Merged Discussion of femr's video data analysis

Rotation of the camera is an issue for the latter.
Not for my purposes. I just need two different angles. Ideally with 90° difference, but the bigger the angle, the better, and that one has a bigger angle with respect to Cam #3 than Dan Rather's, if only marginally.


You're planning on performing the tracing yourself then ? With what ?
I need to approximate the camera location. If the video I have to finally use is the second, then I may need to do the tracing. I don't know what I'm going to use for that purpose yet.
 
Is that 'vinetto'??? Its the only one I have used and only in 'one off' situations. Currently Xubuntu 12.04 on this machine.
Yes, Vinetto. Judging by the file format I doubt there's any other such tool. There can only be one crazy guy that does that kind of wacky job :D


If you could get the directory tree populated only with the thumbs.db files for brevity/bandwidth - it should be possible to simply script the extraction.
I can do that myself, but it's not trivial. I was trying to exhaust the other options. find(1) and sed(1) will do the magic, I expect.


(mu)torrent is only in alpha for linux but there are plenty of torrent options.
I used BitTornado myself. It offers the option to download selectively too, but not sure if in a practical way for this case. Anyway, I have the whole download here and can select Thumbs.db files only.
 
If the video I have to finally use is the second, then I may need to do the tracing. I don't know what I'm going to use for that purpose yet.

As you know I'd recommend SynthEyes (~£400)

For free you could use something like Tracker, but its tracking engine is far less accurate than SynthEyes and you may have feature latch issues.

ETA: The latest version of Tracker has certainly improved things, and uses region based pattern matching similar to SynthEyes.
 
Last edited:
There is no detectable drift/slew of fuselage horizontal or vertical entry location during impact.


I would tend to doubt you could detect it. The effect would be more of a rapid change in direction as the aircraft entered the building. (think letting go of a string on a rapidly swinging ball).

For example...for a 13 degree trajectory with 6 degree orientation (lateral) there would be nearly 20ft difference between nose and tail entry location.
You're using the building as a reference for angle? (again think swinging ball).

Earlier motion cannot practically be used to determine trajectory during impact, so there's only the during impact motion to go on.

No. Sir Newton seems to think otherwise. ;)
 
I would tend to doubt you could detect it.
Again your doubt needs to have some substance to it.

I can absolutely detect a 20ft difference between nose and tail entry point...and there isn't one.

The effect would be more of a rapid change in direction as the aircraft entered the building. (think letting go of a string on a rapidly swinging ball).
You're suggesting the aircraft was "deflected" from a trajectory which did not match the orientation...to a trajectory that did match the orientation...during entering the building ?

If so, rather bizarre, but end result is entry matching orientation

If not, have a think about what you're actually saying, ensuring you include that there was no detectable drift/slew in nose to tail entry point.

You're using the building as a reference for angle?
For trajectory, sure.

(again think swinging ball).
I'd rather think 540mph aircraft.

No. Sir Newton seems to think otherwise. ;)
Reverse-engineering aircraft 3d path during approach is a long and arduous task. Achimspok performed one in 1s increments, leaving out the last second iirc. That would need beefing up to about 30fps up to nosecone/building contact for useful extrapolation of path behaviour.

I'm not inclined to spend the time required.

Bahaviour during impact is quite compelling.
 
Yes, Vinetto. Judging by the file format I doubt there's any other such tool. There can only be one crazy guy that does that kind of wacky job :D
thumbup.gif


...I can do that myself, but it's not trivial. I was trying to exhaust the other options. find(1) and sed(1) will do the magic, I expect.
I'm only half understanding what you need -- but my focus was on extracting the images ss multiple *.jpg within the directory structure so you could browse it. But I have no idea how big the database or how complex......
...I used BitTornado myself. It offers the option to download selectively too, but not sure if in a practical way for this case. Anyway, I have the whole download here and can select Thumbs.db files only.
Understood - now - that was the first step I was focussed on. You are ahead of me. I will leave you with it.
 
I would tend to doubt you could detect it. The effect would be more of a rapid change in direction as the aircraft entered the building. (think letting go of a string on a rapidly swinging ball).

It may be pertinent to quote a passage from NCSTAR 1-2B, which I'm sure has been done before, but perhaps not in a while.

NCSTAR 1-2B said:
It was found that a strict relationship between the aircraft trajectory and orientation needed to be maintained in order to achieve an impact pattern consistent with the damage observed on the south wall of WTC 2.... Larger or smaller angles resulted in projected impact points with the engines spaced too far horizontally or vertically with the tip of the vertical stabilizer in the wrong location.

And so NIST concluded that the lateral fuselage orientation relative to trajectory was -3 degrees clockwise from the lateral approach angle. They don't place a measurement error on that; presumably they considered it to be a fraction of a degree.

So, they estimate inter alia that the trajectory was 15 degrees clockwise give-or-take 2 degrees, and the orientation was 3 degrees counterclockwise from there. They prefer 13 to 15 because of where the starboard engine came out, and because their simulations suggested no "substantial deviation" in its trajectory through the building.

(So, at the risk of belaboring, their best estimate is that the orientation was 10 degrees clockwise.)

I don't see any explicit discussion of whether and how the plane's yaw may have changed during impact.
 
...They prefer 13 to 15 because of where the starboard engine came out, and because their simulations suggested no "substantial deviation" in its trajectory through the building..
that's the part that always raises red flags with me - and frustrates me because I cannot get my head around the basic concepts. Must be age.

On the surface the idea that the engine was a relatively high density compact mass and therefore continued in a straight line looks reasonable. But even that raises questions of impact deflections - was the engine separated at outer wall impact? Did that separation introduce a deflection from pre contact direction of motion? Could there be anything in office partitions/contents sufficiently robust to cause deflection during transit through the building? If "no" to those then the NIST assumptions could be correct.

But the aspects I cannot get my aging brain around are two related factors which arise from "was the engine still rotating?" and "when did it stop?" - therefore gyroscopic precession if it was rotating and deceleration torque if it swallowed debris and quickly stopped spinning. And the time frames very short.
 
It may be pertinent to quote a passage from NCSTAR 1-2B
Why ? The passage you quoted has little to do with the the text you are responding to.

It was found that a strict relationship between the aircraft trajectory and orientation needed to be maintained in order to achieve an impact pattern consistent with the damage observed on the south wall of WTC 2
There are a range of orientation/trajectory combinations which allow a match.

The NIST wingtip penetration points are a bit off, especially the left'un.

Their orientation is off.

Their trajectory is off.

They prefer 13 to 15 because of where the starboard engine came out, and because their simulations suggested no "substantial deviation" in its trajectory through the building
Yet the angle between the NE corner and where it landed is 20 degrees...

And the simulated engine never even made it out the building.

Highly accurate then... :rolleyes:
 
Yet the angle between the NE corner and where it landed is 20 degrees...

So, what do you think caused this change in trajectory (the engine) while allowing it to maintain enough momentum to land where it did. I'm assuming you've run the numbers.
 
So, what do you think caused this change in trajectory (the engine) while allowing it to maintain enough momentum to land where it did. I'm assuming you've run the numbers.
Hows aboot deflections occuring due to crashing into a rather large building, with a pretty large explosion going on.

It was deflected.

It's pretty obvious why, though I doubt anyone will be able to determine exactly how many impacts deflecting the engine or when. It doesn't really matter.


As I've said, for NIST to SET their lateral trajectory angle (for all impact simulations) based upon an assumption about linear behaviour of the engine from impact to an exit point (which is a totally different angle to where it landed) was not a great decision imo.
 
It was deflected.

This I without a doubt.



It's pretty obvious why, though I doubt anyone will be able to determine exactly how many impacts deflecting the engine or when. It doesn't really matter.

True.


As I've said, for NIST to SET their lateral trajectory angle (for all impact simulations) based upon an assumption about linear behaviour of the engine from impact to an exit point (which is a totally different angle to where it landed) was not a great decision imo.

This angle is less however than the trajectory you chose based on a video image. You speculated it was a staircase structure that accounted for this and you predict more deflection, correct?
 
Last edited:
This angle is less however than the trajectory you chose
Calculated, not chose.

based on a video image
No. Several video's. Many images.

You speculated it was a staircase structure that accounted for this and you predict more deflection, correct?
I've suggested it could have been involved, but who knows.

I'd suggest a series of deflections resulting in somewhat of a curved trajectory through the building, emerging something like 20 degrees clockwise of the tower orientation.
 
femr2:

I believe the actual numbers are somewhere between your's and NISTs'. I also wonder where your numbers would be if you didn't start out with the objective of proving NIST wrong. Please don't say you didn't do this, we can all read your posts from a few years back.

I don't really care what you believe today but, have you considered a compromise using the flight path as a moderating factor in your trajectory? (this would be yaw)
 
Last edited:
Why ? The passage you quoted has little to do with the the text you are responding to.

Little to do with whether the plane changed direction upon impact, but relevant to the broader discussion of orientation and trajectory.

There are a range of orientation/trajectory combinations which allow a match.

The NIST wingtip penetration points are a bit off, especially the left'un.

Their orientation is off.

Their trajectory is off.

So you've averred.

Yet the angle between the NE corner and where it landed is 20 degrees...

So, if you trace one line indicating your belief about the trajectory of the engine on the way into the building, and another indicating the apparent trajectory on the way out, do any hypotheses present themselves? (You may have done this somewhere in the last 80 pages.)

And the simulated engine never even made it out the building.

Highly accurate then... :rolleyes:

NIST makes a plausible case about why making it out of the building is possible within their error bounds. I'm not sure how troubled to be about where the engine landed. At any rate, since they use the apparent exit point of the engine to help them choose between 13 degrees and 17 degrees, whereas I believe you're arguing for 6 degrees, this is not so much the point, is it?

ETA: I.e., NIST settles on 13 degrees trajectory, 10 degrees orientation. I'm actually not sure whether you are arguing for 6 and 6.
 
Last edited:
NIST makes a plausible case about why making it out of the building is possible within their error bounds. I'm not sure how troubled to be about where the engine landed. At any rate, since they use the apparent exit point of the engine to help them choose between 13 degrees and 17 degrees, whereas I believe you're arguing for 6 degrees, this is not so much the point, is it?

This goes along the lines of "femr2 proves NIST wrong" more than anything else. He does not allow for a degree of error, instead uses ~.

If only it wasn't about being right and more about the data. :D
 
So, if you trace one line indicating your belief about the trajectory of the engine on the way into the building, and another indicating the apparent trajectory on the way out, do any hypotheses present themselves? (You may have done this somewhere in the last 80 pages.)
Sure. NIST was wrong to assume it wasn't deflected.

NIST makes a plausible case about why making it out of the building is possible within their error bounds.
Only had another 1400ft to travel afterwards...simulation should have indicated it pinging out of there at a rate of knots.

I'm not sure how troubled to be about where the engine landed.
Where it landed provides you with information to assess the NIST assumption that it was not deflected during its journey through the building, and their related decision to set trajectory of the aircraft based upon that, rather than upon checking the orientation at the moment of impact (their calculations beyond those decisions are based on earlier approach vector)

At any rate, since they use the apparent exit point of the engine to help them choose between 13 degrees and 17 degrees, whereas I believe you're arguing for 6 degrees, this is not so much the point, is it?
The point is that the engine was deflected. NIST assumed it wasn't. That 13 is closer to 20 than 6 is is certainly not "the point".
 
I believe the actual numbers are somewhere between your's and NISTs'
"Believe" what you please.

I also wonder where your numbers would be if you didn't start out with the objective of proving NIST wrong. Please don't say you didn't do this, we can all read your posts from a few years back.
The initial objective, as I've already said, was a response to claims that the Carmen Taylor and Hezarkhani footages were faked and from "impossible" viewpoints.

Such origins can be easily proven, regardless of what you "believe".

I don't really care what you believe today
Oh noes :cry1

but, have you considered a compromise using the flight path as a moderating factor in your trajectory? (this would be yaw)
:rolleyes: Been there, done that. RADES data stops short, and is of little use. Have looked in detail at behaviour during impact, and as I've said several times now...there's no detectable drift/slew during impact.

Why are you going round in circles ?

Prove difference between orientation and trajectory with real-world data and imagery and I'll accept it of course.
 

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom