As promised, here's part 2 of the tutorial discussing the conversion of footage from 30p to24p. I left off on part 1 by showing the difference between frame blending and Pixel Motion algorithms in Adobe After Effects. Pixel Motion seemed to be a slam dunk, but as I eluded to, it isn't. See the picture below.
What I found is that the Pixel Motion algorithm will produce these defects that look like someone dropped water on your footage. It happens occasionally with fast motion in the scene and always with large changes in brightness. I noticed it whenever my footage showed a camera flash or a stage light suddenly coming into the shot. The picture completely "freaks out" with camera flashes, as if CMOS sensors and camera flashes weren't incompatible enough. On top of that I was recording athletes who tend to move around a lot.
So again, the question is how to mask this error and the answer is quite simple. You use track opacity to temporarily switch back to frame blending whenever Pixel Motion fails. See the picture below. What keeps this method on from being obvious is that there is likely some type of fast motion or huge change in brightness going on, so our persistence of vision masks the temporary change to the blurry frame blended picture. Also the failure generally occurs in small blocks of frames, so if you're switching over to frame blending for, let's say 5 frames, that's only 0.2 seconds and your brain doesn't really pick up on the change. The picture below shows my time line where the temporary switch over to frame blending happens and the frame blended picture to contrast with the same Pixel Motion-ed frame above.
Is this as good as recording at 23.976 frames per second in the first place? Heck no. However, this method seems to work for the most part and it's better than being stuck with 30p footage. Given the choice between an obviously screwed up Pixel Motion frame and a blurred frame blending frame, guess which one I'm going to choose?