I was curious about this so I found a
video of a guy lying prone and firing an AR-15. I took the 1 second from the Butler clip, repeated it 3 times in slow motion, and overlaid the three shots taken by this guy. I reduced the size of the overlaid video to better match the size of Crooks, but even then you can see how much higher-resolution that video is. (To better approximate the amount of compression on the figure of Crooks, manually change the resolution in YouTube to 360p.)
Main points:
1) The AR-15 has very little recoil, even in high-def. (For a hilarious video in response to a hysterical journalist, see
this vid.) And very little smoke/muzzle flash (the puff you see in the initial video is dirt.)
2) The Crooks video is zoomed in and
highly compressed. You can't make out any distinct features that are greater than probably 1 inch. Due to the nature of video compression, such a small movement as produced by the AR-15's minimal recoil probably wouldn't even register. (
Cool video on video compression and how it basically interpolates frames to account for motion, and can ignore tiny changes as a result, in addition to all kinds of other information. I don't know enough about it, but it seems to me that if a blurry figure makes a tiny change for a frame or two, the compression would just "smooth" that out and show what appears to be a static figure.)
We've seen the opposite problem with video compression
a couple days ago. In that one, what looked like a fast and violent movement of perhaps a purse (as if it got shot) turned out, in higher definition, to be a woman moving her arm. In reality, she had moved her elbow back, and then jerked it forward when she heard the shot. In the compressed video, it looks like her arm snaps backward.
Basically, there isn't enough visual information in this video to say for sure one way or the other whether there was any recoil. If there was recoil, we probably wouldn't be able to see it, and the video would look the same as if there was no recoil. We need more video!