is this photoshopped/altered?

Discussion in 'Digital Photography' started by mcblade, May 30, 2017.

  1. mcblade macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #1
    i'm not asking about image filters by the way. do any of you expert eyes with photoshop think this photo has been shopped/altered? if you need more information, please ask, however, i assume it's probably best if you see it with as little information as possible so you're not predisposed one way or another.
     

    Attached Files:

  2. kenoh macrumors demi-god

    kenoh

    Joined:
    Jul 18, 2008
    Location:
    Glasgow, UK
    #2
    I hope this isnt to try and support a claim on someones inheritance! :)

    I cannot tell. At best I think the lighting / WB may be different on the boy on the left. Other than that, I cannot see anything but I am not an expert.
     
  3. Stefan johansson macrumors 65816

    Stefan johansson

    Joined:
    Apr 13, 2017
    Location:
    Sweden
    #3
    Looks like a phone cam picture with bad light conditions,but of course it can have been altered.
     
  4. anotherscotsman macrumors 65816

    Joined:
    Aug 2, 2014
    Location:
    UK
    #4
    I'm not an expert but using the curves adjustment that i normally use to identify sensor dust (sinusoidal with 4 peaks-and-troughs) to emphasise edges and tonal changes, it doesn't look overtly altered in a major way to remove or add features. Doesn't rule out global changes (grain, colour, contrast etc) though. Looks like it could be a scan of an old photo taken in poor lighting conditions and perhaps a bit faded.
     
  5. kenoh macrumors demi-god

    kenoh

    Joined:
    Jul 18, 2008
    Location:
    Glasgow, UK
    #5
    Is this is where the OP says it is the same boy taken at three stages baby -> toddler -> junior :)
     
  6. anotherscotsman macrumors 65816

    Joined:
    Aug 2, 2014
    Location:
    UK
    #6
    Almost certainly :)
     
  7. mcblade, May 31, 2017
    Last edited: May 31, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #7
    it's not. nothing like that going on here :)

    by global, i take it you mean a change applied to the entire image, not specific sections of it, right?

    funny, but no :) i have no knowledge of that being the case. i'm not pulling a prank here.

    thanks for your responses so far! despite you saying you're not experts, they've been pretty good. hopefully we can answers from a few more people in here. maybe an expert will even show up :)
     
  8. Varmann macrumors member

    Joined:
    Jan 3, 2010
    #8
    As the others I am not an expert but I am suspicious about the head size of the child to the left. The distance from top of the head to tip of the chin is exactly the same for the children to the left and right. They are that different in age that you should have a certain difference. The perspective may play a role here but, but it looks like they are not that far apart regarding distance to the lens. Kids at that age increase their head size about 1 inch per year (circumference).

    In addition you have that difference in lightning on both. The proportions of the face compared to the surrounding head looks a bit off as well. If it is photoshopped it is pretty well made, but my guess is that the face of the kid to the left is put on the head of an older kid.
     
  9. mcblade, May 31, 2017
    Last edited: May 31, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #9
    as for the lighting it might mean there was a problem with the flash that got corrected after the picture was taken but it sort of washed out the kid on the left's face.

    as for head size maybe the kid just had a big head for their age and their face hadn't quite caught up? and the kid on the left does seem a little closer to me. though i see what you're saying. that does look unusual now that you mention it. doesn't seem conclusive though.

    perhaps some of the good people here have some explanations for this. i know that strange optical illusions can sometimes happen.
     
  10. anotherscotsman macrumors 65816

    Joined:
    Aug 2, 2014
    Location:
    UK
    #10
    Exactly
     
  11. Janichsan macrumors 65816

    Janichsan

    Joined:
    Oct 23, 2006
    #11
    ...or it's simply an old analog photo.
     
  12. kenoh macrumors demi-god

    kenoh

    Joined:
    Jul 18, 2008
    Location:
    Glasgow, UK
    #12
    Hmm, looking at it again, I see the face scaling questionability. My points would be:

    1. White balance on boy on left looks different to baby and older boy. Seems cooler
    2. The level of the cushion on the back of the chair is different a la Mona Lisa on either side of the boy on the right - this could just be a play of positioning mind you.

    How are we doing?
     
  13. dwig macrumors 6502

    Joined:
    Jan 4, 2015
    Location:
    Key West FL
    #13
    To start with, this copy of the image is way way too low resolution to be certain whether there were any significant alterations done, period.

    Things that can be said:
    • It is an older film photograph and either the original film or a decent print was scanned using some type of device (scanner, copy camera setup, ...) and software to create a digital file. That, to some degree "alters" the original image. In fact, printing a negative "alters" the image as the operator of the printer make decisions about color balance and density.
    • At some point in its life, the digital file was processed by Photoshop v3 (that's version 3.x from 20 years ago and NOT CS3). I don't know Ps's "fingerprinting" (there is text in the header of the file which can be read by opening the image in a good text editor) well enough to know whether v3.0 and v3.5 (the first available on Windows) use slightly different entries or whether they both use just 3.0. I suspect that this was actually 3.0 and not 3.5 which would indicate that it was done with a Mac as 3.0 was Mac only.
    • Close examination shows very uniform noise patterns. If any retouching was done, it was done on a much higher resolution copy before downsampling to this small size and conversion to JPEG from the original scanner format, probably TIFF.
    • Global (that is, uniform across the entire image) adjustments for color and densities could have been applied when the scanned image was done. The image looks like the original film or print had faded and some correction has been done, though better correction is likely possible.
     
  14. mcblade, Jun 3, 2017
    Last edited: Jun 3, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #14
    You people on the forum here noticed everything I was wondering about, i. e. why I asked.
    What do you mean by a 'play of positioning'?

    960x720 is the highest resolution I have but for future reference with other images, what resolution do you recommend people obtain images in? Thanks for telling me about resolution. I thought 960x720 would be good enough. I admit I don't know much about photography.

    Things that came purely from color balance and density decisions during scanning isn't the kind of alteration I'm talking about. But that is a type of alteration, good point.

    You talked about Photoshop 3.0 being in the header. Looking at it, I also see 'FB' in the header. Processed by Facebook, I thought? I uploaded a different photo to facebook today to test this, and it completely reworked the header and put both 'FB' and 'Photoshop 3.0' in it! I wasn't expecting that.
    It looks like facebook runs people's photos through a batch process with really old Photoshop from the 90's for compression. That's crummy.

    EDIT: I just found out facebook, for a long time, had a default max image size of 960x720 unless the uploader remembered to hit the high quality switch, so if it went through facebook, someone could have uploaded a file with a higher resolution but facebook cut it down.

    But no retouching was done at this resolution, anyway. Good to know, thanks. How high would the resolution have to have been, do you estimate?

    What is 'better correction' as opposed to 'some correction'?
     
  15. kenoh macrumors demi-god

    kenoh

    Joined:
    Jul 18, 2008
    Location:
    Glasgow, UK
    #15
    By play of positioning, it could be the design of the chair/sofa. If the back is higher than the arm, and the boy is wedged in the corner to optimise support when holding baby, then maybe one side is the back cushion and the other is the arm so it would legitimately be at different heights. However the picture makes it look flat background and so I am struggling to work out if that is right or whether it is a Mona Lisa moment (the horizon on each side of her head are different heights).
     
  16. dwig, Jun 3, 2017
    Last edited: Jun 3, 2017

    dwig macrumors 6502

    Joined:
    Jan 4, 2015
    Location:
    Key West FL
    #16
    The best resolution for image distribution varies with the type of distribution, but is not, as a rule, the best resolution for editing or an "forensic" evaluation. I would suggest creating "master" files at resolutions of 2400x3600px or higher Retouching and adjustments done on higher resolution files can become undetectable when a lower resolution copy is made for distribution. Also, the original scans and master files should NEVER EVER be JPEGs.

    If you are reproducing an existing print (e.g. scanning an old photo) you have to pick between 3 "targets"; match the existing print, match what the existing print would have originally looked like, match the original scene. Only then do you have a base for deciding what "better" means.

    The image in the original post looks very very "warm" (orange/red). Based on the shadows, it was not taken on a daylight balanced film with tungsten lighting, and thus the warm cast, but instead was taken with some type of small flash (note the sharp shadows) which would result in proper color balance. The blanket in the lower right appears to have been rather white. On that assumption, a simple one-click Photoshop curves adjustment produced this "better" color balance:
    3kids.jpg
     
  17. mcblade, Jun 4, 2017
    Last edited: Jun 4, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #17
    I see what you mean.

    Thanks, dwig. That's fascinating. Yeah, analog pictures taken with a flash that are scanned do tend to get red/orange, I've noticed that with some pictures I've scanned anyway. I guess in those cases, it was me forgetting to optimize the color settings. Could be what happened here with whoever scanned it.

    Also, I have seen some people's software make their master files JPEGs. I see now that's bad software. I was reading today one could use something called "error level analysis" on JPEG's to estimate how and if a JPEG had been resized since it was converted to a JPEG. Something about how after it was resized it would create a second artifact grid. Is that something you ruled out when examining noise?
     
  18. dwig macrumors 6502

    Joined:
    Jan 4, 2015
    Location:
    Key West FL
    #18
    I didn't notice any significant JPEG compression artifacts.

    The big problem with JPEGs is that their compression is lossy with distinct artifacts and that it is impossible for any app to actually edit or reprocess a compressed file without decompressing it first. The reprocessed file would not longer be a JPEG, at least while in memory, and saving it again as a JPEG would generate new additional compression artifacts.

    It is not necessarily a bad thing that some scanning software defaults to saving as JPEG. JPEGs are easily viewed any device or computer. It's only a "bad thing" when a more advanced user scans with the intention of saving a master file for later alteration and fails to change the save format to TIFF or some other uncompressed lossless format.
     
  19. filmbufs macrumors 6502

    filmbufs

    Joined:
    Sep 8, 2012
    Location:
    Oklahoma
    #19
    Can we start making a case for why it should be photoshopped? :)
     
  20. mcblade, Jun 4, 2017
    Last edited: Jun 4, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #20
    dwig, I see. So it would be easy to tell roughly how much a JPEG was saved, even if not modified. Thanks.

    Fun fact: I found that image was last saved at 82% JPEG compression, a little less lossy than most Facebook photos today (71% is usual these days). In the old days, with lower resolutions they were saving it at 85% sometimes.

    filmbufs, what do you mean by that? You think some things should be done to it?
     
  21. mcblade, Jun 8, 2017
    Last edited: Jun 8, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #21
    It just occurred to me this could be a mixture of tungsten and flash... say there was an incandesant light source to the right of the picture, just off screen, like a regular living room lamp. That could mess up the white balance and might explain why the boy farthest to the left is 'cooler' looking, as he was farther away from the lamp. Right?
     
  22. dwig macrumors 6502

    Joined:
    Jan 4, 2015
    Location:
    Key West FL
    #22
    I doubt that this is the case. The two light sources don't truly "mix". Each would cast its own shadows and there will be color balance differences in the shadows. In this image, the "tint" on the wall and the tint in the sharp shadows on the wall match too well for the light filling the shadows to be radically different color than the flash.
     
  23. mcblade, Jun 8, 2017
    Last edited: Jun 8, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #23
    The tint on the baby's face does seem to be some point between the kid on the left's face and the kid on the right's face though. It seems to me that the further to the right you go in the image, the less the flash color temperature. What do you think could cause that?

    IMO perhaps the picture was taken too closely, and the flash hit the boy on the left the hardest, and the other people in the image didn't have as much flash on them. The picture could then have partially faded before the scan, leaving only the brightest parts with the correct color cast, the other parts being easier for the scanner to "orangify".
     
  24. kenoh macrumors demi-god

    kenoh

    Joined:
    Jul 18, 2008
    Location:
    Glasgow, UK
    #24
    So @mcblade, are you going to tell us what the score is with this image now? and the impetus for the question?

    So we can see if the experts on here are full of it or whether they are as expert as they think they are? :)
     
  25. mcblade, Jun 9, 2017
    Last edited: Jun 9, 2017

    mcblade thread starter macrumors newbie

    Joined:
    May 30, 2017
    Location:
    somewhere in time!
    #25
    Well, I would say that dwig doesn't seem to think this picture has anything that we can see from the resolution we have that shows it was modified in a non-global fashion. I think it was not photoshopped/altered significantly and I'm going with that because that's the best answer I can get with the information available. @dwig seems on the level and knowledgeable.

    As an update, I have found from another picture of these people that the "head size" thing is not an issue. Also there is a plaid something folded up on top of the back of the couch in this picture which is making it look uneven a la Mona Lisa but it's not actually part of the couch.
     

Share This Page