A frame of video (NTSC and PAL) are made up of two fields. The upper field consists of lines 1,3, 5,7 etc. The lower field consists of lines 2,4,6,8 etc.. DV is lower field dominant, which means, the lower field is the first to be recorded/played back. Each field represents 720x240 pixels of information. When combined, they create the 720x480 frame (using DV-NTSC as an example).
In NTSC video, the fields are refreshed every 1/60 of second making one FRAME = 1/30 sec. A still image (1 FRAME) captured or frozen from video will contain two fields. If there is much movement in the scene, the difference between the first and second fields can be noticeable. This difference is sometimes described as combing (the edges of vertical elements look like the teeth of a comb). This is because of the time difference between when the fields were recorded (1/60 sec) the element is in a different position in each field.
In order to get a clear image from a video still that exhibits combing, you need to deinterlace, that is, throw away one field. Since a field represents 1/2 of the image information (720x240 for DV-NTSC), whatever program you use must re-create the missing info in order to maintain the pixel count of the video (720x480 for DV-NTSC). Some programs do a better job than others. In any case, you will end up with a softened image as the newly generated pixels are usually an AVERAGE of the difference between the lines. Much sharpness is lost.
Still images from scans or digital photographs DO NOT have fields. Nothing is gained by deinterlacing them since they *were never interlaced*.
You really need to read the manuals and find some books that explain the basics of the technology you are using. Understanding this material will make you a more effective editor.
x