Temporal Preprocessing - A Practical Guide to Video and Audio Compression - page 541

Figure 33-4 Field averaging.

240 pixels, and
unless some line-doubling or aspect-ratio correction takes place, this looks rather strange.
Consider the scale of the output when doing this. There are significant implications if you
want to retain a full-sized raster.
Now look at the example of what happens to our four-frame sequence (Figure 33-5)
if we discard a frame.
In the second and third frames there are some residual noise artifacts. This may be
due to the format used to deliver the video. In this case, the original video was presented

So an image size that has a resolution of 720 × 480 becomes 720 ×

Frame 1 Frames 1 & 2a Frames 2b & 3 Frame 3

Figure 33-5 Field-dropping artifacts.