Bizarre digital photo - how did this happen?

[william_sturtevant]

<p>"Second there is No Air Force 1 or 2 unless the President is on board and it is up in the air." Technically true, although posting 'Here is a picture of what was previously designated at AF2 two hours prior" becomes rather cumbersome, don't you think? :)<br>

Bob, no problem. I understand the skeptism. Heck, my own friend was skeptical! <br>

Phyllis, I actually did email Samsung at your suggestion. I'll post their response if I get one.</p>

[phyliss_crowe]

<p>Which model Digimax do you have?</p>

[perreira]

<p>@Tom: sure about the black holes? CERN could be now mass producing them...<br>

I'm a physicist too and agree with Tom: there are two possible ways this could have happened: <br />- some algorithm in the camera went haywire and produced funky effects<br />- or this could be a photoshop work. However all the "there's some funny pixels" and "the shadows are wrong" sound like wild guesses and the usual "it's a fake" cry.<br>

It would be interesting if you can reproduce it: take some similar situation where somebody stands in front of a line with light coming from the back and check if you get that effect again...</p>

[william_sturtevant]

<p>It is a Samsung Digimax 301, 3.2 Mega Pixel, Focal Length 5.1mm Samsung Lens. </p>

[phyliss_crowe]

<p>Wow. That's....old. Like 2004. </p>

<p>Well, I don't know how much use it's seen prior to this incident, how it's been stored, etc., same for the card, but knowing its age now, that camera and maybe the card, too, if you were using one, have long since reached end of life and are failing IMHO.</p>

<p>Treat yourself to a brand new sparkly clean camera and card.</p>

[sudip_choudhury]

<p>Strange picture..This is my thought. I could be wrong as I am not an expert in optics etc...We all know about optical illusions. Our eyes are being tricked to believe something which actually is not. The same thing happened here. The Lens is like our eye..Here the lens got tricked..<br>

You can try one simple experiment here. Just cover the line in all the picture except the pants. You can cover with the same color as the ground. You will note that the lines through the pant look much lighter. In fact if you do not observe carefully, you might not even note it. where as when you see as it is, the lines look so prominant. <br>

Not sure if this can be replicated with all lense. It probably has also got to do with the particuler camera model.</p>

[daverhaas]

<p>Tomm -</p>

<p>Jpeg errors - when they occur don't typically replicate patterns like the artifacts on this photo, nor do they look as natural as these do.</p>

<p>If it were a jpeg artifact or a file error - I would expect it to be much brighter and random than what we are seeing here.</p>

<p>As for the PS / Manipulation of this photo - to what end? It would make no sense at all to take a photo, ps / manipulate it then come onto a forum of strangers and ask what happened to the photo... unless the date were April 1...</p>

<p>Radiation Heat - I think what the poster was referring to was the heat from the tarmac itself, not from the plane's radar. As pointed out - heat reflections tend to be viewed / observed from a distance - not 10-20 feet... anyone that has ever driven on an interstate in the summer can attest to that. </p>

<p>I'm sticking with my first guess - flare or an error in the software processing the jpeg in camera.</p>

<p>Dave</p>

 

[pearse s]

<p>This is another reason why film is superior to digital.</p>

[phule]

<p>And yet trolls are just as ugly regardless of what camera they use, Pearse.</p>

[Lou_Meluso]

<p>Could this simply be a reflection bouncing off the sensor to the rear of the lens and back to the sensor? The lines seem to match but in close inspection they seem to be off slightly and the over all appearance of the lower half of the frame is one of optical flare. Just my guess.</p>

[rafael_s]

<p>About as strange as my Nikon D200 with over 16 million shutter actuations (as stated by every shutter count software I've used.) Electronics can do funny things.</p>

[oscar_van_der_velde]

<p>14.6 EV just means the Exposure Value was 14.6, which corresponds well to nice sunny circumstances. It is not exposure compensation.</p>

[oskar_ojala]

<p>The explanations that this would be an electronics/software problem do not convince me. The problem is that the pattern fits the subject way too well. Neither hardware nor software care about the scene contents when they go bad. And there is no way the camera could have seen behind the two guys, unless the camera was on the tripod and got some ghost image before the guys walked in, which is not the case.<br>

So the cause would be more likely to be found in the world of optics...</p>

[tom_mann1]

<p>Oscar -</p>

<p>Your're probably right about the 14.6 number being the EV. </p>

<p>However, EXIF tool names that field, "Exposure Compensation" NOT "EV", and correctly shows compensation values in the range of minus a stop or so to plus a stop or so in units of one-third of a stop, and the numbers it displays match up exactly with the settings on my d200 and d700 cameras, as well as a bunch of P&S cameras that I own. EXIF Tool also displays reasonable values in this field on a random sampling of on-line images that I checked which included (as I recall) Pentax, Canon and a couple of other brands. This is the first time I have ever seen a number this high in this field.</p>

<p>My conclusion from this is that the Samsung embedded SW loaded the wrong number into that field. If Samsung can make an error like that, my confidence in them doing other things correctly is reduced. </p>

<p>FWIW, I seem to remember considerable discussion (a couple of years ago) about the need for correctly formatting the data in the EXIF fields, and a warning that bizarre things happen if you do it incorrectly (ie, don't pad unusually short number with zeros, overwrite other fields by storing too large of a number in a field, etc.). We might be getting to the heart of this matter.</p>

<p>Tom M</p>

[tom_mann1]

<p>I tried three experiments, each of which intentionally corrupted the file under consideration in different ways:</p>

<p>1. Using EXIF tool, I substituted several different values into the Exposure Compensation field. Some values were reasonable, eg, "0", while others were clearly impossible, eg, "1000" and "-1000". These had no effect on the photo in several different viewers that I tried.</p>

<p>2. Using the original, large file, I changed the byte at offset 0004e000-00 from "DF" to "11". The effect was to insert a pale red veiling glare in the lower 40% of the image. This is shown in the attachment to this message.</p>

<p>3. I re-encoded the image using the "progressive" option and relatively low quality setting. I then changed the byte in the new file at offset 0000a0000-00 from "78" to "11". The effect was to insert a jagged, stairstep effect in the strips in the tarmac in the lower central part of the image. This is shown in the attachment to the next message.</p>

<p>My conclusion is similar to Oscar O., namely, bit rot would be extremely unlikely to produce such a precise match-up between the corrupted image and the original, but the above experiments show that it doesn't take much, just the twiddling of one byte, to either put a veiling flare over part of the image or, with a progressively encoded file, to induce corruption of the image in one area. I'm not saying that I think bit rot is the cause of what we are seeing, but the effects are not necessarily extreme, as suggested above.</p>

<p>Cheers,</p>

<p>Tom M</p>

<div></div>

[tom_mann1]

<p>Experiment #3</p><div></div>

[tim_ziegler]

<p>Note also that the line of the stripe as it runs through the legs is somewhat lower in resolution. I've tried to PS in an overlay with lower opacity, but it does not lower the sharpness of the line as in the original. Also note that if you take a look at one of the other photos - the "Lukas at AF2" photo - and pay close attention to the stripe (overlay it on the original as close as possible for a photo shot from a different location) you will note that the particular area behind Tiny's legs was actually missing some of the black paint from the near side. This does not show up in the band either. Both of these indicate to me a flair or sensor burn instead of a transparency or double exposure.<br>

Tim Z </p>

[wieland_willker]

<p>I know this guy. I searched the net and found this one:<br /><img src="http://666kb.com/i/bituo1gzw0z42y6yp.jpg" alt="" /></p>

<p>Seriously, I think what <a href="http://www.photo.net/photodb/user?user_id=4578930">Bob Irvine</a> wrote above is so far the best explanation.</p>

[tom_mann1]

<p>OK ... I think I have part of the answer. Open Bill's image, "Lukas at AF2". Look at the shape of the flare caused by the reflection of the sun from the upper RH corner of the rear window of the car. It's one dimensional, and spreads in exactly the same direction as the yellow line in the image under question. Whatever is causing the light to spread like that will affect everything in the picture, albeit at a much lower level. However, if there is some object in the picture that is reasonably intense and in the same direction as the flare (eg, the yellow line), the flare will faintly fill in any missing sections.</p>

<p>For Nils Pickert (the other physicist) - regard the shape and size of the flare in the "Lukas" image as the Green's function response of the OP's camera, ie, it's spatial response to an intense point source of light. Convolve that specific Green's function with the scene that was photographed by the OP, and I bet you would see exactly the "filled-in line" artifact we have been discussing, but its effect on other parts of the image (which don't have structure that is aligned parallel to the flare direction) will only manifest itself as a reduction of the contrast.</p>

<p>Note: There is an area in the image we've been discussing (aka, the "Transparent Legs" image that has a hot spot, and that is the very bottom front of the engine. If you look carefully, you will see faint flare coming from that spot, and, as a confirmation, the flare is in the same direction as flare in the "Lukas" image and the yellow strip in the image under discussion. In fact, I am not the 1st to notice this - Bob Irvine noticed it first and commented on this above.</p>

<p>With the right software, it is very easy to simulate the effect of a Green's function and test this theory. I'll try to do this when I get back to my office next week.</p>

<p>Of course, the real question is what produced lens flare in this very unusual shape? It's not at all like the multi-lobed, star-shaped flare we are used to seeing. Since it is almost one dimensional, the object causing the diffraction must also be approximately one dimensional, eg, a screen. Bill S - did you put anything in front of the lens of your camera? If not, are there any scratches on your lens?</p>

<p>Cheers,</p>

<p>Tom M.</p>

<p> </p>

[patrick j dempsey]

<p>The following image is just to debunk the debunkers. The only thing more annoying to me than people Photoshopping images for attention, is people saying images are "shopped" for attention. For everyone of you who said things like "the shadows don't match up" or "where is the shadow of the engine?" or "why is there no reflection of the guys on the car?" ... guess what? You just failed basic Geometry. For your education, light travels in straight lines, even if it's inconvenient for your brain:</p>

<p>http://i16.photobucket.com/albums/b15/patrickjdempsey/notshopped.png </p>

<p>The black lines are sun rays. The orange dots are the locations the sun's rays grazed objects in the image. The gray dots are where those rays hit the ground making the edge of a shadow. The green line is the horizon (making the person taking the photo either about 2 inches shorter than Mr. Sturtevant, or they tilted the camera up to get more of the airplane). The orange lines are the lines on the tarmac heading in a direction we can almost bet is due West. The blue lines and dots represent the centerline of the shadow of the engine assuming the jet is perfectly perpendicular to the tarmac. The purple oval on the car is the reflection of the two gentlemen in the foreground, they are shortened by the curvature of the body panel.</p>

<p>Amazingly, all of the black lines converge slightly as they go up as if they are going to one point... the sun maybe? Amazingly the shadow of the engine begins directly behind Tiny right elbow, barely misses appearing beneath William's right arm and rejoins the wing somewhere in his left lung, however the trailing edge of the engine is clearly visible, the blue dot. If someone was to go through all of the effort to generate this image with that level of accuracy... why the heck would they put a big yellow line through the subject's legs?</p>

[perreira]

<p>@Tom: wouldn't you expect some intensity variation over the the large area of the secret service guy's legs - shoudln't it bekome less prominent further away from the edges?<br>

It also looks a bit like the shoe of the guy at the car also shows the effect. It is only the diagonal yellow-black line, no other lines shine through.</p>

[15sunrises]

<p>It almost seems as though the person in the background has the line going through their shoe as well. At first I thought it was just a reflection, but the part of his pant leg where the black part of the line would go though, also seems darker. Maybe my eyes are starting to play tricks on me (as long as it's not my camera it's cool!).</p>

<p>This is really very strange, hopefully somebody can solve this mystery.</p>

[ron831]

<p>Ok.. I can't stand it no more. The cargo plane in the background is not a C-141 or a C-5A. It's a C-17.. Now I feel better.. </p>

[accystan]

<p>Fascinating post. Amazing really, because the most obvious thing has been missed - the big guy ("Tiny") in the left f/g ............ IT'S ELVIS!<br>

Dave D</p>

[michael_roberts2]

<p>The lines that go through the people do not line up with the line on the ground. It's pretty close on the right, but it's noticably off on the left. It's sitting high. So we can safely rule out magic x-ray vision.<br>

What seems VERY strange to me is the shadow of the airplane. The airplane's shadow is behind it while the shadow of everything else is moving to the lower right.</p>