More Fake Pictures of Oswald

[Jerry Freeman]

Show us that you know anything about light & shade. And that you even know what I'm talking about.

How can someone know what Chapman is 'talking about' when he doesn't even know himself?

[John Mytton]

OK ...apparently, Oswald must have discovered Rogaine before Minoxidil was developed years later
Take a look again at the various hairlines and the cosmetic looking eyebrows.
Mo-jo motion ain't working.

What the heck are you looking at, Oswald's dramatically receding hairline at the temple and resulting combover is clear in the arrest photo.





JohnM

[Jack Trojan]

Matt, this procedure is based on the same principle as the iphone facial recognition software and is virtually infallible.

JohnM

  This is absolute BS and confirms that you know squat about image analysis or facial recognition. Else pls explain how a "fade" effect is based on the same principle as facial recognition. This should be good.

Hint: Here's a fade algorithm with 14 lines of code.

    function getfadedcolor(Red1, Red2, Green1, Green2, Blue1, Blue2, fs, fn) {
        fa1 = (Red2 - Red1) / fs;
        fa2 = (Green2 - Green1) / fs;
        fa3 = (Blue2 - Blue1) / fs;
        cx1 = Red1;
        cx2 = Green1;
        cx3 = Blue1;
        for (var n = 1; n <= fn; n++) {
           cx1 += fa1;
           cx2 += fa2;
           cx3 += fa3;
        }
        return RGB(cx1, cx2, cx3);
    }

I used this simple algorithm to make the following animated gif (what gives with his body BTW?)



All fading/morphing algorithms are based on this simple methodology of interleaving images.

How many lines of code do you think are required for infallible facial recognition technology?

[John Mytton]

Hint: Here's a fade algorithm with 14 lines of code.

    function getfadedcolor(Red1, Red2, Green1, Green2, Blue1, Blue2, fs, fn) {
        fa1 = (Red2 - Red1) / fs;
        fa2 = (Green2 - Green1) / fs;
        fa3 = (Blue2 - Blue1) / fs;
        cx1 = Red1;
        cx2 = Green1;
        cx3 = Blue1;
        for (var n = 1; n <= fn; n++) {
           cx1 += fa1;
           cx2 += fa2;
           cx3 += fa3;
        }
        return RGB(cx1, cx2, cx3);
    }

I used this simple algorithm to make the following animated gif (what gives with his body BTW?)

I used this simple algorithm to make the following animated gif

 

The code you presented is a simple fade to white, how do you think that is relevant?
And please explain how your code applies to your gif which is just swapping one set of RGB values for another?



JohnM

[Jack Trojan]

|   133-A   |   133-B   |   133-C   |

Oswald's left shoulder appears to be further back in space relative to the chest-midline in the 133-B picture than in the 122-A picture. This accounts for the difference in the shoulder width.

I also believe the morph...



... has the 133-B head a bit smaller than the 133-A head, which would in turn make the left shoulder more narrow to the camera. Not saying the animation sizing was intentional, but there's an amount of subjectivity that could induce error. Since the shadow-cast on the facial features in the two photos vary (due to head tilt), the width of the head might be a more accurate thing to align.

Nope. The head is distorting but from ear to ear stays consistent. It might be undersized by no more than 1-2%, which is far less than the body is oversized. Here is the proof. I scaled the same photos relative to Oswald's leg length instead and applied the same algorithm.



There is only 1 conclusion I come to, and that is CE 133-A was shot with a different camera than the rest. Otherwise, the spherical aberration of the lens and quality of this photo clearly does not match the others. Oswald immediately claimed it was a fake, which probably meant a darkroom creation where a print of CE 133-A was photographed with Oswald's Imperial Reflex camera. This camera was not originally found with Oswald's possessions by the DPD. And this is not to be confused with Oswald's Minox spy camera or his superior 35mm Russian Cuera 2 camera, which was probably the camera that took the original CE 133-A.

This along with all the other shenanigans that the DPD where up to in the darkroom, including re-enactments and a curious cutout of CE 133-C, which was never submitted into evidence nor found with Oswald's possessions, makes the BYPs highly suspect. The fact that the DPD only leaked CE 133-A, which was the money shot where you could actually read the name of the commie lit, means they were an integral part of the conspiracy. They sheep-dipped Oswald to be the patsy with the BYPs, planted the gun in the TSBD, arrested him in record time, interrogated him in private and took no notes, then led him to his assassin who did the killing for them to gave them plausible deniability. This is why they didn't kill him in the theater. I could go on and on but you get the picture. 

[John Mytton]

|   133-A   |   133-B   |   133-C   |

Oswald's left shoulder appears to be further back in space relative to the chest-midline in the 133-B picture than in the 122-A picture. This accounts for the difference in the shoulder width.

I also believe the morph...



... has the 133-B head a bit smaller than the 133-A head, which would in turn make the left shoulder more narrow to the camera. Not saying the animation sizing was intentional, but there's an amount of subjectivity that could induce error. Since the shadow-cast on the facial features in the two photos vary (due to head tilt), the width of the head might be a more accurate thing to align.

Hi Jerry, yes of course you are right and the original head ratios are indeed different. The wide-angle effect makes Oswald's head in 133-A stretch vertically as compared to the more centered head in 133-B thus exposing Trojan's gif. Btw imo this alone makes any image manipulation virtually impossible because not only do you have to match Oswald's lighting you have to match his position within the frame to allow for the appropriate image distortion.





JohnM

[Jack Trojan]

 

The code you presented is a simple fade to white, how do you think that is relevant?
And please explain how your code applies to your gif which is just swapping one set of RGB values for another?



JohnM

You also don't write code I see. RGB1 is photo 1 and RGB2 is photo 2. A color is interpolated between 2 corresponding pixels according to the frame number (fn). Say you want 12 intermediary frames for your fade. Divide (RGB1 - RGB2) / 12 then multiply this increment by the frame number and add it to RGB1. This is the pixel color for your fade frame. Get it?