Video on digital TV will be compressed using a scheme called MPEG-2. It takes advantage of how the eye perceives color variations and motion. Inside each frame, an MPEG-2 encoder records just enough detail to make it look like nothing is missing. The encoder also compares adjacent frames and only records the sections of the picture that have moved or changed. If only a small section of the picture changes, the MPEG-2 encoder only changes that area and leaves the rest of the picture unchanged. On the next frame in the video, only that section of the picture is changed.

MPEG-2 has some problems, but it's a good compression scheme and it's already an industry standard for digital video for DVD-Videos and some satellite television services. One problem with MPEG-2 is that it's a "lossy" compression method. That means that a higher compression rate gives a poorer picture. There's some loss in picture quality between the digital video camera and what you'll see on your television. However, the quality is still a lot better than an average NTSC image. And using these compression schemes, MPEG-2 can reduce the amount of bits by about 55 to 1!

With that ratio, there's a lot of information that get's thrown away, but there's still enough to look like everything is still there. The human ear isn't as easy to fool, though. It's much more sensitive to subtle changes in sound. Digital TV is going to improve the sound over today's television using advances in digital sound developed over the last two decades.

Digital Sound