Friday, November 6, 2009

Is Dartmouth Professor's Analysis of Oswald Backyard Photo Flawless?


Above is a "backyard photo" -- a high resolution rendition (in pixels, because it's digitally created from a scan, then conducted over the Internet) of one of the photos of Lee Oswald that is supposed to show him holding the "killer rifle" and "The Worker" and "The Militant" --newspapers of opposing Communist/Socialist/Marxist persuasions. The original photo is an emulsion printed on photo paper, of course, with no pixels. Nevertheless, it's surprisingly fuzzy, because the lens and camera taking the photos were both sub-standard.

There are some surprises in the photos, however, and they are unsettling to those with open minds.

The fingers holding the newspapers are too fat and stubby to be Oswald's. The index finger appears, too, to be the same length as other fingers. The chin, despite recent protests that Oswald's chin could 'look' squarish under certain shadow conditions, is too 'square' to be seriously considered as Oswald's. And Lee Oswald himself said his head had been pasted onto the body of whoever posed for the photos, two of which he was shown by Dallas police.

Researchers independent of the government have concluded almost unanimously that the backyard photos were faked, for numerous additional reasons, most of which are listed in this article. The newspapers (size known) are too big for the supposed height of Oswald (height known); see the correct sizes superimposed on the backyard photo here:



..............................................
Note that the index finger and ring finger are stub-cut to the same length: somebody was a little careless with an Xacto blade, or airbrush, it seems. Oswald had normal fingers.
Difficulties in the photos such as this are not addressed by Professor Hany Farid. Therefore, his statement that the photos are genuine because his computer program found no 'alterations' in the photo itself is subject to criticism.

Other problems also exist: the rifle in the photos has a different kind of sling mount than the rifle in the National Archives (on exhibit as the original killer rifle). You can find both rifles on YouTube --and see for yourself the different slings. Researcher Gil Jesus makes the point: "..the bottom sling mount, so clearly visible in the backyard photos, does not exist on the rifle removed from the Texas School Book Depository...it is not the same rifle that is currently in the National Archives." Go to YouTube and see for yourself:

http://www.youtube.com/watch?v=vq25loEpBro



There's more. Fingernails are missing from the strangely stubby fingers, too, as if the newspapers were slipped in under a cutout set of fingers--cut out with a sharp little Exacto knife, as all photographers used to use to 'insert' extra things into photos, before computers came along.

Significant, too, is Marina Oswald's testimony regarding the backyard photos. She said she looked "into" the camera's viewfinder to take the photos, but in fact, the camera had to be held at the waist, as it did not have a viewfinder. She later also stated that she took photos from a different direction. Also, deceased Dallas police officer Roscoe White's possessions included an Oswald backyard photo that nobody had ever seen, and the Dallas police had a strange additional 'backyard photo' that shows a figure cut out. Oswald was shown two photos --it's on record-- before the photos were officially 'found' at the Paine garage in Irving. These photos were blow-ups from the original 3" x 3" format--showing that the originals had already been copied in a photo lab somewhere and then given to the police. Why wasn't Oswald shown the originals? It was only one day after the assassination, and already, these photographs, said to be of Oswald, had been copied and enlarged in photo labs.

The camera itself would not be found for another month and a half. It was discovered in the Paine's garage(the Paine residence is where Oswald's wife was living -- Oswald lived separately from her). However, that garage had been scoured for Oswald's possessions by the police three times by then. Lee Oswald's brother, Robert, "found" this camera, soon after he was caught in a compromising situation with his brother's widow (Mellon's 2007 book, Mrs. Paine's Garage, which is opinionated against Lee Oswald, revealed that a sexual liason occurred). Robert Oswald soon after went with others to the Paine garage, where he 'found' the camera(how convenient!). He said he knew it was Lee Oswald's camera, because Lee had given it to him before going to Russia. Then, after returning to the US, at some point, Lee asked for it back. It was a cheap, inferior camera (an Imperial Reflex, made in the USA). Yet Oswald owned some advanced, expensive cameras and even called himself a professional photographer, describing himself as such on employment forms, etc. He even had equipment to make 3-D photos. In other words, it is unlikely that he would have asked for the cheap camera to be returned, and it would be even more unlikely that he would use the camera for any photos, which would be inferior to what his other cameras (Kodak, Minox, etc.)could do. And how would he have even thought of the cheap camera at all, after three years away, when he, himself, owned several fine cameras?




In an official Dallas police photo of possessions belonging to Oswald, there is no Imperial Reflex camera. Nor is the flash attachment seen in this photo associated with that camera. The camera only showed up much later, when people started wondering where it was. Note the fine camera equipment, in this photo, that belonged to Oswald:


We have a recurring pattern: a cheap camera --a cheap rifle -- a cheap handgun -- owned by a man with expensive cameras who insisted he never ordered the cheap rifle or the cheap revolver, either.

The HSCA later displayed a photo supposedly made by Oswald of his toddler daughter, June Lee, standing in the front yard of his New Orleans residence (Ref: http://www.history-matters.com/archive/jfk/hsca/reportvols/vol6/html/HSCA_Vol6_0083b.htm HSCA Appendix to Hearings - Volume VI, p. 160). The photograph was used to help prove that the camera that took the backyard photos had been used by Oswald previously to take photos of his family. Here's the photo:



This exhibit is a photo of the photo, and there are problems with it.
The Imperial Reflex had a fixed focus. You couldn't take a close-up photo with it and expect it to be in focus. The original "backyard photos" are famous for showing quite a bit of detail --and the originals are in focus. But this photo is not in focus--nor are the fringes of the other two visible on the roll of "family photos" in this exhibit. If a photo OF a photo were taken by the Imperial Reflex, even if the original had been blown up, it would still be difficult to photograph at a distance sufficient to allow it to be in focus. The fuzzy result, however, taken by the Imperial Reflex, could be used to 'prove' Oswald had the camera, and used it for other photos. Is this why this "family photo" is so amazingly out of focus?

Because the Dallas police "found" these photos, and others, we will take a look at one "backyard photo" the Dallas police finally were compelled to admit that they themselves made (after it was found much later) -- featuring a cutout where Oswald's figure could be placed. We will also look at a similar photo found among deceased Dallas Police officer Roscoe White's possessions, where the figure of Oswald is in place, because the two photos look quite similar in pose, as if one had been inspired by the other, though found years and miles apart.

But first, we need to look closer at the more familiar photos, one of which graced the cover of Life Magazine. (Life editors eventually admitted they modified and altered portions of the rifle on that photo. Why?



Researcher Ralph Thomas (EXAMPLES OF DIGITAL
EVIDENCE ANALYSIS AND PRESENTATION: BACKYARD PHOTOS EVIDENCE OR FAKERY?
www.pimall.com/nais/news/backyard.html) tells us:
"The photographs were found at 2515 West Fifth Street in Irving, Texas, the home of the Paines. On two searches on the day of the assassination of this address, the Dallas police did not locate the photographs. However, another search was made the following day and it was this third search in which the Dallas police say they found the photos. But the two photos were never listed on inventory sheets of Oswald's possessions. Neither was the black shirt and the black pants that Oswald had on in the photographs ever located. Officially, two photographs but only one negative were found. Yet, Dallas police Gus Rose says that there were also two negatives. Until 1967, these photos were the only ones known...of the backyard pose."

Thomas adds that "The Warren Commission said that the photos were taken by Marina Oswald, Oswald's wife, on March 31st, 1963 at the 214 W Neely address. The photos show a bright sunny day. But, a check on weather reports in the area that day reveal that it was cloudy and rainy all day."

But then, more photos were found: again, we quote from Ralph Thomas;
"In 1976 the Senate Intelligence Committee located a third photograph of Oswald with the backyard pose that was slightly different. The photo was found among the belongings of the widow of Dallas police officer Roscoe White. In the early 1990's Rosco White's son claimed that Rosco White was one of three gunmen that fired at the President...

In 1967 after he returned from Haiti, George DeMorenschildt found a fourth photo of the backyard photograph...[it was]a slightly different pose...On the back of this photograph is written: HUNTER OF FASCISTS, HA HA HA.

The inscription was first written in pencil and then gone over in ink. Handwriting experts have revealed that the writing was not the handwriting of Oswald, his wife, DeMorenschildt or his wife... In 1970, Dallas news reporter Jim Marrs was looking into the backyard photographs when he interviewed Robert and Patricia Hester. The Hesters worked at the National Photo Lab in Dallas. They said they were very busy processing photographic material for both the FBI and the Secret Service the night of the assassination. In 1970, the Hesters told Marrs that the FBI had color transparencies of the backyard photographs the night of the assassination and had one color transparency that had nobody in the picture. Not only is this highly suspicious, this was the night before the photographs were supposed to have been found in the first place."

The researcher then makes a list of the problems that photo expert Jack White found with the backyard photos:

1) STANDING OFF CENTER: White concludes that Oswald is standing off center and outside the weight bearing alignment of his feet. A person could not stand in such a position.
2) PROPORTIONS: When the body proportions are brought into alignment from the knees to the head by adjusting the size of the photographs, one head is much larger than the other.
3) OVERALL BODY SHADOWS: Although the photos were supposed to have been taken just seconds apart, the overall body shadows in the photographs are all different. In 133-A the photograph has a 10 o'clock shadow, 133-B a 12 o'clock shadow and 133-C a 10 o'clock shadow again.
4) ARM AND ELBOWS: White said that the elbow is too high in one photograph and the elbow doesn't show up on the one photograph of the arm were Oswald is holding the rifle. This pose had been attempted to be duplicated but could not.
5) HANDS AND FINGERS: On the photographs the left hand and finger looks normal. Yet the right hand is missing fingernails and the hand looks stubby.
6) WATCH: The photographs reveal that Oswald is wearing a watch but all witnesses have stated that Oswald did not wear and didn't own a watch. No watch was found among the possessions of Oswald and he was not wearing one when he was arrested.
7) RIFLE: When the photographs are blown up to the actual height of Oswald that was 5'9", the rifle in the photograph is too long. When the rifle is adjusted in the photograph to its proper length, Oswald's height is six inches too short.
8) SCOPE: White noted that in the photograph the rear end of the rifle scope is missing and pants wrinkles appear where the end of the scope is supposed to be.
9) FACE: The face shows Oswald with a flat chin but Oswald had a cleft chin. There is a line that breaks up the grain of the photograph that runs across the chin that many say is where the cut took place to paste Oswald's face onto the photograph.
10) PHOTOGRAPHIC OVERLAY: When Mr. White took 133-A and 133-B and adjusted and overlayed them, nothing matched up, which isn't supposed to happen with two slightly different poses. However, the faces on the two photographs did.
11) FACE SHADOWS: Both photos show the same V shaped shadow below the nose. However, on one of the photos Oswald's head is tilted but the shadow does not adjust for this tilt.

[Look at this graphic example:]



12) NECK SHADOWS: On one of the photos there is light on the right side of the neck but the same photo shows the rifle casting a shadow to this angle.

13) COLLAR SIZE: The collar size can be determined from the photograph using a mathematical formula which came out to size 16. Oswald wore a size 14 1/2 collar and all his clothes found among his personal belongings were in the 14.5 to 15 inch range.
14) BACKGROUNDS: White determined that one photograph had the top cropped off and the other photograph had the bottom cropped off which made the photos appear like they had been taken at slightly different locations. However, except for small fractions, everything lines up on both photographs when the two were compared. That is, the camera did not change position and the only way to do this would be with a tripod which was not used.
15) SMALL DIFFERENCES: For many months White was puzzled by the small differences he noted in the backgrounds but they were not off much. After looking at the photographs some more, he determined that on the background of one, the camera appears to be slightly tilted. He then took another copy of the photo by tilting it on a board and everything came perfectly into alignment...
[16)FEET:] During the 1991 JFK Assassination Symposium held in Dallas,Texas of November of that year, computer image processing expert Tom Wilson corroborated all of the White analysis and added that he inspected the feet on the man in the backyard photograph as to light refraction and compared this to official records of the day concerning the position of the sun. Wilson stated that the photograph was taken at 9:12 A.M. if it was taken on the day it was alleged to have been taken. But Marina Oswald's testimony stated that the photographs were taken in the early afternoon which is completely inconsistent with the Wilson study."

Other scientists who had not previously been associated with Warren Commission critics also have questioned the authenticity of the photos. For example, Malcolm Thompson, a British forensic photography expert, questioned the authenticity of the photographs in a 1978 BBC television documentary.

Marina Oswald herself not only gave two different versions as to where the backyard pictures were taken, she also gave different versions of the number of photos taken. At first, she testified she had taken just one photo, but later changed it to two. Finally, she said she took the photos from a different direction. In addition, Marguerite Oswald testified that soon after the assassination, Marina showed her a third photo, which she had hidden in her shoe. In that photo, Oswald was shown holding the rifle over his head with both hands. This is the same pose that can be seen in the Roscoe White photograph.

But since Marina did not know how to use the camera, at first admitted to taking only one photo, then said two -- then changed the very direction in which she said she was taking photos -- as with so many other things Marina said under duress and fear of deportation, her testimony in this matter is as difficult to reconcile with reality as her testimony that Lee Oswald planned to shoot former Vice President Richard Nixon -- a story even the Warren Commission didn't believe. It seems Marina had to say whatever she had to. Nobody can blame her for quailing under pressure.

The issue with Marguerite Oswald, however, is more serious. The photo could have been planted, and Marina could have snatched it up, afraid the police would see it. Perhaps she will someday explain how she had it on her person, for the photo is not the kind one keeps handy while living in the home of a Quaker (Ruth Paine) who was by religion supposedly extremely anti-weapon/anti-war/anti-violence.

The photo offered by George DeMohrenschildt has its problems, too: it seems to have been created as a "first generation print" straight from one of the originals, as determined by the HSCA (HSCA 381-382, Figures IV-20, IV021). How did DeMohrenschildt get it? And who wrote "Ha ha ha!" on the back? George would later be found with a shotgun shoved into his mouth a few hours before the HSCA's intrepid investigator, Gaeton Fonzi, could question him about that photo -- or anything else. Though ruled a suicide, as so many 'mysterious deaths' in the case have been, researchers have noted that in the room where DeMohrenschildt was found with his head blown to pieces, in the background of a tape recording that was being made at the time, one can hear the security alarm beeping: an intruder had entered about a minute before the shotgun's blast was heard. Here is the photo the poor fellow "found":



Our Dartmouth professor does not consider any of these provenance difficulties. He supposedly answers all questions and problems for all time with his one dummy head, his computer program, and counting pixels (pixels that did not exist in 1963).

The provenance and deliberate creation of these photos has, in other words, been a problem to consider from the very beginning, and the professor's pixel-counting isn't going to make it go away. That won't halt continuing efforts to find yet another way to 'prove' the backyard photos weren't faked, as so much other evidence was faked.

Our professor used a digitalized computer program (but none of the problems we've mentioned were mentioned by him -- just the concern as to whether the photo was authentic). His method also leaves much to be desired. His computer works with pixels, and the originals are emulsions, no pixels involved. Any photo made with pixels of the originals would not show tampering unless the pixel photos had been tampered with. Of course that didn't happen. That leaves us with the outside study materials, of which, apparently, photos were taken-- materials that excluded, it seems, the whole body length that showed shadows extending in a different direction from shadows on the dummy face set up for the 'outdoor shot.' Flawed, flawed!

Our Dartmouth professor nevertheless concluded for all time that the backyard photos were not faked, based on his computer program's result. In the YouTube video, the Dartmouth professor (who inaccurately calls Oswald a "communist") presents the computerized dummy's head showing shadows that he states are the same as in the backyard photos. The 3-D game dummy on the screen, he says, provides digitalized evidence that the photos are not faked. But what part of 'faked' are we talking about, here?

We are not told much about how the result was created. Of concern is the fact that apparently only a 3-D head was used, not an entire 3-D body, plus "portions of the backyard scene." What portions? How did he choose the portions? As we know, shadows can seem to be going in various directions, according to how leaves, etc. are positioned. What 'portions' were selected? This cherry-picking kind of data selection has been done before. But let's read the article, below, and then look at examples, below the article, of some disturbing anomalies that the good professor failed to address: (Bold face = my emphasis).

"Dartmouth Computer Scientist Hany Farid has new evidence regarding a photograph of accused John F. Kennedy assassin Lee Harvey Oswald. Farid, a pioneer in the field of digital forensics, digitally analyzed an iconic image of Oswald pictured in a backyard setting holding a rifle in one hand and Marxist newspapers in the other. Oswald and others claimed that the incriminating photo was a fake, noting the seemingly inconsistent lighting and shadows. After analyzing the photo with modern-day forensic tools, Farid says the photo almost certainly was not altered.
“If we had found evidence of photo tampering, then it would have suggested a broader plot to kill JFK,” said Farid, who is also the director of the Neukom Institute for Computational Science at Dartmouth. “Those who believe that there was a broader conspiracy can no longer point to this photo as possible evidence.” Farid added that federal officials long ago said that this image had not been tampered with, but a surprising number of skeptics still assert that there was a conspiracy.
The study will appear in a forthcoming issue of the journal Perception.
Farid and his team have developed a number of digital forensic tools used to determine whether digital photos have been manipulated, and his research is often used by law enforcement officials and in legal proceedings. The tools can measure statistical inconsistencies in the underlying image pixels, improbable lighting and shadow, physically impossible perspective distortion, and other artifacts introduced by photo manipulators. The play of light and shadow was fundamental in the Oswald photo analysis."

Note that the computer works with digitalized photos. The original photos, of course, were not digitalized. Reproductions of such photos, using pixels, of course will show no manipulation unless they have been additionally doctored. To continue:

“The human brain, while remarkable in many aspects, also has its weaknesses,” says Farid. “The visual system can be quite inept at making judgments regarding 3-D geometry, lighting, and shadows.”

At a casual glance, the lighting and shadows in the Oswald photo appear to many to be incongruous with the outdoor lighting. To determine if this was the case, Farid constructed a 3-D model of Oswald’s head and portions of the backyard scene, from which he was able to determine that a single light source, the sun, could explain all of the shadows in the photo.
“It is highly improbable that anyone could have created such a perfect forgery with the technology available in 1963,” said Farid. With no evidence of tampering, he concluded that the incriminating photo was authentic.
”As our digital forensic tools become more sophisticated, we increasingly have the ability to apply them to historic photos in an attempt to resolve some long-standing mysteries,” said Farid."

===Response===

Prof. Farid tells us that we can be fooled by what we see. Indeed we can. Note that "as our digital forensic tools become more sophisticated, we increasingly have the ability to apply them to historic photos..." --- HOW they are applied depends upon the person running the program: the program can be tweaked to produce what is desired. Because there are computer programs available that can come to such a conclusion does not mean that the data used was good. Once again, it seems that ONLY the head was used--not the body, with its attendant shadows. Why?

The ONLY acceptable procedure would be to analyze the shadows along the entire body and the head, as well as shadow advances between photos along the ground, as these obviously (and unaccuntably) differ. Once again, no such stringent set of factors necessary for a truly comprehensive analysis was conducted. Though portions of the backyard photo may have been selected that were compatible with the shadows on the head itself, we were not told what other portions of the photo were selected to set up for 3-D analysis. Nor are we told why only portions, instead of everything in the photograph, were analyzed. Surely the professor had the time and equipment to use a full-sized dummy, didn't he? So why didn't he?
The only truly scientific methods now available are two: (1) analyze the entire scene in full 3-D, and then analyze the digital photo taken of the scene in all calculated angles as in the originals -- not just selected portions -- using the digital analysis program involved here, and then see if the results can be duplicated elsewhere. The results should come out the same for everyone. The program used should be able to handle a variety of photographs made of such 3-D scenes the same way, with the same results for all. Monitoring the method of analysis would have to include knowing how the computer program was set up, and it should be tested immediately before and after use in the project, on standardized photos with known parameters, concerning shadows, to ascertain that the program has not been manipulated.

Of importance is that the analysis is able to be reproduced by others using the same methodology in every aspect. Unfortunately, we have evidence abounding that computer data can be manipulated. Just to remind you once more: look at the Dallas police cutout photo, and the one found in former Dallas Police Officer Roscoe White's belongings:





What is this photo about? I used to work for Steck-Vaughn Publishers in Austin, Texas. we would cut out an area and place dark red cellophane in that spot. Then we could slip a different photo in the cut-out area. The cut-out is created to place a different photo in the space. Why did the Dallas police have such a cutout in their possession? They said they were just testing to see if a cutout was possible. Of course, the answer is "yes."



..........................................................................

(2) By far the easiest and most trustworthy method to determine the authenticity of these photos is to use a method where data cannot be manipulated. Method (obvious, but not pursued by government invesigators): use an Oswald lookalike (or lifesized dummy with the same features as Oswald) of the same height, dressed similarly, etc. leaning so awkwardly at precisely the same angles (if possible -- researchers have reported great difficulty keeping their balance posing at the calculated angle of presentation)-- with the TRUE size papers and rifle, and with all objects held at the exact angle as in the originals (tests have shown that an elbow seems to be missing in one backyard photo -- that the rifle cannot be held, as shown, without the elbow showing--but it is, nevertheless, missing). Finally, the light source must be solely the natural sun, at the proper angle, with the same weather conditions and time of year to create the long shadows as shown. The fence, plants, etc. are easily put into the proper places, using simple geometrical measurements. Then take the photos. This simple experiment was nevertheless apparently impossible for either the Warren Commission or the HSCA to properly conduct.

---Using only the head ---or only the body--- is not a new idea---

Warren Commission exhibit 748, below, shows the 'model' (not leaning enough) --with no head --a graphic example of the difficulty the FBI experienced trying to create the same shadow effect. Rather than show the "wrong" shadows on the face, we must conclude that the Warren Commission/FBI decided to avoid the problem by blocking out the head.



Next, we have an attempt by the HSCA to show how the head shadows are possible--but no body is shown. Note that they do not show any light on the (our) right side of the neck, as is shown in actual Oswald backyard photos in question. One must ask why.




Note the dummy head only is shown (because the light on the whole body would never have created shadows stretching out behind it?). Now look at an Oswald backyard photo:



Note the angle of the sunlight on (our) right side of neck:compare to HSCA photos showing no light at that angle (would the nose shadow have changed?).


So should you trust the computerized results offered by the Dartmouth professor who says Oswald was a communist? Remember, we didn't see an actual photo reproduced by the professor. We didn't see what parts of the backyard photo(s) he selected to match up with the dummy head and its computerized shadows. We only saw a computerized 3-D game dummy model reproduced on a screen, with nothing behind it. Below is a common shadow processing program assignment for computer students, where shadows can be manipulated any which way on a dummy replica:
----------------------------------------------

"This assignment investigated, and implemented common approaches used in creating real-time shadow effects within a 3D game environment under the DirectX API. The assignment consisted of a written investigative report and two demonstrations of real-time shadow rendering using the planer shadows and shadow mapping techniques."



Planer Shadow Volumes: computer programs allow you to adjust shadows just as you please:





Why would an ignorant (as to details of the case, e.g. that Oswald was a 'communist') Dartmouth professor go out of his way to create the YouTube video and get his name in the papers this month of November ($$$$$$?). Computerized results can be manipulated. I'm not saying that's what happened. But as far as I am aware, the Oswald photos' shadows have not been, have never been, and cannot be duplicated under natural conditions with the sun as the sole light source, as shown, in those damning photos that Oswald told us from the beginning were fakes. Out there has to be one honest scientist who will conduct a serious study using a real person, the real sun, and an accurate representation of time, place, background details, and true solar position. It hasn't been done: government-sponsored investigators danced all around the problem. Is the professor's story yet another dance, full of sound and fury, signifying nothing?