CompuServe Messages

You have sort of made my point. The red square you drew was done in RGB, by paint. It was encoded into NTSC when you saved it. That process is LOSSY. The red may have been an illegal color for video, which would possibly introduce artifacts when it was corrected/encoded. When you re-loaded it, it was decoded from NTSC to RGB, which does not recover the spatial color resolution lost by encoding, or restore lost saturation. If you save it again, you lose some more… etc. The generation loss comes from the RGB->NTSC encoding step. If you grab a frame from video and save it as a framestore, you won't have any loss. If you bring it into paint, it will be converted to RGB, and if you re-save the framestore you could get some loss, though not as much as you would from an RGB original, which had so much more to lose in the first place. Once you go into paint, saving as RGB will avoid all generation loss, as RGBs don't change at all with loading/saving.