I've already proven to you that I can make fake videos on my little laptop that you can't tell are fake.
No, you didn't, because as I said at the time, I don't do QuickTime movies. Sorry, just an irrational dislike of the format. Try making something that will play in the standard Windows Media Player next time (I'm a neanderthal that way).
Besides, you didn't actually create a fake, you just cloned a real aircraft. That's not the same thing at all.
There are little details of motion that betray the fakery. That's what my paper is all about.
What's that old saying... "You can't see the forest for the trees." That's exactly the trap your study fell into.
You went frame by frame and tried to analyze each one individually to try and discern some sort of fakery. When you try going that deep into a moving image, and ending up with individual images, you actually lose the essence of the motion you're studying. Sometimes the best way to detect incorrect motion is by simply watching a video in real-time, and watching the suspect sequence over and over. That can be far more effective, since you can see how the motion looks.
I wish I had a clip from
Aliens when Ripley first drives the powerloader so that you could watch it. Then I'd ask you what, if anything, looks odd about the way it moves. It's only by watching the clip at normal speed that the motion can be studied and the oddity noticed (it's very subtle; I doubt most people caught it). If you went frame by frame trying to analyze what's wrong with the motion you'd fail to see the problem because frame by frame you're not studying motion anymore - you're studying each frame as a self-contained entity.
To catch problems in motion, the image needs to be, well, in motion.
Here's another example: consider the case of a reaction shot close-up of an actor. In the shot the actor doesn't move or speak, but just blinks a couple of times. If that shot was shown to you in normal motion and with the shot reversed so that it played backwards without knowing which was which ahead of time, could you tell the difference? Could you tell which shot was forwards and which had been reversed to backwards? Remember, there's no speaking and no movement other than eye blinks. In other words, there's very little in the way of visual cues to differentiate between the forwards and backwards versions.
I could tell the difference. How? By the eye blinks. The way a blink looks in normal forwards motion is different from how it looks in reverse. It's subtle but definitely different.
That's the kind of fine detail in terms of motion I'm talking about.