My measurements suggest that the 4:3 frame is 640 by 480 pixels not 720 by 480 as suggested on the Watchdog site. Am I missing something here? Because I am making precise scientific measurements from individual frames, I would like to calibrate the lens (convert pixels to angular distances) and check for possible lens and CCD distortions. Does anyone have any experience in doing this that they would care to share? Specifically does anyone know if the true angular field of view, at full zoom, changes with focus?

Hi Edward,

DV is 720 x 480 and uses non-square pixels (i.e., rectangular pixels) for proper viewing on a standard NTSC monitor such as the television in your living room. Your computer uses square pixels for its display, so you're looking at a frame of DV which has been resized to 640 x 480. It's simply an issue of square pixels vs. non-square pixels. Hope this helps,

Edward, if you are serious about "precise", I suggest you check the link below regarding video aspect ratios. The 720x480 DV standard is not actually for a 4:3 aspect ratio frame because when a DV frame is converted to NTSC analogue video, the left and right edges of the picture are in the overscan area. So, the viewable area is closer to 710x480 and it is this area that is defined as having the 4:3 aspect ratio.

On a computer monitor which has square pixels, all 720x480 pixels are visible, but the software used to view the video usually converts the pixel aspect ratio so that it looks correct (i.e. no stretching). If this conversion is done accurately, the 720x480 pixel image should be slightly wider than true 4:3.

Richard

http://www.uwasa.fi/~f76998/video/conversion/