Pentagon video analysis

[Somerandomguy]

I emailed the guy and asked him about whether he thought the sheer speed of the plane could not produce the incompability of Frame 20. Here's the reply, again, in poor english: (I wonder if he uses babelfish for them)

The /object /speed it's a false problem because the two CCTV
were far from filmed object (210/250 meters - 700/820ft) and the
relative representation of speed (in pixel/frame) of the plane it's
measurable in 6/8 pixel for frame. In every second of a NTSC video
system there are 30 frame (subdivided in 60 fields) and the Time Lapse
Recording save only one frame (on only field for frame) per second. So
the plane in frame 20 frame introduce a position change of only 6/8
pixel but the real 757-200 have a dimension in pixel of 37px. It's not
possible to cancel a 37pixel object in 6/8pixel movement. The shutter
speed of cctv cameras are very efficient and their speed come from
1/60sec to 1/500sec. If you look to the high speed debris in the air,
you can see that their representation in pixel is very clear. There is
not blurred trails. There is not blurred movements.

The 3D model of 757 that you have seen in my report are rendered with
real speed and simulated shooting cameras. Every virtual camera
introduce shutter speed and real frame rate so the things you see it's a
high detail reproduction of dynamics representation. No way. Frame 20
doesn't represent AA77.

It seems to me that he mainly answers whether the speed could have an effect on the visualisation of the object, not whether a slight dissynchronisation could cause the plane to be further along in one frame as than the other. However, if you look at frame 21 and 22 under the "Sync Comparison", surely there are objects moving at higher speeds than 20 mph in the debris?

I'll email him the follow-up.

 

[ktesibios]

The claim that two cameras in a security system will be synchronized is wrong.

Unlike studio video cameras, the cameras sold for video surveillance systems have no provision for genlocking, that is, syncing their horizontal and vertical scan to an external source. Look on the back of a professional video recorder, switcher, effects device or camera and you'll find a SYNC IN jack, to which you connect blackburst from your master sync generator. You won't find anything of the sort on a surveillance camera. They just run on their own internal timebase. Date/time window burns on a security cam recording are generated by the recorder, which syncs to the sync information in the incoming video signal.

Two such cameras that start up in sync will probably stay in step over the short term; They generally use crystal oscillators for timing and nowadays even garden-variety crystals have +- 100 ppm tolerances and good stability.

However, the time difference between when camera 1 starts scanning line 1 of field 1 of a new frame and when camera 2 does the same could be anywhere up to an entire frame interval (approximately 33 ms for the NTSC system) and will depend on the timing relationship when the cameras were powered up, the absolute accuracy of both of their timebases and the history of any timebase drifts in each since they were first powered up, all of which are random and unpredictable.

Any analysis which is based on the assumption of frame-accurate synchronization between surveillance cameras fails at the outset because that is not a valid assumption.

Unless, of course, evidence establishes that the security cameras at the Pentagon were in fact genlocked to a master sync source. Has anyone ever seen any evidence of that?

 

[Somerandomguy]

Brilliant post Ktesibios. No, I suppose there is no evidence of that. I guess that's what he's trying to prove, showing that all other frames are in perfect sync, but considering how fast the plane moves it's probably inadequate.

 

[Somerandomguy]

Here's his second reply:

EDIT: I was the one who brought up cable length (I'm clueless about these things)...

In multiplexed CCTV system all the camera are connected to the converter
A/D that switch from one camera to another in "real time". There is not
any kind of problem relative to cable length. Cable Length introduce
quality signal problem (if you use very long cable without signal
amplifiers) in their electrical component, not in transmission speed and
not in sync. Every camera shoot the same time at same speed and in same
standard. Every signal are converted by Time Lapse Recording into the
multiplexed system. The only "offest" sync possible in this kind of
systems it would be measurable in 1/60sec (one field) between camera
signals accessing trough converter.

In this time (1/60sec) the plane at 530mph introduce one movement in
pixel of 3/4px. So the maximum offset plane position into cam1 and cam2
it would be measurable in 3/4 pixel.

Can anyone confirm that the offset can only be 1/60 second? Sounds like he's just making an assumption. And would the movement only represent 3/4 of a pixel? Hmm...

 

[Larry Lovage]

530mph at 1/60th of a second is 13 ft. What is that, a cockpit's length or thereabouts?

 

[Anti-sophist]

That 1/60th figure seems to be completely fabricated, as best as I can tell. I am unfamiliar with the specifics of this CCTV system, but, for me, any claims that don't reference the datasheets seem to be unfounded, at best.

That 1/60th of a second equals 3/4 of a pixel claim seems even more suspect, to me. I'd like to see the math.

Furthermore, he has failed to prove empirically that the simultaneity of the cameras ensures equivalent frames for objects moving 500mph. His analysis only shows similarities for objects moving 1/25th of that speed. Therefore, he is left to rely on the theoretical differences which seems conspicuous in it's lack of technical documentation of the specs of the system.

 

[ktesibios]

That 1/60th figure seems to be completely fabricated, as best as I can tell. I am unfamiliar with the specifics of this CCTV system, but, for me, any claims that don't reference the datasheets seem to be unfounded, at best.

That 1/60th of a second equals 3/4 of a pixel claim seems even more suspect, to me. I'd like to see the math.

Furthermore, he has failed to prove empirically that the simultaneity of the cameras ensures equivalent frames for objects moving 500mph. His analysis only shows similarities for objects moving 1/25th of that speed. Therefore, he is left to rely on the theoretical differences which seems conspicuous in it's lack of technical documentation of the specs of the system.

An NTSC-standard camera will start scanning a new field 59.94 times each second. That's probably where he gets his 1/60th of a second.

It takes two fields to make a complete frame. Field 1 contains all the odd-numbered horizontal scan lines and field 2 the even-numbered lines. This is called "interlaced scan". Frame boundaries, which are what's important to most video recording equipment, occur 29.97 times per second.

I agree with Anti-sophist. Lacking detailed technical information about the system that made the original recordings and also about how the video files found on the 'net were derived from them, all this hair-splitting about timing between two different videos is pointless. So is zarking around with vectorscope displays. It's a fertile field for finding "anomalies", if you use the customary CTer definition of "anomaly" as "any physical evidence which violates my expectations", because it's impossible to have any informed expectations.

They're zarking compressed video files on teh Intarweb, for Zarquon's sake. Endlessly poking at them in the hope of finding some "anomaly" which will magically overthrow the massive preponderance of evidence about exactly what hit the Pentagon is simply an exercise in JAQing off. One might just as well sing "The Anomaly Song" for all the good it will do.

It was a praying mantis!
From Atlantis!
Well, it was a lot like a praying mantis from Atlantis
But, that's the thing
That's the anomaly
It is, because it isn't not, and that's the part that goes up.

D'ooooohhhhhhhh!

Fish works in the shipyard
Building submarines
Teaching all the workers there
What the "dialectic" means
But, that's the thing
That's the anomaly
It is, because it isn't not and that's the part that's traumatic

-edo, the Anomaly Song