It is demonstrated that audio information can be extracted from silent high-speed video with a simple and fast optical technique. The basic principle is that the sound waves can stimulate objects encountered in the traveling path to vibrate. The vibrations, although usually with small amplitudes, can be detected by using an image matching process. The proposed technique applies a subset-based image correlation approach to detect the motions of points on the surface of an object. It employs the Gauss-Newton algorithm and a few other measures to achieve very fast and highly accurate image matching. Because the detected vibrations are directly related to the sound waves, a simple model is introduced to reconstruct the original audio information of the sound waves. The proposed technique is robust and easy to implement, and its effectiveness has been verified by experiments.