Deïnterlacing
Video quality is heavily affected by the way a graphics card can deinterlace the images. The importance of deinterlacing is one that stems from the past; when the older standards (NTSC for the USA, PAL for Europe) were implemented, the technology was not quick enough to render all lines on a screen at once. The solution was to send the images interlaced, meaning that in stead of transmitting 30 (NTSC) or 25 (PAL) full images per second, 60 (NTSC) of 50 (PAL) half images are transmitted per second. Such a half image is called a “field”, containing either the odd or even lines of the full screen. An old CRT TV works interlaced and will show these fields in turns. As the human eye is less sensitive when viewing moving images, this was a smart and effective solution, as this way higher resolutions were possible.
LCD-TV’s and PC’s however work non-interlaced, so the whole image is placed on the screen every time. To show an interlaced video on a non-interlaced screen, the video has to be converted to non-interlaced or progressive video. This process is called deinterlacing, quite an important process, as you can not just combine the two half screens because there is a little time difference between the two halves; PAL has a difference of 1/50 of a second, NTSC of 1/60 of a second.
The image below shows one of the problems that can occur when deinterlacing. At the top a zoomed image is shown of a white line moving across a black background. The bottom part of the image shows that in interlaced video the odd and even lines are shown in turns, field 1 shows the odd lines, field 2 the even lines that are transmitted 1/50 of a second later.

Deïnterlacing is less trivial than it looks.








