3D interactive virtual reality museum exhibits should be easy to use, entertaining, and informative. If the interface is intuitive, it will allow the user more time to learn the educational content of the exhibit. This research deals with interface issues concerning activating audio descriptions of images in such exhibits while the user is navigating. Five methods for activating audio descriptions were implemented and evaluated to find the most effective. These range roughly on a passive-active continuum. With the more passive methods, an audio explanation was triggered by simple proximity to an image of interest. The more active methods involved users orienting themselves and pressing a button to start the audio. In the most elaborate method, once the visitor had pressed a trigger button, the system initiated a "tractor-beam" that animated the viewpoint to a location in front of, and facing, the image of interest before starting the audio. The results of this research suggest that the more active methods were both preferred and more effective in getting visitors to face objects of interest while audio played. The tractor-beam method was best overall and implemented in a museum exhibit.