Ed Applying luma curve to photos and footage

P.J. Denyer said:
Anyone else feeling a palpable sense of relief that we won't have to witness nt1 "analysing" the Air India footage?
Click to expand...

For clarity, I'm not making a joke or a dig here. I honestly would have found it stomach turning and while I'm sure he is, or shortly will be, posting his enlarged still frames with their glowing halos and telling people the film's been doctored, I'm glad it won't be in this little corner of the internet and I won't see it.
 
P.J. Denyer said:
For clarity, I'm not making a joke or a dig here. I honestly would have found it stomach turning and while I'm sure he is, or shortly will be, posting his enlarged still frames with their glowing halos and telling people the film's been doctored, I'm glad it won't be in this little corner of the internet and I won't see it.
Click to expand...
Tru dis. You have to be a special kind of paranoid to think some mysterious powers are faking reality, while at the same time disparaging the real people and their loved ones who died. Sandy Hook conspirators spring to mind.
 
By now, every halfway veteran here (and on plenty of other fora) can spot a crankadingdong after just a few posts. It's a capability that develops whether you like it or not.

Now, in my capacity as World's Leading Expert Around Here, I will venture a generalization about the belfry commandos we attract: They're all extreme egocentrics. Not narcissists, just self-engorged personalities to whom other people are objects, not really separate from the all-containing Me.

As long as poor souls like that can be the center of attention, they'll hang on. We see this in their IRL relationships, where there's at least some variety in the objects' input to the Me: sound, touch, maybe sex, a sandwich, etc. But, unfortunately for them, a forum is a monotonous environment, mostly just a whiteboard on which irritating irrelevancies can appear, i.e., opinions that don't coddle their egos. Bad business! No wonder they almost always fly off in rage and incoherent denial. (There, I ventured another generalization. Good thing I said "almost." I fear the fallacy of the absolute.)

So Empty One is worth at least a footnote in the meta-literature of the poor old forum.*

* This is it.
 
JayUtah said:
No. Your method involving adjusting the luminance channel creates artifacts in photos that you speculatively attribute to image tampering.


Professionals have examined your images and reviewed your methods. They do not produce the effects you claim, and you have decided to ignore the advice of professional.


Your methods are not valid. There is no obligation for government or any other person to respond to it beyond what has already been done.


Yes, you have done this repeatedly without listening to any responses to it. It proves nothing.


Gibberish.


Gibberish. No generative AI for images was available in 2001. Your screenshots are from convenience sources. They are inevitably enlargements from sources that end up being oversampled and thereby overly susceptible to your wrong methods.


Ego has nothing to do with it. You do not have a foundation of understanding suitable to the claims you are making. When people show you that this is the case, you complain that you are being ill-treated.


You are not better or smarter than your critics. You do not have any better information. Your critics' statements are not coming from a place of bewilderment or ideological entrenchment. Your critics simply know more about the subject than you do.


No one is interested in indulging your self-aggrandizing claims if you are unwilling to address well-informed criticism.
Click to expand...
All this Luma Curve nonsense is rather llike Jack Whit tweaking the
colors in Apollo's images to prove they are/were faked.
 
Andy_Ross said:
For the avoidance of doubt.
Apart from one being converted to monochrome and the beach shot having the contrast increased nothing has been done to the pictures.
One of them is a stock library shot the others are mine
Click to expand...
Seems the test proved his gullibility and his inability to answer a straight question.

If it had been a fair test, if three were tampered with, would he have passed? We probably wouldn't know because of his inability to answer a straight question.
 
Andy_Ross said:
Jay found a couple of haloes with just a quick look in pictures that haven't been tampered with.
Click to expand...
Before the Air India crash occupied my attention, I was able to run some DCT resolution comparisons using one of our more powerful computers. The monochrome snowy pond image got a mild hit for possible use of a smoothing or healing tool. And the non-mill seaside scene got a mild hit for possible use of a clone tool.

DCT stands for discrete cosine transform, and is one of the mathematical tools used in JPEG image compression. It turns out that compressibility is a good proxy for overall resolution. You treat the image as a set of 8x8 patches and investigate the DCT compressibility of each. Then you shift by one pixel and do it again, hence the supercomputer. The fidelity of a compressed patch to its source in a local region is normally distributed, so when you find an outlier patch, it's something that has either more or less resolution than the rest of the image. That is often an indication of whether certain kinds of common image manipulation tools like smoothing, dodging, burning, or healing have occurred.

The monochrome snowy lake has a slightly suspicious (z = 1.72) smooth patch.

Similarly the vector of DCT coefficients for an image is a sort of proxy for its contents, If you take every possible 8x8 patch and compare it to every other possible 8x8 patch (again, cough, supercomputer) you can often identify cloned image elements. Each set of DCT coefficients is a 64-element vector (i.e., a vector in 64-dimensional space). You can use normal vector arithmetic (i.e., dot product) to reckon vectors that lie in similar directions and therefore image elements that compress to the same DCT representation.

The seaside image has a slightly suspicious duplicate signpost (normalized dot product = 0.640).

This analysis directs the analyst to portions of the image to examine using other tools or with a well-trained Mark 1 eyeball.
 
I was still entertained by this whole thing. This was an angle I hadn't seen before. I mean, all the footage of 9/11 was fake. That's, well, different.

At the risk of telling, you what you all know. It would have been easier to hijack planes and fly them into the buildings than do whatever nt1 was suggesting happened.
 
I mean, something good came out of this thread. I might make this work of art my new profile picture at some point. GSG will get to stay for a while longer, though.

1750992950339.png
 
You must log in or register to reply here.

Back
Top Bottom