http://covertoperations.blogspot.com/2006/03/further-analysis-on-whether-speed-of.html(hotlinks are in linked article)
"...three significant problems with their analysis:
1) they stopped measuring before the plane was all the way in,
2) they didn't account for variation in the speed per frame-- for instance, they show a decrease in speed of the plane in just three frames compared to four frames before impact: how do we know the decrease was not due to random sampling errors? When I previously analyzed the CNN footage, I noticed there were significant variations from frame to frame in terms of how much the plane moved, and there wasn't a consistent trend towards slowing (more on this below).
3) they seem to miss several frames of the Fairbanks video (which they used for the speed measure) that show the plane entering the building. If you look at the video here, with Quicktime, you can click through the frames easily -- and I count 10 impact frames. So they are measuring only one third of the total data available.
So with the Fairbanks video, I took two sets of measurements.
A) pre-impact movement of the plane-- how many millimeters the plane moves from left to right before it impacts the building, per frame. There were 14 frames here. I got an average of 1.6 mm per frame +/- 0.5 mm (S.D., Standard Deviation).
B) post-impact movement of the plane-- how many millimeters the plane moves INTO the building, per frame. There were 10 frames here. I got an average of 1.5 mm per frame +/- 0.7 mm (S.D., Standard Deviation).
Right off, these numbers didn't look very different, but just to make sure, I did a Student's T test to determine probability, and got a value of P = 0.624. This means that the probability that these two sets of numbers were the same by chance is well over 50%. Generally, P < 0.05 is considered statistically significant. So, the Fairbanks footage didn't really support the idea that the plane slowed upon impact.
But to be thorough, I analyzed two other sets of impact footage.
The second one I ananlyzed was the CNN footage, broken into frames here.
For this analysis, I simply measured the amount the plane length changed as it entered the building. If the plane slows upon impact, the plane length should decrease much less after a few frames than the early frames.
As the plane impacts, starting at frame 2, I measured the decrease in plane length in millimeters (on the video).
Here is what I got starting with the decrease from frames 2 to 3, and ending with the decrease from frames 12 to 13:
2, 3, 4, 7, 4, 8, 3, 7, 5, 5, 10
So, although the numbers jump around a bit, if anything, the length of the plane decreases by GREATER amounts (not lesser amounts as expected for slowing) the more the plane goes in. Clearly there is NO SIGNIFICANT SLOWING following the initial impact in this video.
Next I looked a third set of frames of the second hit, from Scott Myers:
This one is a little trickier, because we never see the whole plane, and have to adjust our measuring reference points a couple of times. First, I measured how much the nose advanced into the frame, then how much the bright spot on the wing-fuselage advanced, and then finally how much the tail advances into the building. This analysis is also tricky because of the angle (apparently the southeast CORNER of the building), and so we can't see exactly when the plane impacts. But since there not very many frames, it is safe to say the plane has impacted by frame 8.
Again, I got two sets of measurements.
A) pre-impact movement of the plane-- how many millimeters the plane moves from left to right before it impacts the building, per frame. There were 6 frames here. I got an average of 8.7 mm per frame +/- 2.3 mm (S.D., Standard Deviation).
B) post-impact movement of the plane-- how many millimeters the plane moves INTO the building, per frame. There were 5 frames here. I got an average of 6.8 mm per frame +/- 0.8 mm (S.D., Standard Deviation).
The P value was 0.115.
So, there was a slight trend towards deceleration upon impact in this footage, but this is not statistically meaningful.
CONCLUSION: THREE SETS OF VIDEO FRAMES ALL FROM DIFFERENT ANGLES SUPPORT THE IDEA THAT THE SECOND PLANE DID NOT SLOW UPON IMPACT. RATHER THE PLANE ENTERED THE TOWER AT FULL SPEED AND CONTINUED AT THIS SPEED FOR THE LENGTH OF ITS BODY