Photoshopped? need pro opinions

Bennie-Mac · 20 Aug 2011 at 15:04

Kristoph said:
lol.

I can tell by the pixels.

Jokester · 20 Aug 2011 at 15:04

Kristoph said:
You know this doesn't work right?

Here's with the edited one.

http://errorlevelanalysis.com/permalink/326b97c/

Errr, take a look at both images... there's a huge difference, the edited one shows massive amounts of errors compared to the other photo.

Kristoph · 20 Aug 2011 at 15:06

Raikiri said:
That's because it's all from the same image, it works based on jpeg compression differences in images which will vary depending on how many times each has been saved.

I realise that.

But I'm just pointing out it cant really be used as a tool for photoshops where things have been edited, eg. size of a fish.

Kristoph · 20 Aug 2011 at 15:06

Jokester said:
Errr, take a look at both images... there's a huge difference, the edited one shows massive amounts of errors compared to the other photo.

It's all brighter, even the background.

The face of the guy that's been moved isn't exactly bright.

Look at this example I've just seen,

Original Image - http://imgur.com/Apm1Ol&HbFwt

Edited version being checked on error level analysis - http://www.errorlevelanalysis.com/permalink/8411259/

Jokester · 20 Aug 2011 at 15:09

Kristoph said:
It's all brighter, even the background.

The face of the guy that's been moved isn't exactly bright.

That's the entire point! The second one is a good example of an image that's been processed and saved, the original one, shows little evidence of it being a composite of different images.

Edit:

Kristoph said:
Look at this example I've just seen,

Original Image - http://imgur.com/Apm1Ol&HbFwt

Edited version being checked on error level analysis - http://www.errorlevelanalysis.com/permalink/8411259/

See, this is a good version of what should appear of that pike had been added to the image and the image edited for things like shadows and so on. The pike photo has none of that in it at all.

Le_Petit_Lapin · 20 Aug 2011 at 15:11

simulatorman said:
Only found this one

http://forums.overclockers.co.uk/showthread.php?t=18130582

Oh, well I was close wasnt I? Same colour

cu3ed · 20 Aug 2011 at 15:13

Tiles along the ground where the fishes tail has been altered have been cloned.

Kristoph · 20 Aug 2011 at 15:14

Jokester said:
That's the entire point! The second one is a good example of an image that's been processed and saved, the original one, shows little evidence of it being a composite of different images.

The second one isn't a composite of different images. Like Raikiri said, it's just been resaved.

Resave any jpeg without editing it and you'll get a brighter result in error level analysis (more so if it's not saved at 100% quality).

Obviously the shutter and the car have been photoshopped in this!

http://www.errorlevelanalysis.com/permalink/2fe0881/

lol

And personally, I don't think the original photo in this thread has been photoshopped.

Jokester · 20 Aug 2011 at 15:20

Kristoph said:
The second one isn't a composite of different images. Like Raikiri said, it's just been resaved.

Resave any jpeg without editing it and you'll get a brighter result in error level analysis (more so if it's not saved at 100% quality).

Obviously the shutter and the car have been photoshopped in this!

http://www.errorlevelanalysis.com/permalink/2fe0881/

lol

And personally, I don't think the original photo in this thread has been photoshopped.

But the original photo doesn't have that brightness in it. It's comparable to the edited version which we know has been edited...

Kristoph · 20 Aug 2011 at 15:42

This is getting confusing now.

This error level analysis tool claims to be able to show alterations through bright areas on the analysis overlay.

Well there's things in these photos that haven't been edited, but are showing as bright areas, suggesting they have been edited.

It also fails to highlight obvious photoshops.

It's a mess.

Here's a resaved version of the OP's image.

http://errorlevelanalysis.com/permalink/a0cd97e/

It's brighter, must be shopped.

Dr. Pain · 20 Aug 2011 at 15:47

Its real surely.

Not one of you has mentioned the fact that you can see the weight of the fish is dragging his tshirt down as hes holding it against him.

Jokester · 20 Aug 2011 at 15:48

Kristoph said:
It's brighter, must be shopped.

What it's specifically looking at is the image degradation due to JPEG compression, so everytime you save it, it'll become brighter and brighter. It'll also highlight differences between composite images that may have been saved different numbers of times. Less brightness means that it's likely it hasn't been edited in anyway (requiring another save).

Edit:
The original image has very little areas of intense brightness, but even just that one recompression of the unedited image makes the whole image substantially brighter which gives credence to the orginal image not having been edited in anyway.

byteagig · 20 Aug 2011 at 15:48

wont it only work with the original edited pic, not a saved then posted on the internet pic? or is i been fik :confused:

MonkeyBasher · 20 Aug 2011 at 15:53

Bennie-Mac said:
Are Pikes not freshwater fish? Looks like a saltwater environment to me.

The OP pic was taken at Salford Keys in Manchester. Its where a river connects with the Shipping Canal, so its all freshwater.

Fishing has only been possible here the past 5 years or so due to the oxygen content of the water. If this helps anyone decide, can fish grow that big in 5 years?

Kristoph · 20 Aug 2011 at 15:58

Jokester said:
What it's specifically looking at is the image degradation due to JPEG compression, so everytime you save it, it'll become brighter and brighter. It'll also highlight differences between composite images that may have been saved different numbers of times. Less brightness means that it's likely it hasn't been edited in anyway (requiring another save).

Edit:
The original image has very little areas of intense brightness, but even just that one recompression of the unedited image makes the whole image substantially brighter which gives credence to the orginal image not having been edited in anyway.

I agree the original hasn't been shopped. I'm just saying I don't think that site can really prove it.

Look at this;

This has about as many areas of 'intense' brightness as the original photo

http://www.errorlevelanalysis.com/permalink/ba01879/

vs.

http://errorlevelanalysis.com/permalink/f68213e/

In fact, I'm fairly sure the original was taken on Omaha beach. There's too much brightness in the jetty photo.

Jokester · 20 Aug 2011 at 16:02

Kristoph said:
Look at this;

This has about as many areas of 'intense' brightness as the original photo

Ooooh that's cheating

try it with the pike in colour.

Bald-Eagle22 · 20 Aug 2011 at 16:07

Deffo shop ,

You would be covered in crap lifiting a 31Lb Pike , and you would also have to support the fish with your right hand under and around its gills showing your hand under the fish.

Dj_Jestar · 20 Aug 2011 at 16:10

It's real ffs.

I die inside every time someone picks out the smallest imperfections on an image using a COMPRESSION ALGORITHM.

Really, would someone bother to go and make such a perfect photoshop just to claim they caught a 31lb pike?

Dj_Jestar · 20 Aug 2011 at 16:11

Bald-Eagle22 said:
Deffo shop ,

You would be covered in crap lifiting a 31Lb Pike , and you would also have to support the fish with your right hand under and around its gills showing your hand under the fish.

He's holding the fish correctly. His right hand will be gripping the gill/jaw like you are supposed to with pike.

And look at the ground he is stood on.. it's not a muddy bank, it's a dock. There is no crap to stick to him...

Kristoph · 20 Aug 2011 at 16:13

Jokester said:
Ooooh that's cheating try it with the pike in colour.

http://www.errorlevelanalysis.com/permalink/b4f2188/