Sunday, January 15, 2017

Faked Photos and photo-analysis, continued

--Leslie <;)))><

Going on with "Abu Ghraib Abuse Photos" at --

The second picture from the top appears much more graphic.  On a gray concrete floor with a tan wall to the left and a large wooden box to the rear right, a naked man with his face clearly visible lies stretched on the floor, grimacing, apparently in pain.  He has thick dark hair on his head and chest, and a close-shaven beard.  His crotch is coyly blurred so as to completely conceal his genitals;  this is such a common media convention that the average reader wouldn't stop to think that this shows that the picture is at least partially photo-shopped.  

The man's arms are behind his back, and the one we can see is held -- with a splayed-finger grip -- by a man standing bent over him.  This man wearing tan military-style boots, uniform pants, a black jacket and a helmet covered with a white cloth cover.  We can't see the right arms of either man, but the position of the visible arms -- and that splayed-finger grip -- suggests that the naked man's wrists are cuffed behind him.  The naked man's left leg is held at the foot, just below the ankle, by a third man -- of whom we can see only his left hand, with a wedding ring, and arm, clad in a white sleeve.

The naked man's left leg is smeared with red fluid which appears to be blood, and wrapped with a strip of white cloth tied in place with a strip of black plastic, just above the knee.  His right leg, not wrapped with anything, is banded with a symmetrical strip of matching-colored red fluid that expands into a shapeless smear on the side near the floor -- as if the blood had spilled evenly over his right leg and pooled under it while the man was lying on his right side.

In the foreground are some scraps of paper and black plastic, and smears of what looks like more blood -- but in different amounts, shapes and colors.  The largest, darkest, and closest smear, with a piece of paper lying in it, has an interesting shape: the leading edge is somewhat rounded and amorphous, but the rear edges are straight lines, meeting at a distinct angle -- which any fan of CSI would recognize as a "void".  If the blood came from the naked man, then it could only have spilled out of his left leg while he was lying on his right side, with his right leg bent.  This is also how that symmetrical stripe of blood came to spill over his right leg and pool under it.  Also, he lay in that position long enough for that amount of blood to run out of him.

So how did he come to be lying a couple feet behind that spot, and where did that small, thin, slightly paler smear between them come from?  The only possibility is that his legs at least were picked up and carried backward for a few inches, but then the side of his right leg was dragged along the floor the rest of the way to his current position.

Further details: there is a thin, faint smear of red on the side of the wooden box behind the naked man.  To the left, behind the uniformed man, extending several feet back, the floor is stained with a pool of mottled watery pink and green patches, the same green as the paint along the baseboard of the wall. This could be water reflecting the paint and thinning out more of the red fluid;  if the red fluid is blood, then it was spilled at a different time from the blood in the foreground.  Standing in the pool, a few feet behind the uniformed man, are what appear to be three mismatched bottles against the wall.

Now, given the example of the modestly blurred crotch, we must first ask "Is it real, or is it PhotoShop?"

Evidence for fakery includes the color of that red fluid.  Anyone who has studied Biology, or worked in the medical business, can tell you that the color is subtly wrong.  Oxygen-loaded arterial blood is a bright scarlet, blood returning in the veins is the purple-brown of prune juice, and spilled blood drying in the air turns steadily brown.  At no point is it the cranberry-juice purplish red of the fluid in that picture.

On the other hand, given the state of digital cameras -- particularly phone-cams -- ten years ago, it's reasonable that the coloring could be off.  Looking at the rest of the picture we see that, yes, the color contrasts are a little too garish, and with a slightly blue tone.  Yes, the color of those blood-smears is realistic.  The clutter on the floor is too random to have been posed.  Also the lighting is naturalistic, coming from the direction of the camera, and the resolution is very tight and clear.  We can assume that this picture is real.  So what does it really show?

The caption reads (emphasis mine): "An Iraqi detainee appears to be restrained after having suffered injuries to both legs at Abu Ghraib.  It is unclear whether his injuries are from dog bites."

There's an obvious fallacy in these statements from the start;  only one of his legs is bandaged.  The blood on his right leg clearly spilled on it, and wasn't shed by it, as can be deduced from the pattern of the bloodstains. There is only one injury: to his left leg.  The amount of blood spilled -- not counting the amount in the earlier, watery stain further up the hall -- indicates that the injury severed a good-sized (therefore deep) artery, even though it's small enough to be covered by that single narrow bandage.  A single deep, small, narrow wound is not characteristic of a dog bite but of a gunshot, and there is no dog in evidence.  Whoever wrote that caption was, at best, a poor observer.

From the evidence we see it's more likely that someone else produced the earlier, hosed-down bloodstain further down the hall.  The prisoner we see was shot in the leg and left to bleed for a few minutes -- why? -- then picked up for a moment while his leg was quickly pressure-bandaged to stop the bleeding.  Before the wound could be washed or treated any further, the prisoner was quickly set down again and dragged backward across the floor.  At some point his arms were pulled behind him and cuffed, and his uninjured leg was seized.

At this point it helps to review the army's and the Red Cross' investigative reports, which tell an interesting tale.  It seems that well before the scandalous Abu Ghraib pictures were "leaked", there was an attempted prison-break.  Local imams, allowed in to provide "spiritual comfort" to the prisoners, had sneaked in a few loaded handguns.  The prisoners, with or without clothes, had shot at the guards and made a break for the doors.  The guards quickly put down the attempt, usually by shooting the prisoners in the legs rather than kill them, but a few of the prisoners died anyway -- usually of bleeding out before medical help could reach them.  Once the fighting was over, the army medics and assisting guards did their best to treat the prisoners' wounds -- although some of the prisoners resisted, preferring to die as holy martyrs, which obliged the guards to restrain them for treatment.  Even the Red Cross couldn't fault the guards' behavior.

The lack of clothes was part of the prison intake procedure.  When first brought to Abu Ghraib the prisoners were stripped naked and put in cells that were bare of anything but toilets, with the lights turned off.  They were left in that state for the first 24 hours, and if they behaved themselves -- no screaming curses for hours, no throwing food or feces or urine, etc. -- then the next day the lights would be turned on.  Another day of good behavior and they'd get bedding.  It took at least three days to earn clothes, and those clothes were chosen by the staff.  This means that any prisoner who appears naked in any Abu Ghraib picture had consistently misbehaved.

From all this we can conclude that the prisoner shown in this picture was a bad actor who had participated in the jail-break and been shot in the leg.  From the evidence we can also conclude that he resisted medical treatment and had to be restrained -- and dragged away from the blood-smear -- to be treated.  More evidence for this conclusion appears in the next picture, the third down from the top.

In this picture we see a medium close-up of the lower body of an otherwise-naked man, who looks very much like the prisoner in the previous picture, lying on his back with an orange cloth spread over his crotch,  his arms behind his back underneath him, and a small but blood-smeared wound on his left leg a little above the knee.  At the top of the picture we see booted feet and a lower left leg, clad in the same pale-tan military boots and camouflage pants that we saw in the previous picture.  One of those booted feet is planted on the nearly-naked man's chest, firmly holding him still on the floor.  It's reasonable to assume that these are the prisoner and guard we saw in the previous picture.

In the foreground crouch a man, to the left, and a woman, to the right.  Both of them are wearing knitted black watch-caps that cover their hair, and rubber surgical gloves: the man's blue, the woman's white.  The man is also wearing a camouflage-patterned shirt with epaulet straps, and the woman is wearing a brown civilian shirt and a tan military flak-vest.  The man is holding a thin thread in his left hand and pulling it taut, while his right hand is pressed to the wound on the nearly-naked man's left leg.  The woman, who has a black symmetrical tattoo partly visible on her right wrist, is holding a pair of fine medical pliers of the sort called a hemostat in her right hand, at an angle which indicates she is pulling something taut, something too slender to be seen by the camera.  The woman is smiling at the camera and holding her left hand in a "thumbs up" sign of success.  The wound on the prisoner's leg is a straight-sided gash with a round hole in the center, still oozing blood but much cleaner than in the previous picture.  It looks exactly like a gunshot wound, from which the bullet has been extracted, being sutured closed.  If the patient is the same man as in the previous picture, then the woman's "thumbs up" indicates that she and the other medic have successfully closed up the torn artery that caused the large blood-smears in that picture.  The resolution of the picture is sharp and clear, and the shadows show that the light is coming from above and right.

The caption for this picture reads (emphasis mine): "A US soldier gives the 'thumbs up' sign as she appears to be stitching up a prisoner's leg wound.  It is unclear whether the injury was from a dog bite."

This is an obviously biased notation, since there is far more evidence that the man and woman are stitching up a deep wound than there is for any dog bite.  The author(s) of these captions appear to be obsessed by dog bites for which there is no evidence.

The Red Cross investigators reported that prisoners "vigorously" complained to them about being bitten by guard dogs, and showed the investigators what they claimed were dog bites.  On observation, the investigators found that the bites had been made by human teeth, in fact the prisoners' own teeth.  Other wounds which the prisoners likewise displayed proved to be made by human tools.  Why, then, did the media editors who supplied the captions on these pictures obsess on dog bites?  No one else -- except the prisoners, their sympathizers and their fellow Jihadists -- did.

So why were these two photos taken in the first place?  Clearly to document that the military guards did indeed make great efforts to save the lives and treat the wounds of the prisoners, even the bad actors who had participated in the prison-break, regardless of how hard the wounded tried to refuse treatment.  Note particularly the lighting and resolution of these pictures, which were intended to be seen clearly -- perhaps used as evidence in a legal investigation.  Compare these technical details with those of pictures which will be shown later in this series.  Also note the faces that we've managed to see clearly so far;  we'll be seeing them again as the illustrated story unfolds.  And bear in mind that the prison-break occurred several weeks before the pictures of the "prisoner abuse" were leaked to the media.

There's more of the story to come.  Stay tuned!    




Monday, January 2, 2017

Fake News, Faked Photos, and how to identify them

--Leslie <;)))><  

Professional news reporters, to say nothing of editors, are supposed to check out their stories: examine their sources, verify their facts, before publishing -- let alone editorializing.  Editors, at least, are supposed to do the same with photographs.  Journalism professionals are supposed to have the skills and the training to do just those things.  Seeing how many of these professionals have gotten sloppy on the job lately, I think it's a good idea if more people learn more critical thinking and verification skills for themselves.

As a training exercise, let's look at a famous scandal.  Google search "Abu Ghraib Photos", carefully set aside the initial prejudices created by the media, and view with a critical eye.  ...Hmm, it would also help to do one's homework;  read the investigative reports on Abu Ghraib done by the US Army and the Red Cross.  A little knowledge of 2006-level digital photographic technology would also be useful.  Okay.  Ready?  Hit the search button and let's go.

The first site, "Images for Abu Ghraib", features a chaotic mix of photos ostensibly taken at Abu Ghraib prison, political cartoons, photos clearly PhotoShopped, photos purported to be from the brig at Guantanamo Bay, pictures of political protests, and portraits from news reports -- a blizzard of them in no particular order.  The only unifying narrative is outrage over US military abuse of prisoners, and the only verification is the citations of where the pictures were published.  Since many of these are re-posed and/or PhotoShopped versions of each other, they clearly can't be trusted for much accurate information.  Let's move on.

The next entry is "Abu Ghraib Abuse Photos - by news -", which labels its editorial position clearly, cites the Washington Post as at least one of its sources, and shows that it would necessarily exercise some editorial caution.  Going to the site, we find the photos dated to February 17, 2006 and updated to June 11, which gives us a reference point.  The headline reads: "The Washington Post has released new photos along with new information about the use of dogs on prisoners."  This phrase neatly shapes our expectations while giving precious little information.  Think: just where did the WP get those "new photos" and "new information"?  How are they verified?  We know what that phrase makes us expect, but just how are the dogs used on prisoners?  Let's take a clear-headed look at the first photo.

The picture shows a wide, plain, concrete corridor lined with multiple plain metal doors, one partly open and the rest shut.  There appear to be bundles of cloth part way down the corridor and squarish light-sources above three of the doors and the end wall, but the picture is so grainy and the resolution so coarse that we can't see any detail to be sure.  There are three men positioned in the middle of the corridor, the two nearer wearing army-style boots, desert-cammo pants and dark jackets, one sleeveless, and what seem to be knit caps -- but again, the picture is so grainy that we can't be sure about the right-hand man.  The man on the right has his hands in his pockets, and the man on the left is holding a dog by a short leash.  The dog is black and looks somewhat like a Alsatian;  its nose is pointed toward the third man, its ears are up, and it looks curious or eager.  The third man, positioned between the other two and just under ten feet further down the corridor, is crouched over with his hands raised defensively, looking toward the dog.  He appears to be either naked or wrapped in clear plastic;  again, the picture is so grainy that the viewer can't tell.  There are no injuries visible on the third man.  Although there are no windows, the corridor is surprisingly well lighted -- yet the shadows are very soft, vague, and non-directional.  That's what we see.

The caption under the photo reads, cautiously (emphasis mine): "An unmuzzled dog appears to be used to frighten a detainee at Abu Ghraib prison in Iraq.  Two military dog handlers told investigators that intelligence personnel ordered them to use dogs to intimidate  prisoners."  Obviously the viewer is supposed to conclude that the dog is threatening to bite the prisoner -- but what are we really seeing?

The devil is in the details -- such as, why is the picture so very grainy, or so well lit, and exactly what direction is the light coming from that casts those muzzy shadows?

According to an old friend who used to do professional photography, that graininess is consistent with digital cameras -- including phone-cameras -- of ten years ago, especially if used with an integral flash-device.  The flash had to come from the camera, but in that case the shadows would have been more clearly directional and sharply defined -- unless washed out by those over-the-door ceiling lights.  If those lights were bright enough to wash out the camera-flash, there would have been no need for the flash in the first place -- unless the whole point was to create that extreme graininess, the bad resolution that ruins the details in the picture, and incidentally makes the face of the prisoner unidentifiable.  The faces of the other two men are concealed.

By the way, Mohammed hated dogs, especially black dogs, so good Muslims are supposed to regard them as "unclean", only a little less than pigs.  A pious fundamentalist Muslim would fear any contact with the animal, not necessarily bites.  Of course, the reason Mohammed hated dogs, especially black ones, is that he made a lot of his early fortune leading a robber-band to attack caravans, particularly at night.  If the caravan included dogs, the animals would smell Mohammed and friends sneaking up and sound the alarm.  Even if Mohammed and cronies could shoot the dogs with arrows to keep them quiet, black dogs would be hard to spot in the dark.  This paints a less-than-virtuous picture of Mohammed, but the pious can always come up with an excuse for their hero.

More details later;  there are plenty of other pictures to examine.