The last frame in my sequence appears to correspond roughly in time with the frame that you referenced in NIST 1-6 (the third frame on that page). So you can see that the top section has already dropped noticeably by the time of that frame. Whether there is enough tilt to make a difference (or any tilt) by this point looks debatable to me. http://www.
livevideo.com/media/playvideo_fs.aspx?. ..