certain size to another without changing the aspect ratio (and thus the proportions) or to curtail it. We'll see when Google uses this technology for YouTube. ° = VY 8qMovlLlr_0 more infos at bei googleresearch.blogspot.com