Best Practices for On-Demand Streaming
Part 3b. - Deinterlacing and Inverse Telecine
Deinterlacing
Television systems use a technique called interlacing to paint a picture on the TV screen. With interlacing, a television raster “draws” every other picture line on the screen. Then it goes back and inserts lines between the first set. On older televisions this was very necessary. If the raster painted the image down the lines of phosphors of the TV screen, the top pixels would begin to fade before the raster painted the bottom pixels. The result was a kind of flicker. If the raster interlaced the process, painting one set of lines from top to bottom before painting the second set, the flicker was reduced. The result is an odd and even field of information that goes by so fast that all you see is a blended single image.
Desktop video doesn't have to worry about interlacing. In fact, it improves if you can remove the interlacing. Some camcorders offer a "progressive" feature which records video already deinterlaced. This saves time but may not be the best look for your footage if you are also putting it to tape. Some capture cards offer a deinterlacing feature. Or you can usually deinterlace in your digital editor. Deinterlacing usually isn't an issue if your video falls below half size at capture. But if you grab full screen, keep an eye out for it.
Inverse Telecine
Film is shot at 24 frames per second. Depending on the video standard, television might run at 25 to 30 frames per second. To account for this discrepancy, extra frames are added to the film. This process of converting film to video is called telecine. When you are encoding your digital content from film, there is no need to encode these blank frames. If available, use an Inverse Telecine conversion to bring the clip back down to 24 frames per second.
