This paper describes an interface which helps people maintain a sense of spatial context while navigating virtual real-world scenes. First, a single panoramic image of the entire space is constructed from the separate partial, but detailed, images which constitute the original video sampling of the scene. The user can then navigate through this real-world data by manipulating either the panoramic overview or the original detailed views appearing in a separate window. Clicking or dragging the cursor over regions in the panoramic overview updates the corresponding detailed view. Using the panorama in this way frees the user from the traditional linear modes of interacting with virtual real-word scenes. In addition, interacting with the detailed view highlights the corresponding region in the panoramic overview and leaves a "trail" of the user's path through the space. These methods of visualizing and interacting with digital video described in this paper can also be appli...