For the past couple of years, people wishing to create ultra-high-resolution panoramic photographs have been able to do so, using their own digital camera and a GigaPan robotic tripod.
The device slowly pans the camera back and forth across a user-determined vista, triggering it to take up to several hundred shots in the process. The included software then stitches all the photos together – side-to-side and top-to-bottom – creating one big panorama, which retains its resolution even when details are zoomed in on, much like Google Earth. So, what could top that? Time-lapse videos created using GigaPan Time Machine software, as it turns out.
GigaPan Time Machine was developed by Carnegie Mellon University’s CREATE robotics lab, and NASA.
Like the existing GigaPan system, it creates Google Earth-like panoramic images by stitching together a multitude of photos. Staying in the same location, however, the tripod “recaptures” the same panorama many times over a period of minutes, hours, days, or whatever interval the user wishes. Those multiple panoramas, played in sequence, become a time-lapse video of the scene. Viewers can then zoom in on or pan across various details within the scene, watching as they change over time, forwards or backwards. Any one frame of the video could potentially contain billions of pixels.
Needless to say, such video files would be a little too large for YouTube, so users can instead upload their videos to a dedicated GigaPan Time Machine website. It was created using a special feature of the HTML5 language, that allowed for the development of software architecture which made it possible to smoothly transition from one video section to another, without requiring plug-ins such as Adobe Flash or Quicktime.
