Ed Applying luma curve to photos and footage

[grunion]

Thank you for your wonderful input. It is greatly appreciated. I understand what a strong subject it is. I agree it deserves the correct amount of understanding and consideration. Again. Thank you for your reply.

Which parts of the story that I laid out do you dispute? What conclusions could be drawn from your suppositions? Was the photographer part of a conspiracy to hide what really happened? Did he take pictures of the towers and then superimpose the falling man images? Were the dozens of eyewitnesses succumbing to some kind of mass delusion, or part of the conspiracy?Was the chain of custody over the photo broken prior to publication and altered somehow? So if we looked at the original photos taken that day they would be different?

 

[Axxman300]

Undeniable fact:

A luma curve adjustment does not delete or create pixels—it remaps existing luminance values, meaning the data being revealed was already present in the image.

That’s the base truth. claiming otherwise must demonstrate how tone mapping itself removes source image data—something not supported by how digital image processing actually works.

Its there. It cant be explained away.

You don't understand the technology, and you are not interested in understanding this technology. Jay and others have explained how it works, and how it doesn't show what you claim it shows. You have come here to claim 9-11 was some kind of staged incident, and you've chosen (poorly) some crappy photo app to prove your foolishness.

You've ignored Andy's photos, which would go a long way to proving your point if you can demonstrate with of his pictures have been altered using your app in the way you're claiming it works. The fact you continue to ignore Andy suggests that your app can't figure it out, and it blows your claims out of the water.

You then claim that those who investigated 9-11, all 15,000+ of them, should have spent more time on the internet instead of Ground Zero, and the Pentagon where the physical evidence was recovered. That's just dumb.

 

[Axxman300]

Which parts of the story that I laid out do you dispute? What conclusions could be drawn from your suppositions? Was the photographer part of a conspiracy to hide what really happened? Did he take pictures of the towers and then superimpose the falling man images? Were the dozens of eyewitnesses succumbing to some kind of mass delusion, or part of the conspiracy?Was the chain of custody over the photo broken prior to publication and altered somehow? So if we looked at the original photos taken that day they would be different?

Oh yeah, he says ALL the images are fake, and claims wrecked cars were staged in the streets. He's a piece of work.

 

[P.J. Denyer]

I explained how it does. Since you don't seem to understand, I'll explain it again.

If I have an original image in which one pixel has luminosity Y=0.45 and another pixel with luminosity Y=0.55, an inverted paraboloid luminance map centered on Y=0.50 will map those two different pixels to the same luminosity, say Y′=0.95. If two pixels have different luminosity in the source and the same luminosity in the adjustment, the adjustment has lost data.

Luminance maps such as those you've employed are useful only as the first step in edge detection. They are band-pass filters that selectively amplify certain specific luminosities, generally those statistically determined to be transitional pixels between light and dark regions. Your method is nothing more than crude edge detection.

Quoted as this is the data, not 'pixels' that is lost.

 

[JayUtah]

Undeniable fact:

A luma curve adjustment does not delete or create pixels—it remaps existing luminance values, meaning the data being revealed was already present in the image.

Asked and answered. No one is claiming pixels are created or deleted. However, your method is lossy because there is less luminosity information in pixels in the the adjusted image than there was in the original image. This was shown to you mathematically.

Its there. It cant be explained away.

Your method is lossy. That's a mathematical fact.

Applying a luma curve does not invent data or artifacts—it reveals the relative distribution of luminance already present in the image.

No. On the contrary it is lossy, The luminosity distribution in the image is effectively cut in half by your adjustment. More importantly, it creates the halo artifact by eliminating the visual difference between midtones that are a fixed distance in luminosity above or below your apex domain value. If your luminance channel map is represented by the function L(y) with apex at domain value a, there exists a number b and a single value y´ such that L(a-b) = y´ = L(a+b). In fact, there exist an infinite number of the tuple (b, y´ ) for which that property holds. Your method is lossy as a matter of simply verified mathematical fact.

Further, your method was empirically demonstrated to create a false positive according to the understanding expressed above.

 

[P.J. Denyer]

You don't understand the technology, and you are not interested in understanding this technology. Jay and others have explained how it works, and how it doesn't show what you claim it shows. You have come here to claim 9-11 was some kind of staged incident, and you've chosen (poorly) some crappy photo app to prove your foolishness.

You've ignored Andy's photos, which would go a long way to proving your point if you can demonstrate with of his pictures have been altered using your app in the way you're claiming it works. The fact you continue to ignore Andy suggests that your app can't figure it out, and it blows your claims out of the water.

You then claim that those who investigated 9-11, all 15,000+ of them, should have spent more time on the internet instead of Ground Zero, and the Pentagon where the physical evidence was recovered. That's just dumb.

QFT

 

[ahhell]

Its all disingenuous, he claims his evidence is easy to find so he doesn't need to provide evidence. He claims his manipulations of 20 year old video shows manipulation and ignores anyone who provides explanations for why it doesn't. Then says we are the ones not acting in good faith. Nope, don't buy it. Just trolling at this point, which is probably a violation of the rules but I'm pretty sure I'm right about that.

 

[Andy_Ross]

Are you going to show us the results of applying your methods to the six example pictures I posted in the thread?

 

[P.J. Denyer]

Are you going to show us the results of applying your methods to the six example pictures I posted in the thread?

 

[JayUtah]

Are you going to show us the results of applying your methods to the six example pictures I posted in the thread?

I would like very much to see how @nt1's method performs on this task. A properly blinded test can be pretty convincing. Right now we're left with a tightly-spinning circular argument.

 

[P.J. Denyer]

I would like very much to see how @nt1's method performs on this task. A properly blinded test can be pretty convincing. Right now we're left with a tightly-spinning circular argument.

I'd sooner see him try to refute your maths, but that's not likely to happen.

 

[nt1]

That is a screenshot of a video camera filming the paper. Its in there and never coming out.

Its embedded into that photo. What that means? We wont know. but it has been touched.

Im moving on. Next photo.

 

[Andy_Ross]

No.

Move on to the 6 pictures posted in the thread. Like you agreed.

 

[nt1]

That is a computer following the collapse. That is a masking line.

It just is not a natural artifact. We can spend as much time as needed on this

 

[JayUtah]

That is a screenshot of a video camera filming the paper. Its in there and never coming out.

This does not help your case. You are interposing two new processes: newspaper halftoning and video color gamut reduction. Put differently, you're now insinuating that the residual effect of digital image compositing that you insist (against all mathematical certainty) that you can detect simply by egregious maladjustment of the luminance channel is now also preserved through these processes.

Newspaper halftoning resamples the color space, rendering it into CYMK, where the K channel at best approximates the luminosity derived from an RGB color space. It also resamples the raster into one specifically jittered to avoid coincidence. There is essentially zero chance that anything you alluded to previously in terms of digital image composition will survive that process.

NTSC broadcast television systems in 2001 used the YIQ color model, similar to the YUV model I alluded to for PAL. This is a much smaller color gamut that RGB or sRGB that a composed image would have had, and is additionally being applied to the halftones image. A smaller gamut means the video camera is unable to reproduce all the color information from a CYMK image—it is lossy. There is little reason to believe that the subtle luminosity effects you wrongly claim to be able to amplify using your maladjustment will persist through these compounded effects.

Its embedded into that photo. What that means? We wont know. but it has been touched.

No, this is entirely circular reasoning. Given the new impediments I have outlined to the process you insinuate is being applied here, you are simply unwilling to consider that the halo is produced by the proximal effect of your misguided adjustment whose mathematical effects we have already demonstrated. You seem unaware of how the various imaging processes you allude to affect the image. Now you're paradoxically suggesting that evidence against your claim is somehow even stronger evidence for it.

Im moving on. Next photo.

Moving on without addressing the criticisms already on the table is dishonest. Spamming the thread with more examples of your circular reasoning will not square the circle.

That is a computer following the collapse.

Gibberish.

That is a masking line.

No, it is the same halo effect as in all the other photos, only this time applied to the stairstep aliasing created by undersampling.

It just is not a natural artifact.

Correct, it's an artificial product of aliasing, combined with and amplified by the artificial effects of your lossy and ill-formed luminance map.

We can spend as much time as needed on this

Repeating your wrong claims ad nauseam is not productive. I'm spending far more time on this than you are, and it's exceptionally rude of you to ignore feedback.

 

[steenkh]

@nt1: When are you trying out your method on Andy_Ross’ six images?

The impression you give is that you don’t have much confidence in your method, and avoid having it tested. Instead you keep piling up your own selected pictures in a move that you think should impress us.

It doesn’t work.

 

[grunion]

OK, I’m done. It’s either significant delusional behavior or blatant trollery.

 

[Thermal]

View attachment 61632
That is a computer following the collapse. That is a masking line.

The video editor itself says its not, and their program doesn't identify "masking lines" with Luna curves.

It just is not a natural artifact. We can spend as much time as needed on this

It's the same ◊◊◊◊ you posted days ago. We've already spent too much time on this.

 

[P.J. Denyer]

OK, I’m done. It’s either significant delusional behavior or blatant trollery.

TBH I think a lot of trolling comes from people who are delusional.

 

[Thermal]

Asked and answered. No one is claiming pixels are created or deleted. However, your method is lossy because there is less luminosity information in pixels in the the adjusted image than there was in the original image. This was shown to you mathematically.


Your method is lossy. That's a mathematical fact.


No. On the contrary it is lossy, The luminosity distribution in the image is effectively cut in half by your adjustment. More importantly, it creates the halo artifact by eliminating the visual difference between midtones that are a fixed distance in luminosity above or below your apex domain value. If your luminance channel map is represented by the function L(y) with apex at domain value a, there exists a number b and a single value y´ such that L(a-b) = y´ = L(a+b). In fact, there exist an infinite number of the tuple (b, y´ ) for which that property holds. Your method is lossy as a matter of simply verified mathematical fact.

Further, your method was empirically demonstrated to create a false positive according to the understanding expressed above.

I kept reading "lossy" as "lousy". Worked either way.