Can you detect the "faked" photos?

[Mike Buckley]

A newly published study by researchers at the University of Warwick in England has found that people were only able to spot fake photos among real ones 60% of the time. What’s more, those who correctly identified fake photos could only point out what was faked 45% of the time.

EDIT: The test explains that the person taking it is to determine whether a photo has been "digitally altered." That is unclear, as discussed later in the thread, so I'll explain that I know from an article that I read about the test that you are to look for something that has been faked by digitally changing part of the image to display something that was not there at the time of capture. You are not to look for any global alteration such as a change to the exposure.

You can take the test here.

I accurately determined 7 out of the 10 photos. Two of the photos I was certain had been faked weren't. One that I was certain had not been faked was. When providing the test results, I wish the website would indicate what had been faked.

[skitterbug]

I tried it for interest to see if I could detect anything out of sorts... or not. I managed to get 50% so middle of the road for me.

[JBW]

A newly published study by researchers at the University of Warwick in England has found that people were only able to spot fake photos among real ones 60% of the time. What’s more, those who correctly identified fake photos could only point out what was faked 45% of the time.

You can take the test here.

I accurately determined 7 out of the 10 photos. Two of the photos I was certain had been faked weren't. One that I was certain had not been faked was. When providing the test results, I wish the website would indicate what had been faked.

if my mind is working properly I'd suggest that by checking your right and wrong answers you should be able to discern which were faked and which weren't.

[pnodrog]

I lost interest. They were not particularly good photographs and the changes (if any in the ones I looked at) apart from possibly one were minor. I never zoomed in which I think may have been necessary to confirm.

[Mike Buckley]

if my mind is working properly I'd suggest that by checking your right and wrong answers you should be able to discern which were faked and which weren't.

True. However, doing so doesn't indicate what had been faked to make the photo a fake. As an example (without giving anything away to someone who hasn't taken the test), one photo displayed a shadow that was in the opposite direction of all the other shadows in the scene. Clearly, that part of the scene had been substituted.

[Mike Buckley]

I lost interest.

I found it difficult to maintain interest, enough so that I was glad there were only 10 photos to review.

[Merlin52]

Ok, I fess up, only got 50%, as with most psych studies I guess there was a sub study going on.

[Rob Ekins]

The term used in the study was "Digitally altered" This is pretty ambiguous there are 100s of ways to digitally alter an image, are we talking exposure correction or dodgy cut and paste.

[billtils]

The term used in the study was "Digitally altered" This is pretty ambiguous there are 100s of ways to digitally alter an image, are we talking exposure correction or dodgy cut and paste.

Indeed. Every one of them could have been altered, some more skilfully than others.

Not a very well-designed study (and I thought that before taking it and mulling over my score of 6).

[Dave A]

I did the test and got 8 out of 10 correct. I didn't look for changes to the entire image like exposure for example, but did look at shadows and anything that looked like it was pasted in or cut out.

Dave

[Mike Buckley]

I have added the following to the first post in the thread:

The test explains that the person taking it is to determine whether a photo has been "digitally altered." That is unclear, as discussed later in the thread, so I'll explain that I know from an article that I read about the test that you are to look for something that has been faked by digitally changing part of the image to display something that was not there at the time of capture. You are not to look for any global alteration such as a change to the exposure.

[Mike Buckley]

Bragging rights go to Dave so far. Way to go, Dave!

[tao2]

Where's the pictures?

[Mike Buckley]

You have to take the test to view the photos and there are several web pages requiring you to answer some simple questions before the photos are displayed. The link to the test is in the first post of the thread.

[RandyRFJP]

I got 6 out of 10 correct, and was surprised I did that well. My decisions were made mostly on angles of shadows and what I perceived as "cut and past" changes to the images.

[Panama Hat & Camera]

The term used in the study was "Digitally altered". This is pretty ambiguous because there are 100s of ways to digitally alter an image, are we talking exposure correction or dodgy cut and paste (copied from Robbie's analisys because I agreed 100% with him).
I got 6 out of 10 correct, and was surprised I did that well. My decisions were made mostly on angles of shadows (partially copied from Randy's analisys because I got the same score and I used some of his methods).