The Backyard Rifle Photos, the HSCA, and Two Scientists

[Michael T. Griffith]

Physicist Philip Stahl, who has written books on physics and astronomy, did a pixel analysis of the backyard rifle photos and found that they must have been tampered with:

"Given some original values dx1, dy1, the mapping will yield diffused dimensions that are within about 30% of the original.  However,  +30%  shows that there clearly had to have been tampering such that the images are not the same." (http://brane-space.blogspot.com/2013/11/analysis-of-pixel-diffusion-in-oswald.html)

Stahl decided to conduct the analysis after reading Hany Farid’s study, which supposedly proved that the backyard photos are authentic. Stahl reviewed Farid’s study and concluded it was severely flawed, as he discusses in the above-mentioned article and in his article “Hany Farid’s Pixelated Illusions”:

http://brane-space.blogspot.com/2009/11/hany-farids-pixelated-illusions.html

Former professor of mathematics Richard Charnin has studied the backyard rifle photos and has likewise concluded they have been faked. If anyone is wondering about Charnin’s credentials, in 1965, he earned a BA in mathematics from Queens College. In 1969, he earned an MS in applied mathematics from Adelphi University, and in 1973, he earned an MS in operations research from the Polytechnic Institute of New York. Here is Charnin’s article “Mathematical Proof that the Oswald Backyard Photos Were Faked”:

https://richardcharnin.wordpress.com/2014/06/09/jfk-mathematical-proof-that-the-oswald-backyard-photos-were-faked/

I would recommend Charnin’s 2014 book Reclaiming Science: The JFK Conspiracy.

The HSCA’s photographic evidence panel (PEP) attempted to deal with the long-recognized problem of the variant shadows in the backyard photos, especially the variant shadows under the backyard figure’s nose. The PEP was unable to duplicate those variant shadows without manipulating the model’s head to a position that bore no resemblance to the position of the backyard figure’s head, as I document in “The HSCA and Fraud in the Backyard Rifle Photos”:

https://miketgriffith.com/files/fraud.htm

The PEP also attempted to explain the long-recognized problem of the virtual sameness of the backgrounds in the backyard rifle photos.

The PEP dealt with this problem by falsely framing the issue and then, inadvertently, providing powerful evidence that the photos are fake. The PEP knew that critics specified that the backgrounds had very slight differences in the distances between objects, and that those differences were far too small for photos that were allegedly taken by a top-view handheld camera that was handed back and forth between exposures. The PEP understood that critics claimed that the same background had been used but that the background had been very slightly keystoned to try to avoid making it brazenly obvious that the same background was used in all the pictures.

The PEP set up the strawman argument that critics claimed that the same background was used for all the backyard photos, period. The PEP “refuted” this strawman argument by announcing that they found “very small” differences in the distances between objects in the background of the photos! How small was “very small”? Incredibly small:

The PEP did parallax horizontal and vertical measurements on selected objects in the backgrounds. The horizontal parallax measurements were done on points on the fence at three levels on 133-A and 133-B. There was an “a” measurement and a “b” measurement, each done at three levels. The differences had to be expressed in millimeters:

a-lower: 0.8 mm
a-middle: 0.1 millimeter
a-upper: 1.1 millimeter

b-lower: 0.5 mm
b-middle: 0.7 mm
b-upper: 0.1 mm

The largest difference was 1.1 mm, which equals 0.043 inches. 0.043 inches as a fraction is 11/256ths of an inch. By comparison, 1/16th of inch is 1.59 mm. So 1.1 mm is 30% smaller than 1/16th of an inch. And, again, that was the largest difference.

The vertical parallax measurements revealed equally tiny differences. These measurements were done on two objects on the fence. To account for differences in magnification, the measurements were related to the distance from the left edge of one picket to the left edge of the next, and the scaling distance was measured on the two center pickets of the four pickets on the fence. The differences--which, here too, had to be expressed in millimeters:

Gate bolt to screen: 1.7 mm
Scaling distance: 0.3 mm
Gate bolt to screen adjusted for scaling distance: 0.15 mm (1.96 mm in 133-A vs. 2.11 mm in 133-B)

Here are the measurements as written in the PEP report:

133A: gate bolt to screen =30.4 mm. scaling; dist.=15.5 mm
30.4/15.5=1.96
133B: gate bolt to screen=32.1 mm, scaling dist.=15 .2 mm
32.1/15.2=2.11

You do not have to be a math whiz to grasp that these are very, very tiny differences. Now just try to imagine the odds of the camera ending up in virtually the same horizontal, vertical, and distance position in relation to the target, to within a tiny fraction of an inch each time, after being handed back and forth between each exposure, in order to produce such incredibly small differences.

The PEP admitted that the camera moved only "slightly" between exposures. "Slightly"? Yes, very, very, very slightly--so slightly that the measured differences between objects in the background were almost microscopic. The measured vertical distance differed by only 0.15 mm (1.96 mm in 133-A vs. 2.11 mm in 133-B). 0.15 mm is only 0.005905512 inches. To get a better idea of how tiny this is, consider that 1/16th of an inch is 0.0625 inches. So we're talking about a vertical difference that is 11 times smaller than 1/16th of an inch. The measured horizontal differences were also amazingly tiny: they ranged from 0.1 mm to 1.1 mm. 1.1 mm, the largest difference, equals 0.043 inches, or 11/256ths of an inch, or 45% smaller than 1/16th of an inch. 

When the HSCA PEP did a Penrose analysis of the backyard figure’s face and of known photos of Oswald, the analysis yielded huge differences between the backyard figure’s face and the other faces. This is especially revealing because the PEP omitted the measurements of the backyard figure’s chin, nose, and ear lobes, even though they knew the chin had been identified as problematic by photographic experts, including Malcolm Thompson. Here are the variations that the Penrose analysis found between the clusters:

Dallas Arrest: 0 shape distance, 0 size distance
Marine: 0.5 shape distance, 0.020/0.025 size distance
New Orleans: 1.6 shape distance, 0.06/0.07 size distance
Russia: 0.9 shape distance, 0.19 size distance
Backyard: 1.75/1.8 shape distance, 0.31/0.32 size distance

As you can see, the closest of the Oswald clusters to the Backyard cluster is the New Orleans cluster, but even it diverges by 9% in shape distance and by a whopping 250% in size distance from the Backyard cluster. And, again, this was the closest cluster; the other clusters showed even larger variations, as you can see.

[Chris Bristow]

The shadow under the nose is often considered to be directly under the nose. The shadow does center at the philtrum but the nose is not on center. Oswald's head is turned about two degrees to his left. If you draw a line from the tip of his nose to the tip of the shadow it is a 4 degree angle.
  Oswald had to turn 22 degrees South of West to face Marina. The Sun was at about 35 degrees South of West or Azimuth 235( Azimuth starting in the North). So Oswald was facing about 13 degrees away from the Sun. But his head was turned an additional 2 degrees to his left which lessens the angle of shadow under his nose to 11 degrees. So we should see about 11 degrees of angle if drawing a line from the tip of his off center nose to the tip of the shadow. (Regardless of the Suns elevation it's shadow will fall directly below the nose when a persons turns to face directly towards the Sun.)
 But there is one more factor. Just as turning you head on a horizontal axis to the direction of the Sun causes a shadow that matches a high noon shadow, the same effect happens if you tilt your head to the side. If you tilt your head towards the sun and the angle of you head matches the elevation angle of the Sun, you will again have a shadow that falls directly below the nose the same as high Noon. This happens when you rotate horizontally or tilt your head on the vertical axis.
  Oswald facing 13 degrees away from the Sun should create a little less than a 11 degree shadow under his nose. (The change of shadow relative to the change of facial angle is not a one for one relationship). We should see 11 degrees but because his head is rotated 2 degrees left it cancels out 2 degrees and we should see  only 9 degrees. With his nose rotated 2 degrees left the angle from the tip of the nose to the tip of the shadow centered at the philtrum is 4 degrees. The fact that his head is tilted 4 degrees toward the Sun cancels out 4 degrees of the shadow we should see.
  So 11 degrees total shadow is seen in 4 degrees of shadow angle under the nose, 4 more degrees cancelled out by the head tilt and 2 more degree cancelled by the 2 degree horizontal head rotation toward the Sun. That pretty much accounts for the position of the nose shadow.

When people attempt to dupllcate Oswald's lean they generally hold the hips at a 35 degree angle to the camera(right hip farther back). They also tend to ignore the angle of the right foot and turn it farther than Oswald.
  Last year I noticed that the shadow of the telephone lines crossing Oswald's hips are at a 9 degree angle compared to the direction of the shadow on the ground. This turns out to be a measure of how much Oswald's hip are angled back from the camera. It turns out his hips were facing almost directly at the camera, no more than 7 degrees away from facing directly at the camera.
   It is hard enough to duplicate the stance with the hips at 35 degrees. It becomes very painful and nearly impossible at 20 degrees. If you restrict the hips to 7 degrees it is imo impossible. So if you try to lean over with the right shin and knee aligned vertically and bring your belt buckle or the snap on your pants within one inch that same vertical line it becomes impossible. If anyone wants to attempt it you wwill need to keep your right foot swung out to no more than 45 degrees. !33a will measure 68 degrees out but that is due to perspective. A 45 degree angle seen from above(The accurate measure) will look like a 68 or 70 degree angle from Marinas position at 10'4" away.
 

[Duncan MacRae]

[Michael T. Griffith]

The shadow under the nose is often considered to be directly under the nose. The shadow does center at the philtrum but the nose is not on center. Oswald's head is turned about two degrees to his left. If you draw a line from the tip of his nose to the tip of the shadow it is a 4 degree angle.
  Oswald had to turn 22 degrees South of West to face Marina. The Sun was at about 35 degrees South of West or Azimuth 235( Azimuth starting in the North). So Oswald was facing about 13 degrees away from the Sun. But his head was turned an additional 2 degrees to his left which lessens the angle of shadow under his nose to 11 degrees. So we should see about 11 degrees of angle if drawing a line from the tip of his off center nose to the tip of the shadow. (Regardless of the Suns elevation it's shadow will fall directly below the nose when a persons turns to face directly towards the Sun.)
 But there is one more factor. Just as turning you head on a horizontal axis to the direction of the Sun causes a shadow that matches a high noon shadow, the same effect happens if you tilt your head to the side. If you tilt your head towards the sun and the angle of you head matches the elevation angle of the Sun, you will again have a shadow that falls directly below the nose the same as high Noon. This happens when you rotate horizontally or tilt your head on the vertical axis.
  Oswald facing 13 degrees away from the Sun should create a little less than a 11 degree shadow under his nose. (The change of shadow relative to the change of facial angle is not a one for one relationship). We should see 11 degrees but because his head is rotated 2 degrees left it cancels out 2 degrees and we should see  only 9 degrees. With his nose rotated 2 degrees left the angle from the tip of the nose to the tip of the shadow centered at the philtrum is 4 degrees. The fact that his head is tilted 4 degrees toward the Sun cancels out 4 degrees of the shadow we should see.
  So 11 degrees total shadow is seen in 4 degrees of shadow angle under the nose, 4 more degrees cancelled out by the head tilt and 2 more degree cancelled by the 2 degree horizontal head rotation toward the Sun. That pretty much accounts for the position of the nose shadow.

Nobody has been able to duplicate the variant nose shadows in live reenactments, neither with people nor with models. The HSCA could only duplicate the nose shadows by angling the model's head so far that it was no longer facing the camera and then by moving the camera. When HSCA PEP member McCamy tried to sell this to the committee, Congressman Fithian challenged him, and McCamy admitted that "there would be a number of assumptions necessary . . . to interpret the Oswald photograph from this demonstration of this effect."

McCamy then said, in effect, that it didn't matter that they couldn't actually duplicate the shadows in the reenactment because they had authenticated the shadows with a vanishing point analysis.

But anyone who knows anything about vanishing point analysis knows that many fake photos could "pass" a vanishing point analysis. A vanishing point analysis merely states that a shadow should fall in a line, but it does not establish at what point on the line the shadow should fall. There are countless shadows in a given photo that would "pass" a vanishing point analysis, even though only one of them would be correct, which makes it even more curious that the PEP only did their vanishing point analysis in one dimension, making it even easier for the shadows to "pass" the analysis.

I interviewed numerous professional photographers and photo lab technicians about this, including a school-trained NSA photo lab tech and photographer/photo lab tech who had taught photography at the collegiate level, and and none of them believed that a vanishing point analysis could explain these kinds of variant shadows.

If the backyard rifle photos are authentic, duplicating those shadows should be a simple matter of having someone stand outside and having them strike the same poses that the backyard figure does, and then seeing where the shadows fall. Larry Schiller tried to do exactly this, but his resulting photos did not duplicate the nose and neck shadows, especially the neck shadows.

In 1966, the London Times conducted its own backyard rifle photo simulation in an attempt to duplicate the variant shadows. Its photographers failed to do so. As Stewart Galanor notes in Cover-Up,

"The London Times's test shows that when the shadow of the nose falls straight down, the shadow of the body is behind. When the shadow of the nose veers off to the right, so does the shadow of the body." (Cover-Up, p. 81)

Anyone can see this for themselves in the London Times reenactment photos, which are shown in document 35 of Galanor's Cover-Up (p. 81).

And I would bet my IRA and 401K that no one will ever do a realistic reenactment that will produce photos with only incredibly tiny differences between the objects in their backgrounds. The almost-microscopic differences between the objects in the backgrounds of the backyard photos prove that a forger used the same background for each picture but slightly keystoned the background. There is no way on this planet that you would get such extremely tiny differences in background-object distances in three photos taken with a top-view handheld camera that was supposedly handed back and forth between each exposure.

[Duncan MacRae]

And I would bet my IRA and 401K that no one will ever do a realistic reenactment that will produce photos with only incredibly tiny differences between the objects in their backgrounds.

1967


http://jfk-archives.blogspot.com/2010/06/backyard-photos-part-2.html

[Michael T. Griffith]

1967


http://jfk-archives.blogspot.com/2010/06/backyard-photos-part-2.html

What is this supposed to prove? This only shows the 133-A pose, with the head not tilted. Where is the reenactment photo that shows the 133-B pose, with the head tilted? The nose shadow will fall straight down if the head is not tilted, so we would need to see a reenactment photo with the figure's head tilted, as in 133-B, in order to compare the nose shadows.

British forensic photographic expert Malcolm Thompson explained the problem with the nose shadows:

In photographs A and B you see Oswald's face in a different posture and yet the shadow under the nose hasn't moved or if it has moved it is only fractional compared with the actual movement we see in the face and one comes to the conclusion that, it is the same. picture used for both faces.

So if a reenactment only does one pose, the one with the head upright, it proves nothing about the nose shadows. A credible reenactment must start with doing both poses, the pose in 133-A and the pose in 133-B. Otherwise, no shadow comparison is possible.

And where are the parallax measurements of the distances between objects in the backgrounds of the two reenactment photos (assuming there are two reenactment photos)? If this reenactment produced only photo, one pose, it is worthless.

Ah, I see you got this from DVP's site. Mr. Von Pein does not seem to understand the basic issues involved here, or he would know that this photo proves nothing.

[Chris Bristow]

The reenactment photo with Mr Cappel has some major problems. Their postures are very different, Mr Cappel's hips are over his feet much more than Oswald in 133a. But the larger problem is before you can test it you have to rotate the photo's so they are both level. Looking at either the picket fence or the stairway post next to Oswald it is obvious that 133a photo is rotated almost 4 degrees to the right compared to the Cappel photo. The picket fence should lean at 2.25 to 2.5 degrees right when level. So Cappel needs to rotate 2 degrees right and Oswald 2 degrees left. Oswald is only leaning maybe 6 degrees so even a 2 degree rotation is a big deal.
 One great thing about the Cappel photo is the door jam on the house next door is a sharp clear image. Door jams are a good guide to find level because if they are even a little off the door gets stuck. Most all the vertical lines in 133a are wonky and the door jam is blocked by the open screen door so it is nice to have the door jam visible in the Cappel photo. The slight distortions from a keystone effect due to the camera being tilted down slightly and any pin cushion on the door jam are tiny and can be ignored. But also correcting for them would mean rotating the 133a image to the left which would add to Oswald's lean. So by not correcting he leans a bit less. The way those two images were rotated made for a very misleading comp photo.
 It has been claimed more than once that Mr Cappel is the same height as Oswald even though he looks very lanky and taller than Oswald. They point to his head in relation to the roof in the background. But when you look at that roof and where it intersects with  Oswald's stairs it is clear that the camera was several inches higher than 133a. So if they both appear the same height relative to the post behind the head or the roof line then Cappel has to be about 4 inches taller. you can also just look at how Oswald and Cappel compare to the post next to them or the tread on the stairs. Cappel looks like he may be standing closer than Oswald and that would explain his apparent height difference. But his back leg is in the same place as Oswald's relative to the post. His left leg is farther forward and it makes him look closer overall. His left foot being farther forward would drag his torso a couple inches forward but I don't think that is enough to account for height difference we can see relative to the post.
 
     Michael T. Griffith, I believe my explanation in the last post completely explains the nose shadow in 133a. I have also used a dummy head and reproduced it in natural sunlight. Those images are gone with my last hard drive but I am going to reproduce it.  I have not compared the nose shadows 133a and B but it is interesting and I will look at it. you mentioned that Cappel's head is not tilted 4  degrees right so it is not a valid comparison. But after the image is rotated 2 degrees right to level it Cappels head is tilted 4 degrees right. When measuring the tilt it is important to draw the line by using the center of the mouth and then bisecting the bridge of the nose. Because Cappel is also looking 2 or more degrees to his left, his nose will not line up with the center of his face and so it will give a false measurement.
 Regarding the use of keystoning to fake different camera positions it would not create the effect of the roofline in the background intersecting Oswald's stairs at different places. in133a the roofline hits the post next to Oswald a couple inches lower than 133c and a tiny bit lower than 133b. Keystoning can magniy, stretch or compress one axis of a photo but it will never change were the roof and post intersect. I have keystoned 133a and other photos and you cannot make the positions of the roof and post change relative to each other. You can change relative position if those position are not right next to each other. That is because the effect changes as you move from top to bottom(Assuming the keystone is vertical). But objects next to each other experience the same keystone and simply shrink or expand together. so it would be impossible for a keystone effect to change where the roof meets the post. There does appear to be a real camera height difference in the 3 backyard photos.

[Michael T. Griffith]

Michael T. Griffith, I believe my explanation in the last post completely explains the nose shadow in 133a.

Did you read the articles linked in my OP? Did you read Dr. Stahl's pixel-diffusion analysis of the backyard photos? Using differential equation fractal analysis, Stahl shows that the V-shaped shadow under Oswald’s nose is essentially identical in all four photos. Stahl debunks the pixel analysis of software scientist Hany Farid, who claimed the photos were genuine.

There is no need for an explanation for the nose shadow in 133-A. In 133-A, the nose shadow looks the way it should because the figure's head is virtually upright/not tilted. 133-A's nose shadow is not the problem. 133-B's nose shadow is the problem.

The HSCA could not duplicate 133-B's nose shadow without altering the angle of the model's head so far that "the subject is no longer looking at the camera" (2 H 414). This is part of what Congressman Fithian called out McCamy on, and that is when McCamy said, in effect, "Well, don't worry that we could not duplicate that nose shadow because we confirmed all the shadows with a vanishing point analysis."

I have also used a dummy head and reproduced it in natural sunlight. Those images are gone with my last hard drive but I am going to reproduce it.

Well, again, I'm certain you were able to duplicate the 133-A nose shadow, because that nose shadow looks the way it should given the position of the figure's head. Again, 133-A's nose shadow is not the problem.

I have not compared the nose shadows 133a and B but it is interesting and I will look at it. you mentioned that Cappel's head is not tilted 4  degrees right so it is not a valid comparison. But after the image is rotated 2 degrees right to level it Cappels head is tilted 4 degrees right. When measuring the tilt it is important to draw the line by using the center of the mouth and then bisecting the bridge of the nose. Because Cappel is also looking 2 or more degrees to his left, his nose will not line up with the center of his face and so it will give a false measurement.

The Cappel photo in question was apparently designed to duplicate 133-A, so, again, it proves nothing. Several reenactments have been done, and so far none of them have duplicated the variant shadows seen in the photos.

Regarding the use of keystoning to fake different camera positions it would not create the effect of the roofline in the background intersecting Oswald's stairs at different places. in133a the roofline hits the post next to Oswald a couple inches lower than 133c and a tiny bit lower than 133b. Keystoning can magniy, stretch or compress one axis of a photo but it will never change were the roof and post intersect. I have keystoned 133a and other photos and you cannot make the positions of the roof and post change relative to each other. You can change relative position if those position are not right next to each other. That is because the effect changes as you move from top to bottom(Assuming the keystone is vertical). But objects next to each other experience the same keystone and simply shrink or expand together. so it would be impossible for a keystone effect to change where the roof meets the post. There does appear to be a real camera height difference in the 3 backyard photos.

Then the HSCA PEP stunningly failed to detect that alleged "real camera height difference" with their parallax measurements. The PEP would have been thrilled to report that they found marked differences in the camera height in the photos. The PEP was forced to conclude that the camera moved only "slightly" between exposures. If anything, "slightly" is a bit of an overstatement. To produce the incredibly tiny background-object differences, the camera would have had to move very, very slightly, a matter of a small fraction of an inch, horizontally and vertically.

Hershel Womack, a professor emeritus of photography at Texas Tech University, has identified other indications of fraud in the backyard photos:

If Lee Harvey Oswald cocked the shutter each time for Marina as she supposedly stated, then how did Oswald's leg stay in the same place relative to the dark area next to his left, photo right knee? Compare in two of the three photos. Measure from the line on the building on the right and measure to different parts of his body and I think you will reach the same conclusion. Note the Roscoe White backyard photo.

Other measurements from a fixed object like the stair post to portions of his body or even the pistol appear to be the same or near so. There's no way you could move and go back to the identical spot and take the same position without drawing the image on the back of the camera.

With this in mind then how did he get taller if neither he nor the camera moved. Maybe the camera was on a tripod and lowered which would make him taller but it would do the same to the post which may be a little taller but the height of Oswald seems out of proportion to that of the post or vice-versa. (https://miketgriffith.com/files/faulty.htm)