More photojournalism SNAFU?

This image by Anis Mili is coming under fire for possibly being a fake.

It was first run by Reuters, and then pulled when speculation began that this one might not be on the up and up. Mili has spent a lot of time in Libya and in other conflicts, and has quite a portfolio of impressive wartime images.

But I have to admit, this one looks Photoshopped to me. The blur around the edges of that rocket explosion just doesn't look right. There have been a number of very public controversies over phony photography lately. Several in photojournalism circles.

What do you think? Fake or not?

The focus point is the real problem. The concussive wave from the launch is nearly ALWAYS out of focus simply because the "photographer" has focused on the back of the weapon. The launch wave is about 2.5 to 3 feet closer to the camera. It should be out of focus. THIS shot is in very clear focus on the launch flame. This smells to high heaven.

BWTHDIK

Pete

Fake he is leaning on the crutch and the shadow does not show any horizontal item being held. Among other things, he is also standing way too straight forward to fire a weapon like this. In that stance the weapon would probably knock him on his butt.

I agree on it being fake. There is no shadow to indicate he's carrying an RPG launcher.
I will point out though. These weapons have very little to almost NO recoil. I've fired a few similar weapons in my lifetime. The rear of the launcher is open, so all the force of the launch is contained in the back blast. Keep in mind though, that anything behind the weapon for a distance of 20 to 30 feet is Toast.

In addition to whats been pointed out, there's too much flash out the back end and it would not be that widely dispersed

Mike150 said:

These weapons have very little to almost NO recoil. I've fired a few similar weapons in my lifetime. The rear of the launcher is open, so all the force of the launch is contained in the back blast. Keep in mind though, that anything behind the weapon for a distance of 20 to 30 feet is Toast.

Click to expand...

Really? the biggest thing I have fired was a grenade launcher and it had a little bite to it. Still rather fire a .300RUM

I was going to comment that the left arm does not appear to be up on the weapon but after looking at some other images it appears that some of these weapons do not require a second hand, which further attests to Mikes comment.

I agree.....fake. The shadow doesn't support the object purportedly being held.

Most likely photoshopped - on another forum a member had posted a photo of a mountain biker in an awesome jump - the whole photo looked so good everyone was saying fake, "how come there is no shadow" how come this how come that" Well it turned out to be the real deal he had set up strobes at the jump, so when the rider came by he started shooting and the the lighting and shot were spectacular. So much so he was challenged, however they were for real. So since then I am very careful to judge.

Seen a few videos of the RPG being fired, that whole 'blast' or smoke out of the back just looks wrong to me, somehow the photo doesn't feel right. BUT you do on occaision get a photo thats out and doesn't look natural. I would have though that to capture this very instant that there might be more in a burst? leaning towards fake

After a second look, the shadow arguments are very true but to me as an occasional shooter, what on earth is he supposed to be shooting at? Nah on instinct the picture doesn't look right to me at all

Lighting and Shadows
One major sure-tell sign something is fake is by looking at lighting and shadows. The time stamp on this photo is 11:29:14am. This is close to high noon, so the sun should be directly overhead. That said...is the time stamp when the photo was taken, or when it was uploaded? Or is there a time zone discrepancy?

The sun is coming from the left. You can clearly a beam of light on his ankle (the one w/ the brace), crutch, and the entire right side of his body is lit as well. His shadow cast is long and narrow to the left, which coincides with sunlight coming from the right. However, due to the length of the shadow, I would guess the actual local time of this photo to be closer to evening (6:00-7:00pm) when the sun is much lower in the sky or early morning (8:00-9:00am). This is a contradiction of the time stamp, but again, we don't know if that is local time.

Additionally, any RPG or grenade launcher that I've ever seen is going to be at least 3ft long. Where is the shadow from the launcher?

I'm inclined to believe this photo was taken in early morning when the light isn't as strong. There is no shadow cast from any of the light poles in the background, nor any from those large decorative bowls that line the street.

Firing the Weapon
It's already been stated that launchers of this type don't have much "kick". That is true. The back of the weapon is open for gases to escape, so the wielder isn't absorbing the shock like with a traditional firearm. Still...I think most would brace themselves when firing launcher, and not be standing in this casual posture.

Also, his clothing is loose-fitting. However, it appears as though he is holding this weapon with just one hand. His left arm looks as though it is in front of him, as the sleeve does not seem to be bent at the elbow as if his arm is raised. You can clearly see his other crutch next to his left foot, which leads me to believe he is holding onto it in front of his body.

Then there is the odd explosion. The timing on this shot would have to be perfect, which is plausible. However, I'm quite certain that any photographer in this situation would be shooting in rapid fire mode. Why keep this photo? Why not keep the one of a smoke-trailed projectile sailing through the air?

What the heck is this guy shooting at anyway? The effective range of an RPG is 200-300 meters (220-350 yards). Because of the angle of this photo, he is firing directly in front of him at a treeline. Maybe he has a gripe with the local forestry department?

Look at the launch wave at the back of this guy's head. We have a reversed circle with an opaque bar leading INTO the center of the explosion.

Proof positive... FAKE

I've been playing a bit of devil's advocate on this one. This one is most likely (95%) real:


ELA (Error Level Analysis) compares areas of a photo against each other to determine anomalies. If something were amiss, we would see a sharp outline here or a stark contrast between black and white. We do not. What does stand out here is the result of image sharpening:





This Luminance Gradient test would reveal if something were pasted or blended into the photo, such as that odd-looking blast. Here we are looking for "bumpiness" over the entire photo. Pasted in areas would appear flat, or have a different pattern:





And finally, there is a high frequency noise test. Here the image should be unrecognizable. You shouldn't be able to identify trees, the road, or even the person. It should all be a series of random dots. You can make out a few outlines, but again, this is the result of sharpening.





Aside from sharpening, the only modification of this photo is that the blast was enhanced by making it a little bit brighter. That may be why it seems so out of place. This may violate Reuter's rules for Photoshop enhancements, but I doubt it.

Most importantly, here is a news report of the actual fighting in Sirte that day. While this doesn't confirm or deny the authenticity of this photo, it only adds to the possibility of it being real. Finally, if you look just past the sidewalk and between tree clusters, you can just barely make out the outline of a tank:

I wonder how many different shots that we are looking at here. The shot that was so eloquently analyzed does NOT show the odd appearance at the back of the guys head, and yet it is clearly visible in the last photo.
This shot smells to high heaven.

This shot shows that even the crippled can become warriors. This is just too damn good to pass up because of the "moral" message imparted.

I ain't buying it.

Well at first I was in agreement with everyone else that this was photoshopped. Now that I see the tank, and have looked at the filter tests I would have to agree that this ones real.

Anthony, are those filters / test in CS5? Very interesting to learn you can analyze a photo like that. Might be good to know how to do that one-day for whatever reason.

Still not buyin that this is real. The guy is on 2 crutches and firing an rpg and the shadow is still not there to support what the guy is supposedly doing.

AxeMan said:

Anthony, are those filters / test in CS5? Very interesting to learn you can analyze a photo like that. Might be good to know how to do that one-day for whatever reason.

Click to expand...

You can run similar tests with Photoshop, yes. I use a standalone program that I wrote to run these tests. The extreme level of my nerdiness knows no bounds.

These things are about 3 feet long at least. The shadow should look at least a little like a "t", no sign of anything in the shadow. The fake blast would effect the shadow to a certain degree also, again nothing.

Apart from the shadow to me that explosion from the back of the weapon looks very very wrong it looks like a cropped bomb explosion that has been tweaked, after watching a lot of video footage on these launches from a variety of RPG devices, a lot show no explosive exhaust, some are launched by a gas charge before the 'rocket' part kicks in, the ones that don't use this method have a long exhaust plume, the whole feel of that blast area is wrong, it looks like its spreading out from a central point rather than being expelled from a tube. I would also surmise that if that much has managed to exhaust from the rear of the RPG launcher that the rocket itself should be visible as it heads towards the tank.

I was reminded of this thread again when I found this:
Without a Crutch - The Hacker Factor Blog

Cliff notes: The image isn't a fake.

The author is well known image forensics specialist, Neal Krawetz. You can read about him here: Hacker Factor: About


Edit: I just realized this is the source of what Anthony posted above

Anthony Hereld said:

You can run similar tests with Photoshop, yes. I use a standalone program that I wrote to run these tests. The extreme level of my nerdiness knows no bounds.

Click to expand...

Wait, did you do something with the code that Neal Krawetz wrote?