A team of scientists at M.I.T. are developing a new process for analyzing video for minute, pixel-by-pixel variations in color, then using the information to amplify and exaggerate motion that's captured in the video but not visible to the eye. The developers think it has ramifications in medical imaging, where it could radically simplify quick patient evaluations. For more info, including a video presentation from Siggraph 2012, visit the Eulerian Video Magnification website.