Ghost Photo

russ_watters

4th pic referenced above...

I also noticed that due to the feather area being too close to the border of the cropped pic, there is still a little bit of outline noticeable in what I posted last night. So I went back and copied from the original, uncropped pic for a new "final" version, attached. I see no evidence of an outline in it.

Attached Thumbnails

   

russ_watters

One of the real problems I see with the overall issue of using untraceable photos for evidence of anything is just how easy it is to fake them and just how hard it seems like it would be to spot the signs. Altogether, I spent about 5 minutes creating the sample images for my little tutorial above, so they still aren't very sophisticated. Could you go into some more detail about how one might spot such a fake? I'm not sure there is any way to naked-eye spot the evidence, but I can think of two possiblities with analysis software (speculation):
1. Mathematical analysis of the blending to identify the blending algorithm. Evo noted that dark-on-dark blending gives the appearance of more transparency with the overlay. Maybe ghosts work that way too, but there is probably a way to prove mathematically that it uses a Photoshop blending algorithm.
2. Finding the gradients in the edges, again, mathematically.

Ivan Seeking

Okay...

When I said "raw", I meant the original, unmodified form, wrt to tampering. Can it be determined conclusively that the photo was/was not uploaded from a pc or received by email?

Pythagorean

I don't think it's a ghost, but this complaint can be explained by shadow and color. A brighter color will show up superimposed on a darker color (i.e. both images will show up: the wheel and the face). When both colors get dark (near her neck) they are more indistinguishable because they are closer to resembling each other.

russ_watters

If you can match the photo with a sample by reading the exif data and checking the modification dates, it would be fairly conclusive that it was not an edited photo. If they don't match, it would be certain that it was edited.

Ivan Seeking

By a sample, do you mean a photo that was taken at the same time as the photo in question? That makes it pretty tough to actually test a claim. The claimant could just lie and provide a sample that was taken later, right?

Thetom

You can visually determine fake photos with the naked eye, to some extent, unless the fake is very sophisticated. This one isn't. Can you see the difference in quality of the two images? The same is apparent in your example russ_watters. The image of the girl is of a different quality to the rest of the image. You can see the difference in the amount of aliasing in the image. Notice if you zoom in curved edges in the original are blocky and staggered. You also get bleeding on these edges, along with artifacts due to compression. The edges that describe the girls face are far smoother with no bleeding. I suspect the image of the girl was taken using a traditional camera (or possibly a high-res digital camera). The rest of the picture was taken on a low quality digital camera. When you put a nice, high quality, high-res picture over a low quality picture this is what happens. Dead give-away. If you wish I can post a picture illustrating what I mean?

Thetom

I took the liberty of making a quick illustration. Unfortunately i'm using a tired old CRT monitor that is very, very dark and so I have to turn the brightness up allot. In order for me to get a good view of the photo I had to readjust it and take the brightness way down. If this image is too dark to see on your screen adjust it so that it is clear. Alternatively I can undo the adjustments I made and re-upload so that you can see the picture in its original light.

Hopefully you can see the difference in quality of the two images. It is very apparent. The girls face is smooth, crisp and clear even when magnified (of-course it still has the aliasing and bleeding, just allot less). Comparatively the rest of the picture is jagged and broken. These are clearly two pictures from different sources. Also please note it is impotant to view that illustrtion at its full size (click on it again when in imageshack). If you view it smaller the effects will be dimnished (though still possable to see).

eupeptic

A couple considerations that haven't been mentioned:

Consider the height of the bar stool, TV, coffee table, and baby, as well as the size of the stereo equipment - the ghost seems out of proportion...

And if that's really a ghost looking to be noticed, why is she looking off to the side? Why not look at the camera? (And why have such a plain and somber expression if you're dead and seeking attention?) (This is an easy thing to spot that is often overlooked by people who want others to believe that they have an image of a ghost.)

russ_watters

No, I just meant a photo taken with the same camera so you could compare the exif data. It doesn't need to be at the same time. In fact, it would be better if you took the test picture yourself. Since a cell phone camera and/or internet hosting program will typically resize a picture and often strip the exif data, comparing originals would be a good way to know if a picture you are seeing is really an "original".

YinepuOfSand

i have a picture for you...and it's not fake...and i don't know how to interpret it...
first pic is not modified...was taken last year at the medieval camp in bistrita, romania, at the haggard concert...i just was with my camera on in my hand...i have to say that were some fire games, but you weren't in the middle of them when the picture was taken...i saw it for the first time after i was home...
this is the original one...the second i cannot uploaded because is to big, and i modified only lights and contrast, so it can be seen better...
you can see a 'couple' and some faces...'the couple' is not like any of my friends i was that day...enjoy!

Attached Thumbnails

 

JaredJames

It's nothing special. It is fake (but I'm not saying deliberately).

If digital, it's just a mashup of two photos.

However, I'd say it was taken with a "film" camera (you know, 35mm etc) and it's a double exposure. Simple as that.

This is further backed up by the fact you have a background behind the couple and not just their figures standing alone.

(And this is of course ignoring the question of why two ghosts would stand in the shown pose, at such an odd angle - you're aiming the camera at the floor but they're standing upright as if your legs are the ground they walk on.)

YinepuOfSand

first of all...it was a digital camera
second...none of my friends are looking like the couple
third...if can zoom in you gonna see multiple faces, big and small, that are not human-like
last...i'm a boy
i've asked several photographers and they could not give me an explanation of that...
don't jump with conclusions!

JaredJames

Then it's fake in another way.
So?
The brain sees a lot of things that aren't really there. We try to deliberately find recogniseable patterns. Human condition.
Again, so?
They can't explain after proper computer analysis? Or they can't explain it by accepting you haven't tampered with it? There's a big difference.

I see potentially four faces, but the couple are in modern dress. Of course that doesn't really mean anything (why shouldn't we have modern ghosts?) but then they're posed perfectly as if standing for a camera shot. To add to that, as I said before, this would mean the ghosts where standing on your legs. Why would they do that? What would make them think "ah, a camera randomly aimed at the floor - let's get stuck in!"?

Also, why such a random shot?

JaredJames

Well it looks like a perfect example of double exposure. But then we're told it's digital and so the only way this can occur is by being edited.

I'd love to examine the original, but I somehow doubt we'd get that.

JaredJames

[QUOTE=YinepuOfSand;3207852]nismaratwork@ the photo is not fake you stupid if is on the camera![QUOTE]

We don't know it's on the camera. We don't know anything other than the picture you posted which may or may not be the original.

Cut the attitude. It looks about as fake as possible and there's nothing about it that suggests otherwise.
So what if they don't look like someone from your group. I never said, nor implied they did.

It's random because of where you're aiming the camera - or did you mean to take a pic of someones crotch?

EDIT: In fact, it would be silly to use people you know to fake it. People would spot that far too easily.

YinepuOfSand

is not edited...why do you people think that everything is fake?...how the **** can i do double exposure with so many figures with different sizes?...