Hacker News new | past | comments | ask | show | jobs | submit login
Unbelievable (hackerfactor.com)
226 points by Garbage on May 13, 2013 | hide | past | favorite | 107 comments

I think it is slightly unfortunate that the topic of the photo triggering this debate is a controversial political topic with many contradictory truths, claims and counterclaims of distortion and deceit. In this case there seems to be legitimate journalistic photography of the event that should be the focus of any debate about the event itself.

The real issue with this photograph are whether the digital processing applied exceeded that which is appropriate for a photography competition. Specifically there seems to be a suggestion that it may be composite of different photos from the same event although that seemed far from proven to me.

What do people feel are acceptable digital edits to be applied?

1) Global filters/brightness/contrast/sharpening

2) Similar filters applied to local areas to to ensure all wanted areas are properly lit and visible

3) Actually touching up/smudging or otherwise directly manipulating particular pixels.

4) Actually combining different photos (e.g. you have ten photos but someone (different) has their eyes shut in each one so you grab one face and merge it back onto the correct body).

Personally I don't see a single clear line where you should stop so I think that it should probably be spelled out in the competition rules.

Personally I don't see a single clear line where you should stop so I think that it should probably be spelled out in the competition rules.

I think it's fairly easy to describe either side of the line even if the line itself is a little blurry.

When post-production edits of a documentary photograph change the time and location of relevant people and objects, the journalist has some responsibility to disclose those edits to the viewer/judge.

When post-production edits merely enhance the clarity of the image, or the image is clearly not meant to be documentary, there is less need for such disclosure. Even though bias is present at all stages of the process, that includes even shot selection to begin with so presence of bias alone is not an indicator of which side of the line a particular image might be.

Obviously the first step is to determine whether an image is being passed off as a documentary photograph or an art photograph. Almost no one cares whether a photo of Sheryl Sandberg in TIME is heavily doctored (i.imgur.com/lFwMNia.png). Presumably, the photos in the contest were. This was a photographic report of a real and notable event.

Then the question is whether the modified elements of the photograph are relevant to the associated story. In most cases, filters and contrast are not relevant; although they could be in some edge cases. For touch-ups and smudging, the question is what are you touching up and smudging? A minor blemish on someone's face might not be relevant, but replacing a rifle with flowers could be substantial. A composite to eliminate someone's eyes being closed doesn't change the substance, but eliminating someone's presence entirely would be substantial.

Of course, a competition might have even stricter standards but from a consumer standpoint, that's where the line should be. An artistic collage photo meant to concisely tell the story of an event is fine, even if it LOOKS like it's almost real, but the nature of the photo should be disclosed. It shouldn't be passed off as a documentary photo.

> In most cases, filters and contrast are not relevant; although they could be in some edge cases.

This makes me think of the infamous Time magazine cover featuring O.J. Simpson: http://en.wikipedia.org/wiki/File:OJ_Simpson_Newsweek_TIME.p...

> What do people feel are acceptable digital edits to be applied?

It's about how you represent the image. Is it pristine or processed? When applying filters, it is appropriate to admit that filters were applied. If you combine different photos, it is appropriate to admit that multiple photos were combined. Do you really have a newsworthy photo of an athlete winning the race if you digitaly process a photo of from a previous competition to make it look like it is from today? Is it more newsworthy because it's yesterdays images of dead children instead of yesterdays athletes?

> Personally I don't see a single clear line where you should stop so I think that it should probably be spelled out in the competition rules.

Unfortunately, morally corrupt photographers will claim their manipulated images are following the rules, even when they are not. I see similarities with athletes and doping here. It's not allowed by the rules to enhance your performance with drugs, yet many top-level athletes continue do so.

Cheating will not go away but can and must be combatted. News reporting must be resilient against such cheating.

What about artificial lighting?

I think it's equally valid to be purist as it is to accept post-processing as a part of the modern photographers toolbox.

As long as post-processed photos are presented as such, I think that's fine.

Similarly, when an image is composited I think it should be presented as a composite, otherwise it really is like telling a lie.

The most striking thing about the Gaza photo is that it looks fake or staged. It's weird that this won a competition because it looks like someone set the whole thing up and positioned lighting to get the effect they wanted. This is particularly true of the guy on the left who looks artificially lit.

Congrats, you've done what most top HN comments do, contribute nothing but uninformed speculations and hearsay.

I think you misunderstood the point of te article, it's not trying to prove that the scene was fake nor staged, but the photo was digitally edited and not worthy of the award.

Digital editing aside, the scene seems to be quite real as it was covered by multiple news outlets and with different photos. Here is the link from the article to the time's photo (by the Reuters) : http://world.time.com/2012/11/15/a-new-gaza-war-israel-and-p...

You are reading things I didn't say (but thank you for replying otherwise I might not have realized that people would read my comment as me implying the event was faked).

I'm not in any way claiming that the scene was fake or staged. I was expressing an opinion about how it 'looked to me'. i.e. when I look at this photograph the colours and lighting remind me of pictures that have been taken in a studio or with artificial lighting. I was making that point in the context of this winning a competition for press photography where you would not expect a photo to have the look of something seen in a studio.

That is all I was saying.

If I wanted to make the claim that something was faked I would say so.

I am claiming nothing about the event, just the look of this individual photograph and its merits independent of the event it depicts.

I also feel this photo (not the event) do not seem natural. I don't see why this opinion should be shameful in any way. It has nothing to do with the event itself or the poor people in the scene, it has everything to do with a photographer, his tools and his deontology (work ethic).

The implication that the methods of analysis used in the article have similar or less predictive power than your intuition is dismissive, trivializes the hard work of others, and is probably false to boot (look for false positives by google image searching "HDR").

You did not make this implication but the parent post we are discussing certainly did.

To be fair, the implication was probably not made maliciously, and a more appropriate response would have been "This got past the pros because such lighting is possible to achieve in outdoor photography, although it requires a level of skill, equipment, luck, post-processing, and aesthetic risk that make it uncommon outside studio settings."

Yeah, this "plastic" ("airbrush"?) HDR like look can be achieved regardless of source material.

Though the fact that someone could be this... blind or heartless, I'm not sure what it is... to think this is a good "look" for this source material, might warrant a post on DehumanizationNews or something - while PoliticalNews commenters could point out that this hack job essentially turned a bunch of grieving people into an "advancing mob", and that the photographer is a total asshole for that etc.

Personally, today I learned one new thing, how to figure out the position of an unseen light source via hard shadows. Awesome!

Of course it was digitally edited. Since the advent of the digital photography, the very act of taking a photograph (i.e. digitalised version of it) is editing of sorts.

I think you're confused. An original digital photograph is no more intrinsically "edited" than a photograph taken on film. Would you say a film photo is edited just because of the limits of the film grain? Because the limitations of a digital sensor's pixels are really no different.

So no. Not all digital photos can accurately be said to be edited; there is such a thing as an original digital file, and a distinction to be drawn between that file and an edited version of the original.

You are the one confused. Do you understand the difference between analog and digital? Analog is a direct imprint of the photons that pass through the lense of the camera and have a chemical reaction with the film (simply put). A digital version is a COPY of the image that passes through the lens in the digital camera and by way of that it is already one time manipulated. In any case, I'm not saying I am the only one right here but that it's possible to look at it from different points of views. My overarching goal was to show that the whole claim of the edited photo not deserving a prize was irrelevant - photos are edited daily, every time and that does not take away from their aesthetic, documentary or other value if they are pulled off greatly. Short version: yes, even looking from your angle, the writer could have said the same about ANY photo, analog or digital, arguing that because the light in the photo looked different than what it would look in reality (and don't forget that a human eye is a sort of an imperfect camera itself) it isn't worthy of recognition.

This means nothing. There must be a limit to what one is allowed to edit on a photo, or you would accept the Soviet or Mao era propaganda edits?

I'm not getting into any propaganda issues here as that is proven. The writer only pointed out that MAYBE some lighting has been altered in the photo that won the award. So what. Maybe it was even cropped a little. Gasp.

Meant to say "unproven", heat of the moment typo :)

"The most striking thing about the Gaza photo is that it" is real, those are in fact dead children, digital editing notwithstanding.

There are dead children all over the world, and plenty of pictures of them. A photograph is notable not because it shows dead children, but because it shows an emotionally poignant scene that reverberates with its audience.

If the photo's made up, it's no better than a poster. It can be drawn with skill and make the same point, but it is not a photograph and so does not deserve a photography award.

The emotionally poignant scene draws its weight from the fact that its showing dead children, not the photographer's technique.

The dead children just make it worse ethically. There are thousands ways to make a political point, the ones using dead bodies or suffering kids should be out of question. It is despicable.

This argument makes absolutely no sense. The 'political point' here is that children are being killed. If you can't use the dead bodies of children to illustrate this, then what can you use? Digitial manipulation is irrelevant to the fact that this was a real funeral with real dead children.

A different point is that we are so numb to images of dead children that those photos need to be manipulated so we take attention.

That's pretty chilling.

Whenever I see photos of suffering children in a political context for me it is the "look at those poor children, clearly the [opposing group] are poor evil" argument. Which is not about the children anymore, it's about diverting your attention from the cause of the problem. And this argument is used over and over on any occasion. How can you not become numb to that?

It's not much of a stretch to say anyone killing children is (if not evil) in the wrong.

My argument is that there are some limits to what can be exposed to the public. You won't show on tv pictures of abused children to fight against children abuse. As a father, each time I see dead children I can't stop imagining it could be my kids, it hurts deeply, striking very intimate chords. I don't think it is ethical to strike these chords. These kind of pictures can be used as proofs if needed, but should not be public.

It's not ethical to encourage you to feel empathy? OK. You're entitled to your opinion but it's stupid.

You seem to have things backwards.

It is despicable to kill children.

It is not despicable to point out that they are dead.

What's more despicable about killing children than about killing people of other ages? Yet it's the children that are frequently used for political arguments. And that is disgusting.


What's more despicable about killing children than about killing people of other ages?

Perhaps the fact that they can't defend themselves? Or that you can't say that you just made the world better eliminating some dangerous terrorists? Or the fact that they had their whole life to live? You only have to choose.

Medical (and other) interventions that are effective on children are judged as more valuable than those that are effective on the extremely elderly due to the concept of "quality adjusted life years" (QALYs)[1]. Of course, people's emotional reaction to children being killed is merely based on evolutionary adaptedness, and has nothing to do with whether its morally justified for them to feel a particular level of outrage.

1: https://en.wikipedia.org/wiki/Quality-adjusted_life_year

Frankly, looking at the other fotos I find it quite hard to believe those children are even dead.

Firstly, muslims would never show images of dead people, but brushing that aside (and that would require explaining why hamas would have no problem with showing a few palestinian atheists world-wide. Also would those people behaving against islam's principles be in a large angry mob on the streets of Gaza ? I mean, that is avoided by most pious muslims in Gaza I hear).

Where are the wounds ? A bombardment kills by using shrapnel - lots of it - as bullets, or by crushing. Obviously these children haven't been crushed. The right child's face is full of sand, but the white burial clots is immaculate. They are lying flat ... euhm ... how did that happen ? (In case you don't know there's a reason they call a corpse "a stiff"). If they died through bombardment they wouldn't have been lying, they would have been curled up. Since nobody collects dead bodies immediately after a bombardment (would have to be in 2 hours or less), they would experience rigor mortis in the curled up position. At that point the only way to get them into a lying position would be to break all of their bones, which takes time and a lot of effort. The bodies in the picture are obviously not experiencing rigor mortis, they are flexible and move with the arms of those men ... so what is it ?

Either these kids died in the last 60 minutes before the picture was taken (where's the blood ? I hear even muslims have blood. And ahem, all muscles relax, which would color such an immaculate burial cloth yellow and brown for reasons I won't go into, again where are those discolorations ?). Furthermore the eyes wouldn't stay closed. Or they're past rigor mortis, dead for at least 3 days ... so why is there skin wet ? Why is there no loss of skin tone ? And frankly a corpse 3 days old smells to the point that no emotion will allow you to ignore that, and the smell is easily detected from 10 meters away in open air (no matter how exiting a medical student tells you it is, never open a drawer with a corpse, just don't, you'll still smell it constantly a week later).

The easy explanation for all of this is very, very simple : those kids are alive (they even look like they're holding their eyes closed, although that could be an impression). Lying about it would be perfectly in line with islamic practice (probably important for large angry mobs in gaze, don't you think ?), it would explain the lack of wounds, lack of blood/other bodily fluids, the eyes, the skin tone, the flexibility of the children, everything.

That would of course mean that it is a complete lie, and the photographer is very likely complicit.

>Firstly, muslims would never show images of dead people

What BS is this? I follow world news for 20+ years. There have been countless, literally tens of thousands, pictures of dead people in muslim countries.

>Frankly, looking at the other fotos I find it quite hard to believe those children are even dead. (...) where's the blood? I hear even muslims have blood.

Really? Really?

And it goes downhill from there, with your explanations of why they are not really dead, of the "I've seen a lot of CSI on TV" quality.

Spoken like a true ignorant that has never gone 5 minutes away from his hometown in rural Idaho (or TV-land, which is the same).

All the BS about rigor mortis etc, are not even TV-doctor quality, but what is especially disturbing is the lack of humanity and empathy.

I've heard similar points made before and while I can concede that they are at least somewhat coherent - or "not impossible" - they are almost always riddled with bad grammar and every other sentence is an affront to any sense of compassion.

I'm fine with people using Occams Razor even if the conclusion that brings about is uncomfortable for me. But I'm not OK if the argument made lacks basic human dignity.

> (where's the blood ? I hear even muslims have blood. And ahem, all muscles relax, which would color such an immaculate burial cloth yellow and brown for reasons I won't go into, again where are those discolorations ?)

Yes, they dirtied and wrapped a living child in a cloth to stage a dead child and forgot to simulate piss and shit stains necessary to pass your analysis. Huge slipup on their part. Not possible that a parent or relative would first take care to ensure the dignity of their child to remain intact before bringing them out to show their grief in front of their community.

> And frankly a corpse 3 days old smells to the point that no emotion will allow you to ignore that

You might be surprised.

There are stories around of mothers tending to their stillborn fetus for weeks. I think one even made a blog about it.

I suppose we can conclude that you are falling victim to the limits of your own experience and imagination. Otherwise known as ignorance.

It looks like a digital render to me, it does seem bizarre that it would win.

Context is everything. When presented on a webpage in the context of debunking fake photos, it's strikingly artificial. Whether we'd have the same impression for a printed photo submitted to a competition among many other photos is much harder to tell. One is likely to conclude it's just a very well taken photo that was aiming for a surreal feel.

Yeah that's the Dave Hill HDR effect, it was fashionable a few years ago. Normally it's done by composing several photos with different lighting settings, and then spending a few hours cleaning textures and saturating the hell out of everything in Photoshop

>This is particularly true of the guy on the left who looks artificially lit.

Ever heard of dodge and burn, in Photoshop?

Also try "staging" an angry group of people, protesting and lamenting their dead children and lost relatives after yet another raid.

Would anybody have the gal to say the same things for the Boston victims?

Please see my comment above: https://news.ycombinator.com/item?id=5698353

I am in no way claiming that this event did not occur or that the scene was staged, only that the photograph looks staged and in my opinion isn't a good press photo and therefore shouldn't have won a competition for press photos. I am claiming nothing about the event, just the look of this individual photograph and its merits independent of the event it depicts.

I think you missed the point. jgc is not saying he believes the photograph has been staged, but that it looks like it has been. In other words, how did a photo that is so obviously fake win a photography contest?

Is there a site to compare the other entrants?

A bunch of photographs today "look faked" - sometimes that's great (and lucky) lighting, or gentle tweaking of levels, or outright fakery. And the fakey look is what makes it a remarkable photo. Those faces really pop out.

I'm surprised it won if they didn't get to see the RAW image. But maybe I misread the article.

>I think you missed the point. jgc is not saying he believes the photograph has been staged, but that it looks like it has been. In other words, how did a photo that is so obviously fake win a photography contest?

I'm confused. Your "in other words" conflict with your previous statement. (1) "jgc is not saying he believes the photograph has been staged, but that it looks like it has been". (2) "In other words, how did a photo that is so obviously fake win a photography contest?"

So, is he saying that it is fake (and "obviously" so), or that it merely "looks fake"?

And, may I ask, what is that "obviously fake" about it, besides color adjustments?

Now, it's true that color adjustment is frowned upon in certain news circles, but it's used a lot in magazine reporting, and it's not that different than what you could achieve with your choice of camera (or film in a previous era). That kind of alteration might, or might not, be acceptable for a photo competition.

As for "obviously fake" as in "he staged it by putting the people in a specific position", "had lights" etc, I don't see any signs about that -- or even find it that possible given the circumstances. Those are not some "damn arabs, doing PR work and pretending to be sad and enraged". Those are parents, neighbors and friends, holding their dead children. The way some people commented here, they might as well been subhumans in the picture.

For the composition part, if a photographer is half competent, he can get a very good composition out of such as scene without even trying. And he's not some tourist with his point and shoot.

(Disclaimer: I've done some professional photography back in the day, but mostly travel work and interviews / portraits, not news).

The picture looks unnatural. Whether this was achieved with props or post-processing is not relevant - what is relevant is that it won a competition for realistic photos, despite not exhibiting realism.

It wasn't a competition for "realistic photos" it was a competition for real photos. Photojournalistic work. The difference is important. A photo can look as "unrealistic" as can be, as long as it's real, it's OK.

So a picture doesn't have to "exhibit realism" to win such a competition. It just has to be real. If it looks uncanny, all the better.

The line is draw at actual image manipulation. Compositing people in, changing a face, removing elements etc.

As for color adjustments and light fixes, those are sometimes frowned upon, other time totally accepted in journalistic work. Anytime a photo editor gets a jpeg from the camera RAW, some work of that kind has already been done. Including the camera letting you pick several options, from sharpening to color balance (and more traditional stuff like bokeh, depth of field and shutter speed, each having it's own artistic effect. Does a 1 sec photograph of a running river look "realistic"?)

I've seen pictures made on an iPhone with stuff like Instagram in Afganistan/Iraq by photojournalists that won many awards, despite looking totally unrealistic with regards to neutral tone/color coming out of a camera. E.g:


"Now, you might be thinking "why are there two conversions on two dates"? This is what you typically see when a picture is spliced from two sources."

"The January 4th date is interesting because the deadline for submissions to the World Photo Awards was 17 January 2013. So although the base photo was taken in November 2012, it wasn't edited until about two weeks before the contest deadline."

This whole theory pretty much fails as it was published in "Dagens Nyheter" in November 2012. The lower image is what it looked like:


Here it is on the front page of the newspaper. 21 November 2012.


http://i.imgur.com/aKTAc7d.jpg (Image only mirror)

The photo that won the award (was physically available to the judges) may have been significantly different from the photo that was originally published.

Indeed, I can see several differences. The colors, obviously, but also the child in the foreground is dirtier, the men at the edges are darker, the men at the center are lighter, and all faces are clearer, and the sky is darker.

These children are dead, there is nothing fake about that, and there is also nothing fake in that they are victims of a variety of factors, including Paul Hansen who manipulated their funeral images to cheat his way to an award. There were over 100,000 photos up in the competition, and prizes were awarded to photographers of 33 nationalities (with a notable absence of Israelis). Other photos bringing other stories (and maybe even the same story) from Gaza and elsewhere in the world could have won instead. The whole point of a photography contest is to identify and recognize those photographers who can skillfully identify and catch dramatic moments and angles in split second, not those who take otherwise normal pictures and edit them later.

To give some balance, the day these two brothers and their father died, there were continuous rocket attacks on Israel. At around 7pm, a rocket hit a high school in Ashkelon. A student from the school told the reporter, "We were playing soccer near the school and suddenly there were sirens, and we saw the Iron Dome rocket fire above us, and then we heard a large explosion and we realized it fell just near us. The police came and we directed them." The rocket fell on the rooftop of a nearby school building unexploded and was disarmed by the police. The IDF later released a statement saying it attacked over 50 tunnels and tens of hidden rocket launchers. The Hamas often places the rocket launchers near civilian homes for protection and this might have been the cause of the attack. The civilians, as perhaps the Hejazi family in this case too, have no real say and are sometimes forced to remain near the weapons against their will.

[1] http://www.worldpressphoto.org/content/swedish-photographer-...

[2] http://news.walla.co.il/?w=/2689/2587705 (live blog from the day of the attack, Hebrew)

To my eye, the Gaza photo appears to be illuminated by an off camera flash above and to the left of the camera. The man at the front on the left has a very bright left side of his face, and a very dark right side (also notice the shadow and light around his ear). Everything else in this photo seems consistent with this.

The reason for there appearing to be more dirt on the child's face is simply because it's a higher contrast image.

Maybe it has been manipulated more than is allowed, but I'm not seeing it.

EDIT: I'm not sure that there's only one artificial light source — notice how the front left man's visible ear is red? That's because it's being brightly illuminated from behind. Also notice that the top of his back (his left — photo right — shoulder) is illuminated, and that would be in shadow if the photo was only being lit by the flash that is lighting the left (his right) side of his face. Finally, there's another light source (maybe the sun?) that is causing the shine on the top right section of many of the heads (such as the bald man carrying the dead child).

>> , the Gaza photo appears to be illuminated by an off camera flash

It could also be that there were multiple photographers on the scene in various places shooting with strobes simultaneously which created all the extra light.

Take a look at the image linked here: https://news.ycombinator.com/item?id=5698476

The original version is less heavily edited, but shows the same lighting patterns. He just got a great angle with good ambient light, and sunshine reflecting from the wall on the left.

Because of the wide range he must have done a good deal of shadow/highlight recovering; the desaturation and dodge/burn done for the final submission were over the top and made the image look artificial.

The striking photographic detail about the photo is how the faces appear to be artificially lit. The arguments about the sun's position may actually explain that if you consider the possibility that the light on their faces is reflected light from the left wall plus the photographer's off camera flash.

The faces could be the result of some aggressive shadow recovery.

I find that being overaggressive with shadow recovery can result in fakey, HDR-like images where the lighting on the subjects' faces can seem... off.

Fake photos like these are called "terja" in Sweden after the Swedish wildlife photografer of the year was busted in 2011.

Verb: Terja - to manipulate photos Adj: Terjat,Terjade - fake photo,photos

http://terjadefoton.wordpress.com/english/international-medi... and


In college I lived with one of the greatest people I have ever met, and he was from Palestine. My late father was born and raised in Israel, so the roommate and I definitely had our share of conversations about the situation. Just as a note, his dad was a high ranking official in the Palestinian political movement. He once told me, "If you give the people food, water, shelter, and don't steal their opportunity for a better life, there will be no problems." That always stuck with me and I think it is resoundingly true.

When I went to Israel in 2008 for the 20th time in my 26year life, I learned that in Israel the average income is almost $30000, while in every neighboring country the average income is well below $5000, some under $1000. Palestine included. Israel is a completely different world than the countries that surround it. The situation over there is crazy and though it didn't end to well for my dad here(check my posts to figure that one out) I can see why he snuck out of the army 6 months early to get the f out.

When this was being debated on photography blogs, someone dug up the originally published version of Hansen's photo, published just the day after the event. Here's a side-by-side: http://www.flickr.com/photos/gunthert/8485283411/sizes/o/in/... .

While the earlier version could also have been edited (quickly) before publication, there's no doubt that it had much more believable lighting and coloration. For the WPP submission, Hansen desaturated the photo globally, particularly the blue and yellow. He also locally edited the luminance of the left wall to be consistent with the right. The effect of both changes was to isolate the marcher's angry red faces as the primary point of interest.

The image is 100% real in the thing that matters, that that the photographer captured this scene.

The photographer has played with levels, dodge & burn et al, so the aesthetic output is "fake", but those are actual people, holding their actually dead children.

The editing might mean he is not worth the prize.

But that's it -- nothing more to discuss. Not "where is the blood?", "why aren't they curled", "they're holding their eyes closed", etc. I do hope those are just the opinions of one disturbed individual.

Try reading some World News, from world sources (European, African, Arab, etc), outside the US media, and you would see tons of similar photos. Including tons of similar unedited photos from Gaza.

People would have defended more and have shown MORE respect for the moon landing photos.

No, you're missing the point of the article. According to the article the photographer didn't capture the scene. It says he took the people standing with their dead children from a photo and pasted them into a crowded street. E.g. merging pictures. A bit more dramatic than playing with lighting etc.

No, that's not what the article says. It says the image was probably based on more than one picture merged together, but it doesn't claim the specific things you say it does.

The HDR effect usually implies taking the same photo several times with different lighting and merging them all together. I wonder if this is the manipulation he's detecting, or if there's more as the parent suggests

Not necessarily, you can get the HDR with most image editors from a single image nowadays.




You also don't need ELA to see that. Just by looking at the sharp edges of the hair on the guy at the front and to the right, you can clearly see that this has been composited together.

I don't see any obvious such thing (at least on that part) and I've been doing Photoshop work for 10+ years.

Things to keep in mind:

1) Error Level Analysis is extremely unreliable. Especially if an image has had several saves and editing passes (and given that he did it on a small jpg, not the full image).

2) Sharp edges can also come from the sharpen tool, which can be manually applied selectively on different areas (instead of in the whole image).

3) Lighting changing and artifacts can also some from sloppy dodging and burning work. Both artifacts introduced by (2) and (3) also show up on ELA.

4) The time metadata he mentions in the article is useless. For one, the camera time could be totally wrong depending on the timezone settings. When I did some photo work, I always forgot to set mine when changing timezones frequently.

Wait, are you actually arguing that it's an authentic photograph?

Why, are you arguing against it? This is an authentic photograph, there are no two ways about it.

The only alterations are light (dodging burning) and color.

As far as the subject matter, place of scene, people etc, it's all genuine.

I don't buy his argument about inconsistent left-right lighting in the Gaza photo.

Assuming the sun is really top-right, it is actually expected that people's faces are illuminated from the left, given the setup: the people are between two walls, the left one being illuminated, the right one in shadow. The light you see on their faces is reflected from the left wall.

The light isn't from sunlight reflecting off the wall. It's too... directional? It's also too bright, and the colour might be too warm. I'm not entirely sure how to explain it, but I've done enough photography to recognise that as an artificial light source.

But using an off-camera flash is not the same as Photoshopping three photos into one, and the OP not even mentioning a flash as a possibly explanation for the "inconsistent" lighting leaves me with significant doubts.

If it was a known "event" at the time (i.e., all the photojournalists knew the procession was coming and planted themselves in an "ideal" location), and you imagine a TMZ-like throng of photographers with their flashes firing from all angles, the inconsistency of the lighting might make sense.

Note: I don't have an opinion of whether the image is fake or real, but I am not sure the OP's analysis considers all possible lighting situations.

I know little about image forensics, but am just old enough to have done some black and white film development. What would a skillful dodge-and-burn when making the "Eddie Murphy" print look like in an analysis like this? Or any other creative highlighting by the developer? It seems a bit of a stretch to go from "some regions of the photo are different than others" to "it's a fake". I'd be interested in seeing a similar analysis of some other photos known to be "authentic" for comparison.

I put "authentic" in quotes mostly in homage to an excellent series of Errol Morris' essays on nature of documentary truth and history of photographic manipulation. Worth reading. If you like it, you probably should watch some of his films too. http://opinionator.blogs.nytimes.com/2007/09/25/which-came-f...

The differences he's looking at are digital breadcrumbs left behind by editing software. It's not so much an analysis of the photo as it is of the JPEG file. JPEG is lossy, and repeatedly saving the same image changes the data in a predictable way, such that it's possible to identify regions of "newness" in a photo based on how their compression error differs from other parts of the same file.

Could the title here be changed to something a bit more informative?

The argument about peoples faces being illuminated on the wrong side is flawed. The light on their faces is reflected from the wall on the left or hits them from behind. It cannot illuminate the right side of their faces because the building on the right casts a large shadow and because the sun is in peoples back. There has been some serious dodging and burning going on this picture but you cannot conclude from light alone that it is a composite.

That is also my analysis.

The only character that looks odd to me is the one at the bottom left, and even then, it could be genuine.

It's funny how many people here latch onto the "fake" argument so quickly. I read through the whole piece, and the only thing that sounded plausible was the three sources thing in Photoshop. All other arguments I cannot really follow.

Gaza looks like it follows GMT by about 3 hours, which would make this photo taken at about 11:40am solar time. I think that at least roughly fits with the angle.

The thing that really bugs me about ELA is that while there is validity to the technique, and it can identify certain images as clear fakes, any asshole can plug any image into it, point at the bright points near some edges in the image (which will be there regardless of authenticity), and smugly act like they know what they're talking about.

Yeah, I've seen various people use these types of methods to prove a point, and each of them were able to come to their own conclusions based on the results. Ultimately you can manipulate any photo to look like this in the end, and some photos even just naturally look this way when processed even if they are the real deal. I think we're putting too much credence into one-click web-based tools and unreliable Photoshop methodologies.

Exactly my thought, it all looks like someone is just trying too hard.

See: http://www.hackerfactor.com/blog/index.php?/archives/545-Sta...

To me that looks like some disgruntled redditors who can't take the fact that Morgan Freeman is just too tired of PR to scan and type on IAmA himself. So the most plausible explanation that he's just resting on the couch while some agent is typing has to be debunked by means of pseudo-hackish image forensics which don't prove anything after all.

Exactly. I don't trust this guys analysis. He is unpersuasive.

I'm not going to express an opinion about the photo other than general surprise about how varied a concept of 'faked' appears across all commentators.

What really interested me was the article's description of Photoshop's XMP metadata block, giving a history of saves/exports/edits. I never knew this level of detail was stored with the images. It also makes me think, how long until there are tools to easily edit such metadata? The website takes the metadata as gospel truth. At some point in the future, this is not going to be a reliable position to take.

It is not hard to edit the metadata. I think the current reference point for picking out forgeries is a RAW file - not that they are impossible to manipulate either but it is harder, plus the kind of statistical analyses done in the article would be more meaningful on an image that is supposedly completely unedited.

Well, to be honest, they really just use it as a starting point. I'm sure he wouldn't have just concluded it was totally genuine if he had seen a clean metadata block. Its a good starting point, but by far not the "gospel truth."

I've no problem with how they used it in the article. At the moment, the overlap of 'professional photographers' and 'people skilled in metadata forgery' is pretty slim and (IMO) the data is unlikely to have been altered in this case.

My point is, in the future, you won't be able to even use the metadata as a starting point because all of it could have been changed with easily-available and well-known tools.

It 503s for me and I can't find it in Google's cache. Any mirror?

Coral Cache never fails


Edit: It's failing.

Failed for me.

It failed.

Try refreshing/reopening a couple more times. Worked for me.

Yeah, I did. That persistence eventually paid off. The photo looked faked just to my eye. I could see the lighting was off without needing proof.

EDIT to add: Don't know why this appeared twice. Never happened before.

403, and the offered alternative location is not available either. Is this another salvo in the Middle East propaganda war? Please enlighten.

World Press Photo Award-winning picture is found to be significantly retouched, "cleaning" dirt from faces of dead kids and adding light and shadows in Photoshop. Picture happens to be of Gaza men mourning dead children in their arms, likely as result of some Israeli military action.

I believe the main argument is entirely about photography: clearly composition was great on its own, but lighting was heavily edited; when is a picture still a picture, so to speak? It's an important question for photojournalists and their awards.

Whether this overlaps with the customary Israeli pushback on this sort of material, I honestly don't know.

Why no, the point of the article is that this is not one photo, it is a collage.

The recent history of conflict photography in this region is worth digging into. I believe there was an HN discussion of the following documentary, but I can't find it:


"an auto-critical photo essay showing the paradoxes of conflict-image production and considering the role of the photographer in the events"

Warning, graphic pictures of dead children in this post. Not safe for mind or work. It is horrifying. Posts like these are why I stopped using reddit, because "surprise! Here's a horrible picture, now feel like shit for the rest of the week/month/forever!"

If there were some place else to get programming or startup news like you can get here I would never come back. I wish there were such a place or that HN stopped allowing politics and NSFL stuff like this.

Now the link title and site don't give you any useful information about the topic it is fair but the photo is not a graphic one of the children. I fear that if you sanitise your life to the extent of avoiding such photos you risk a severe shock at some point and possibly in real life.

The graphic part is the visible emotions of those in the crowd.

Looking at your posting history you submitted a before and after [edit: Google Earth satellite] photo of the earthquake in Japan 2011. To me that is far more shocking as I can't help but imagine briefly at the impact on those there, hundreds of children lost and others orphaned.

Yes, that wasn't a good submission by me. I aim to do better in the future.

I wasn't really complaining about that but more commenting on the range of what people feel shocking. The photo in this case was not in my view problematic.

Death, pain and suffering happen all the time and while it shouldn't be fetishised it also doesn't seem a great idea to hide or deny it either.

"Not safe for [my] mind or [my] work. It is horrifying [to my particular sensibilities]."

Fixed. Not every reader of HN shares your specific sensitivities.

How could they not see that this was fake/manipulated?

getting 403 error :(

It's so unbelievable, even the server can't handle it!

Try it a couple of times and it will eventually work. Really fascinating analysis.

Oh, the side is blocked in Russia

The the site is blocked in Russia? Can anyone else confirm? It may just be poor connectivity. Hackerfactor would be an interesting addition to the known Russian blacklisted sites[1] though.

[1] http://en.wikipedia.org/wiki/Russian_Internet_blacklist#Know...

So the photo someone took was later edited in order to look better? Shocking.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact