One more thing to keep in mind when it comes to frame grabs, The way video compression works, unless you make sure you grab the Key(I) Frame, you're likely suffering from not only image comperssion artifacts but also mpeg compress artifacts designed to "be close enough" with motion. Essentially any frame not on the key frame(essentially jpeg) is a composition of the last and next key frame with a series of differential elements applied, which is why on low end video you can see the video go sharp then soft and blurry or macroblock and then go sharp again. The frames that are sharp are key frames and the rest is progressive differences applied to make the in between frames which is where most of the space saving of video compression comes from.
That said High end mpeg-4 has frequent I frames and high bitrate reduces blurring and macroblocking on P and B frames, but it will NOT be as sharp as a still.
