Merged Discussion of femr's video data analysis

You don't understand your own filter.
The Savitzky-Golay filter is not "mine". It has been around since around 1964 and is fully understood by many, including myself. It is clear than many posters within this forum do not understand it, despite it being described numerous times, and copious links to external resources being provided.

I would ignore the old dolt engineer
And ignore your inept and derogitory assertions ? No.

I was lucky to pass Maybeck's class.
Assuming you mean Bernard Maybeck, I see no relevance to a filter invented 7 years after his passing.

Your knowledge of filters cames from google U.
Incorrect, though obviously when you ask for links to help you understand I'm going to google "Savitzky Golay", innit.

Put me on ignore.
No. It is more appropriate to show the failings in what you spam here. A better alternative would be for you to ignore the thread.

It is best since your definition of technical papers is 911 truth paranoid conspiracy papers
The purpose of them being listed there has been explained to you numerous times, namely that each contain some detail on how to perform computational aspects of descent. Their parameter selection and conclusions are irrelevant. It is a point you raise regularly, which given it's on a web page created in 2009 I find quite funny, though very interesting in terms of your motivation.

It is ironic 911 truth nuts on line, on youtube, use your work to push lies about 911, they thank you.
News to me beachnut. I get mostly moronic crap from all "sides" in my experience :rolleyes:

Did not expect an answer in engineering terms.
Dear me...you asked...

"When will you define the high frequency noise you are removing using SG?"

I said...

SG methods result in some smoothing of noise, but it is not "removed". Are you sure you understand ?

All smoothing affects both noise and signal.

If you prefer in W.D.Clingers words... To get a usable picture of acceleration, you have to do some smoothing somewhere.


If you like I'll add...it's not just high frequency noise that gets smoothed either, some "low" frequency noise, some high frequency signal, some "low" frequency signal. Smoothing methods are fairly indescriminate, of course.

No goals, no conclusions, no engineering.
Yawn.

Using methods googled.
Yawn.

Not defending your thesis, unable to explain why.
Read the thread. Some of it is really funny "in hindsight". The incredible number of ludicrous errors by others...amazing. My own personal consistency and integrity...not shabby at all. Pretty much everything I've said in this thread still holds true, which most certainly can't be said for posts by numerous others who have participated.

Hide behind other statements because you can't defend using a filter which removes high frequency noise.
You make no sense. Smoothing is a must. Trend is focus, so smoothing of higher frequency noise preferred over smoothing lower frequency noise, absolutely. Really not interested in the higher frequencies at all, neither noise OR signal.

The noise is the data
Silly thing to say. Also, there's not really any practical way to separate noise from the true signal for the data under discussion. Of course, via processes such as static motion removal the inclusion of noise sources is minimised, unlike the similar data from NIST.

you don't want to remove it
Smoothing is a must. Without appropriate smoothing you end up in tfk's shoes making inept assertions that the data indicates +/- 60ft/s velocity fluctuations, even during periods of no actual building motion. Not wise.

you want it to remain raw
As raw as possible, absolutely. Which is why I utilise the Savitzky-Golay filter, enabling me to retain maximum noise and still expose the underlying trend behaviour.

I don't understand your made up methods
That is not my problem, as I've described "my methods" is detail numerous times. Other people seem to have no problem understanding, so I can only assume the issue is with you. It's really not complicated.

and your inability to defend your work with more than google references used without reason.
Word salad. Untrue. Read the thread again.

No big deal
Certainly seems to be of issue to you beachnut. You've been writing posts like this for years, with very little change in the content. Habitual ?

not sure what engineering school you went to
I've told you numerous times, my personal details are irrelevant, and, as you say, unknown by you (and everyone else who seems interested).

but at the my schools I had to explain in detail and present the reasons with real references. It is your work, defend making data of falling building look like music, low frequency music. It looks better than the data, but it is not the data, and study of the collapse serves no purpose, not stated goals related to 911 truth crazy claims.
Read the thread.

A stop watch and the many videos is enough to destroy the faster than free-fall nonsense.
I have performed a rather more rigourous full descent timing recently, and provided detail for anyone to replicate it.

The purposes of tracing are not limited to your singular focus.

As for tracking objects visually with optical equipment
Say what ? :rolleyes:

you are not and have not developed new techniques
Correct, though within the limited arena I've refined the processes greatly.

and the filters you applied are not better than work already done
Who are you referring to beyond NIST, Chandler, ...?

and being done at grad schools in engineering and optics workshops
Say what ? :rolleyes:

like the engineers I worked with at AFWAL (yes I am old).
Yawn.

Put me on ignore
Again, no.

I don't want to study your work
Ignore the thread then.
 
The purposes of tracing are not limited to your singular focus.

It might help if people actually knew this purpose. If it could be explain how fractions of a second in any direction are important (and to what end), maybe you would encounter less resistance (not just here). ;)
 
It might help if people actually knew this purpose. If it could be explain how fractions of a second in any direction are important (and to what end), maybe you would encounter less resistance (not just here). ;)
I've gone into more detail on that question so many times within this thread it's ridiculous. I'm not stepping into the hamster-wheel. I suggest you (re)read the thread.
 
I've gone into more detail on that question so many times within this thread it's ridiculous. I'm not stepping into the hamster-wheel. I suggest you (re)read the thread.
I have read this thread many times. The point is, how does the world and building safety actually benefit from this? That was the purpose of this investigation after all.

Unless it's shown that these fine details actually effect the main thrust of the findings they are just that, fine details.

I have no problem with your work. Outside of just a hobby though, I see little actual utility for this level of detail.
 
I have read this thread many times. The point is, how does the world and building safety actually benefit from this? That was the purpose of this investigation after all.

Unless it's shown that these fine details actually effect the main thrust of the findings they are just that, fine details.

I have no problem with your work. Outside of just a hobby though, I see little actual utility for this level of detail.

Many have asked none have been answered.

At this point his analysis seems less useful than working crossword puzzles.
 
I have read this thread many times.
You have my sympathies. Would have been so much better without the noise (no pun intended).

The point is, how does the world and building safety actually benefit from this?
Not my area of focus I'm afraid.

That was the purpose of this investigation after all.
No, that was part of NIST's purpose.

Unless it's shown that these fine details actually effect the main thrust of the findings they are just that, fine details.
You are still focussing upon NIST. "femr2's video data analysis" has provided details which confirm elements of the NIST report (which people here don't seem to like), refute elements of the NIST report (which people here don't seem to like), refute elements of Chandlers "work" (which people here don't seem to like), refute elements of Szamboti's "work" (which people here don't seem to like), provide emprirical independant data which refutes theories focussing upon "instantaneous removals" (which people here don't seem to like), identify gross errors in NIST WTC2 simulations, identify probable error in NIST WTC1/2 initiation sequencing (which people here don't seem to like), suggest error in NIST WTC7 initiation sequencing (which people here don't seem to like), ...etc.

There are areas of "9/11" and "WTC" which can be looked at in terms of "video data analysis", which I do, regardless of which "camp" it bothers.

I'm not overly concerned if no-one else thinks issues identified are important to them. If there's an error which is possible to highlight, I'll highlight it.

I have no problem with your work. Outside of just a hobby though, I see little actual utility for this level of detail.
You've said so many times in the past. Why repeat yourself ?
 
Many have asked none have been answered.
Incorrect. Really can't be bothered to wade through this single thread highlighting every instance, but if your memory is so lacking it may become necessary. Or you could alternatively just accept that you are mistaken due to lack of interest/recollection.
 
At this point his analysis seems less useful than working crossword puzzles.
I don't think I'd go that far. If you just were interested in historical accuracy of the collapse, then his work is in-fact useful. I seriously doubt however that it could ever lead to a more comprehensive understanding then we already have (nuts and bolts wise).
 
At this point his analysis seems less useful than working crossword puzzles.

Recently, and from my perspective, his work in this subforum has been among the most useful. Far more useful than yours.

Just my opinion.
 
I've gone into more detail on that question so many times within this thread it's ridiculous...
Correct. And to some of us - including me and I suggest Oystein - your purpose is clear.
...I'm not stepping into the hamster-wheel. I suggest you (re)read the thread.
...probably wise. BUT I already stepped into the hamster-wheel** so I will continue to see if we can make progress. And I disagree with the suggestion that he or they "...(re)read the thread." Being confused at this stage is unlikely to alter if people re-read the read with the same perceptual framework. With respect several of us here need to back off and re-think where we are coming from.

(** Actually not a good analogy. I am about to kick over the hamster wheel. ;))

Let me start by using one of DGM's posts - I respect DGM, I think I understand where he is coming from PLUS his post goes pretty close to the issues. I may go back to one of tsig's posts a bit later - it involves some more abstract issues.

So this one:
I have read this thread many times. The point is, how does the world and building safety actually benefit from this? That was the purpose of this investigation after all.

Unless it's shown that these fine details actually effect the main thrust of the findings they are just that, fine details.

I have no problem with your work. Outside of just a hobby though, I see little actual utility for this level of detail.
Let's take it bit by bit:
I have read this thread many times...
With respect I suggest that you have re-read it with the same frame of reference - see my previous comment.
...The point is, how does the world and building safety actually benefit from this?...
Straight off the rails in the second sentence. No that is not the point of either a) this thread OR b) femr's research and his research tools. This thread is "Discussion of femr's video analysis". It is not "Was there CD at WTC?" It is not "Let's write better building codes". It is not whether or not other members "see the utility of femr's work".

It is about discussing his analysis techniques. And that can include questions of the accuracy of the tools he uses - it does - and that is probably the core focus of legitimate discussion here.
That was the purpose of this investigation after all....
We need to sort out the ambiguity. What investigation are we talking about? I suggest DGM is referring to the investigation into building collapses. That is not the topic of this thread - there are dozens of other threads to discuss that topic. The "investigation" relevant here is "femr's video data analysis".

I'll risk an analogy:
In the overall context of building construction methods a traditional method of joining materials is to use nails which have been put in place using hammers. If we choose to discuss "putting nails in" we can discuss the methods of inserting nails and whether hammers generically are appropriate tools. We could even discuss a specific form or brand of hammer. BUT those discussions do not depend on whether or not a specific building fell down. In more technical jargon the topic hammers and their use in inserting nails is a closed system subset of building techniques.

(OK From experience I should expect nit picking attacks on the analogy. I warn that I will not join in derail discussions. The point of "isolated sub system" is what I am illustrating. People will either see it or not.)

Let's continue:
...Unless it's shown that these fine details actually effect the main thrust of the findings they are just that, fine details.
You just switched from "hammers and nailing" which is the topic to "building collapse" which is not the topic of this discussion and of femr's use of tools and techniques.

So a final apology to DGM for using his post as the launch pad. It was the clearest for specific comments.

I also commend the post by tsig where he raises two valid points. I may address them in a later post.

And I note that Beachnut does not share my sense of humour. :rolleyes: Ventis secundis, tene cursum.
 
Last edited:
Correct. And to some of us - including me and I suggest Oystein - your purpose is clear. ...probably wise.
...

With all due respect: No, isn't clear to me.

But at this point I don't care.
Femr2's work is limited in scope, perhaps limited in purpose, but accurate in its results, and accurately communicated. Thus, it is either useful for my purposes (which may vary from thread to thread), or it isn't.

In this thread, which is probably about his generic video analysis tools (and discusses some examples of data output and smoothing results), I have learned about his tools, and about some criticism thereof. I understand that no one really has much of an issue with the accuracy of his method to gather raw (sub)pixel data, and take that as an indication that this raw (sub)pixel data is probably pretty reliable, and probably the most accurate we can find anywhere. Which is a great result.

Lots has been said about smoothing, and I think I have understood some of what is said on either side, and still don't understand other things. What emerges in my mind is the realisation that no smoothing will result in useful results when it is not guided by specific questions and theories about the underlying process that gave rise to the raw data (which is the actual motion of building features PLUS several sorts of error and noise). "Useful" as in "leads us to an understanding of what happened". Throwing raw data into a S-G machine does not lead to this sort of understanding. I understand though that it may be a good intermediate step that helps to visualize the trends and guide the eye. But analysis should stop there.

I haven't really seen anybody trying to take analysis further.



Do you guys remember the movie classic "Blow-Up" by Michelangelo Antonioni?
 
Femr2's work is limited in scope, perhaps limited in purpose, but accurate in its results, and accurately communicated. Thus, it is either useful for my purposes (which may vary from thread to thread), or it isn't.
Which is fine by me :)

In this thread, which is probably about his generic video analysis tools (and discusses some examples of data output and smoothing results), I have learned about his tools, and about some criticism thereof. I understand that no one really has much of an issue with the accuracy of his method to gather raw (sub)pixel data, and take that as an indication that this raw (sub)pixel data is probably pretty reliable, and probably the most accurate we can find anywhere. Which is a great result.
<tips hat>

Lots has been said about smoothing
I had hoped the discussions would have been much more useful and explorative. Only a couple of smoothing methods have been discussed, most of which is ping-pong about trivial aspects rather than deeper and more useful "analysis" of smoothing methods...which I DON'T already know about (or haven't thought to apply).

Certainly more focussed discussion about various noise reduction methods would have been useful.

The potential for such discussion is drowned out by ... stuff.

What emerges in my mind is the realisation that no smoothing will result in useful results when it is not guided by specific questions and theories about the underlying process that gave rise to the raw data (which is the actual motion of building features PLUS several sorts of error and noise)..
My view is that focussing upon trend (rather than, say, revealing "mini jolts) is the most useful way to guide assessment of the data. Certainly from the Dan Rather viewpoint, there are very real limitations on fidelity that can be extracted two derivatives later. The cam#3 data has been kind of shunted to one side, even though trace fidelity is up to 10 times as high as that from the Dat Rather viewpoint. (which is somewhat of a shame)

Throwing raw data into a S-G machine does not lead to this sort of understanding. I understand though that it may be a good intermediate step that helps to visualize the trends and guide the eye. But analysis should stop there.
There has been little suggested beyond interpretation of that trend. I've been pretty rigid in bounding the usefulness to trend interpretation of derived metrics. For displacement, given the fidelity is effectively greater, and early motion behaviour can be determined with good confidence, additional interpretations can be validy formed I rekn.

I think it important to highlight that focus is not simply upon the velocity and acceleration views. More useful, to me at least, is the insight into building behaviour gleaned from the displacement data...especially when multiple locations are compared...an area that hasn't really touched this (or any other) thread (even though the data exists).
 
In my previous post I said "tsig...raises two valid points. I may address them in a later post".

So here goes. Thanks to tsig for raising two of his three points.
Don't see how this applies to ozeco. Or femr2....
It doesn't but two of the points tsig raised are validly identified issues - provided we make valid use of them. These are the three points:
Heuristic (play /hjʉˈrɪstɨk/; or /hyoo-ris-tik/; Greek: "Εὑρίσκω", "find" or "discover") refers to experience-based techniques for problem solving, learning, and discovery. Where the exhaustive search is impractical, heuristic methods are used to speed up the process of finding a satisfactory solution; mental short cuts to ease the cognitive load of making a decision. Examples of this method include using a rule of thumb, an educated guess, an intuitive judgment, or common sense.

Most engineers prefer facts and figures and maths and graphs to using a rule of thumb, an educated guess, an intuitive judgment, or common sense.

The common truther mantra "It doesn't look right" is based on 'common sense' so you're welcome to crawl into that bed if you want to but don't be surprised if you get a few fleas.
OK Here goes:

Point #1 - about heuristic objectives. For purposes of this discussion we need to distinguish two types of objectives:
1) The first one - the one preferred by the military - has a definite fixed end point plus time to achieve. So an army version could be "Capture hill 208 by dawn on 25 December 2012." I think we all comprehend that form and can translate it into many situations.
2) The second one is the heuristic or "discoverable" objective. Put simply it can have the form "Let's head down this way and see what we find". Pure research goals are often in that form.

(And there are many shades or blends between the two - applied research goals often directed towards broadly defined end point targets are one example - no need to discuss those at this stage.)

And we don't need to know much more about it for this discussion. So we can overlook the technical definition of "heuristic" which tsig posted except to note that it is very broad. And I presume that most members were already familiar with the word :D

The main point is that femr's goals are heuristic and that he has stated them many times. Most members attacking femr on the false claim of "no goals - no objectives" are actually saying "I don't recognise your goals because they are not within my limited scope of understanding of goals and objectives". And I don't think even femr has expressed it in those terms -- but that is a derail we can avoid for now.

So how about folks either stop claiming "No goal". Either accept that "let's explore this and see where it leads and I don't care how long it takes" is a valid goal OR support their counter claim with reasoned argument that the goal is not valid?

Point #2 - about misuse of "engineer" or "engineering".
The point which tsig sort of makes is:
"Most engineers prefer facts and figures and maths and graphs to using a rule of thumb, an educated guess, an intuitive judgment, or common sense."
This statement looks like a truism but it actually relies on innuendo and says nothing about the topic we are supposed to be discussing. So, at this stage, I will limit myself to two comments.

1) The comment about "most" engineers refers to a strawman in the shape of a false dichotomy. The core claim is that one sector of engineers view some things differently to the remainder of engineers. I happen to agree with tsig on this point as it is stated - but I'll wager my interpretation is diametrically opposite to tsig's. Let me illustrate from my experience as an engineering manager.

There is a large subset of engineers who "prefer [technical] facts and figures and maths and graphs". Whether they qualify as "most" I cannot say for reasons I could explore in an appropriate thread topic - not here. The application of engineering skills by these engineers tends to be limited to routine rote application of standard methods. The true comparator is not, as alleged by tsig, "using a rule of thumb, an educated guess, an intuitive judgment, or common sense." It is their inability to get out of their rut of limited experience. And that is precisely the sort of situation we see on these threads where blind spots and skill limitations are clear to see for those who can see them. And I mean that for the known engineers - wait for the following comment please ;).

2) The second comment is "This topic is not about engineering anyway so why try the derail/diversion?" One of the common arguments raised and falsely applied is "You cannot say that because you are not an engineer."

The underlying issue is "Is the claim correct" NOT "Is the person an engineer". I repeat something I have said many times. I will back the none engineer who makes a correct claim against the engineer who makes wrong claim. Having multiple engineering qualifications does not make false claims true and not having engineering qualifications does not make true statements false.

And Oystein has already made the main point on this aspect of the discussion:
...Femr2 uses precisely facts and figures and maths and graphs, and receives an awful lot of flag - often from debunkers who use a rule of thumb, an educated guess, an intuitive judgment, or common sense.
Would you disagree with that assessment?
Thank you Oystein - (Who many would claim cannot say those things because he is not an engineer. :D ) Bottom line is that femr does not make those errors whilst his opponents do. So why try to attribute to him the flaws of the other side? Some of us will spot the transposition every time. :rolleyes:

So thanks to tsig for identifying a couple of relevant issues. However: :(

Point #3 - The "Common Truther Mantra".
The claim is "[t]he common truther mantra "It doesn't look right" is based on 'common sense'....."

The claim fails because this discussion is not about truthers and I did not mention truthers in the quote you are referencing. The latter part of your point is not clear as to what you mean nor demonstrably linked to a valid introduction into this debate. So no comment is needed.
 
With all due respect: No, isn't clear to me....
No problem - I was "inviting" a comment.
...Femr2's work is limited in scope, perhaps limited in purpose, but accurate in its results, and accurately communicated. Thus, it is either useful for my purposes (which may vary from thread to thread), or it isn't....
You seem clear enough about where it relates to what you are interested in. And how you can use it when it fits within the scope of what you are discussing in a thread. Main thing therefore is that you control your scope of discussion and incorporate what femr does if it fits. That is diametrically opposite to one of my key points of criticism. Others are trying to make femr fit their context or "world view" rather than they use femr if his work fits and ignore it if it doesn't. And some going the step further and in effect saying "femr is wrong because his work does not fit my world view". And stating it that bluntly I could be inviting denials. :)

Still we can keep parsing without much benefit so I'll leave it there.
thumbup.gif
 
And some going the step further and in effect saying "femr is wrong because his work does not fit my world view"

Haven't seen that so much. Questions do ring, as to what it all means. Like what outcomes does it change? Fence walkers have his problem...
So do wood spoons!
 
Haven't seen that so much. Questions do ring, as to what it all means. Like what outcomes does it change? Fence walkers have his problem...
So do wood spoons!

Like I said, this depends on YOUR objectives. You can either use the heuristic results that femr2 generates, or leave them be. If they don't fit your purpose, don't use them. No need to go further and burn them.
 
This thread is "Discussion of femr's video analysis". It is not "Was there CD at WTC?" It is not "Let's write better building codes". It is not whether or not other members "see the utility of femr's work".

It is about discussing his analysis techniques. And that can include questions of the accuracy of the tools he uses - it does - and that is probably the core focus of legitimate discussion here.

Maybe it would help if this thread was not in this subforum as it has no conection to any conspiracy theory. Would you support it being moved to science?
So a final apology to DGM for using his post as the launch pad. It was the clearest for specific comments.
No apology necessary. ;)
 
Maybe it would help if this thread was not in this subforum as it has no conection to any conspiracy theory.
It has lots of connection to 9/11. The method is applied to data from 9/11 and has the potential of evaluating claims made by CTists vs. claims made by realists.

Would you support it being moved to science?
...
I am totally against that.

What I support is every member in this subforum staying mindful of the topic and scope of each thread, and arguing the arguments, not the arguer (as intentions and purposes of the arguer, while perhaps interesting, are usually not the topic). In other words, acting in accordance with their MA. If that simple rule is followed, everything can be discussed nicely and productively here, in this subforum, to which the results are relevant.
 
Maybe it would help if this thread was not in this subforum as it has no conection to any conspiracy theory. Would you support it being moved to science?
No - That concept is also not appropriate. The historic purpose of this sub forum was to accommodate all 9/11 related matters - put simply so that the woo does not contaminate the rest of the forum. If the criterion of "it is pure science" had been applied from the outset the flavour of the sub-forum would have been a lot different. Remember the criterion would have to apply to threads such as the current dust analysis threads. And historically to all the pure engineering analyses of the WTC Towers collapses.

Its not that simple anyway - what is meant by "no connection to any conspiracy theory"? The pure engineering discussions of how the towers collapsed are equally "not connected to conspiracy theory" - we constantly rubbish the AE911 strategy of "It was CD therefore inside job" which links a purely technical fact - CD or not - with a political claim - inside job. There is no doubt that "inside job" is a conspiracy theory but the technical claims would have to be sorted into those which said "the towers fell by this mechanism..." which did not explicitly mention conspiracy and therefore went to "science" from those that said "the towers fell by this mechanism which was started by US personnel..." which could stay here.

I don't believe that sort of division could be consistently applied. And it is easy to see how it could quickly be circumvented. It would be a far too subtle distinction for members and staff alike.

And, again, if that had been the original intent the forum would have been a very different place. The nice peaceful residents of the science forums would have had to tolerate a lot of 9/11 woo. :)

BTW the same concept was given a thorough work over about a year or so back so I won't rehash the old debate any further.

No apology necessary. ;)
Thank you.
 
It seems rather futile to have such a long thread about TV footage which was fake anyway.

After years of research and connecting the dots under the most difficult conditions (youtube adverts, disinfo agents masquerading as truthseekers, low connection speeds), we now know that the towers were brought down by nothing else than a precise controlled demolition.
Demolitions require loud explosives, however in the "collapse" videos aired by the government, no such detonations are heard. This proves TV fakery beyond any reasonable doubt.

If the sound was faked, then the picture was faked, too.

/close this thread
 
Last edited:

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