Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Science Technology

Scientific Publication Condemns Photo-Manipulation 85

valdean writes "According to a recent article in the New York Times (registration and short biography required) scientific journals have begun to respond to a growing problem of photo-manipulation in submitted manuscripts. At The Journal of Cell Biology, a test developed in 2002 revealed that 25 percent of all accepted manuscripts had one or more illustrations that were manipulated to the point of violating the journal's guidelines. Examples included the duplication of images for re-presentation as a control experiment, making pictures prettier with the clone stamp tool, use of the contrast tool to hide data, and merging portions of several images so that they appear to be a single image. How were many of these scientists caught? They submitted layered Adobe Photoshop files that showed exactly what they had done."
This discussion has been archived. No new comments can be posted.

Scientific Publication Condemns Photo-Manipulation

Comments Filter:
  • I don't know much about imaging techniques. But how easy is it to tell a picture's been manipulated without having the original source in hand?
    • Re:Question (Score:5, Informative)

      by faloi ( 738831 ) on Tuesday January 24, 2006 @12:24PM (#14548900)
      It depends on the skills of the person manipulating the image, and the image quality. The higher resolution the picture, the easier it is to zoom in and spot anomalies. If the picture is craptastic to begin with, it's harder to see the differences (tough to tell if the blurring is a result of someone mucking with the picture, or just poor quality).

      The biggest roadblock to telling whether an image is real or not is time, in my opinion. If you generally trust the person providing the photos and they're not too unbelievable, you probably won't spend time trying to figure out whether it's reliable or not.
    • Re:Question (Score:4, Informative)

      by whorfin ( 686885 ) on Tuesday January 24, 2006 @12:28PM (#14548939)
      In this case, they had the source, which tells me that the scientists that got caught weren't exactly the sharpest spoons in the drawer.

      Here's a prior slashdot posting [slashdot.org] about mathematical techniques to identify photo manipulation. And another article [oemagazine.com] detailing some techniques.
      • Considering they were submitting articles to the Journal of Cell Biology, it would seem unlikely that they are not sharp. More likely, they didn't think they were doing anything wrong and were just trying to enhance the appearance of their submission.
        • FTFA: While reformatting illustrations submitted in the wrong format, Dr. Rossner realized that some authors had yielded to the temptation of Photoshop's image-changing tools to misrepresent the original data.

          Because all the sharp scientists think misrepresenting experimental data is an integral part of the scientific method. Sorry, but most scientists are not brilliant Einsteins. They are guys trying to make a living and there is significant pressure to get published and the design/conduct "successful" e

          • Re:Question (Score:3, Insightful)

            Also FTFA:

            The editor of the journal, Ira Mellman of Yale, said that most cases were resolved when the authors provided originals. "In 1 percent of the cases we find authors have engaged in fraud," he said.

            So, my original statement is completely true, most of the photos were doctored strictly to make them look better, not for fraud. I never said there was no fraud, only that most of the submitters didn't care if they got caught altering pictures, because they only did it for aesthetics.

    • Re:Question (Score:4, Informative)

      by merlin_jim ( 302773 ) <{James.McCracken} {at} {stratapult.com}> on Tuesday January 24, 2006 @12:44PM (#14549124)
      Others have made good replies to this but thought I'd add my $0.02

      Generally, image manipulation will leave a signature of some sort on the file - do a fourier transform (view the image as frequency data as opposed to spatial) and you can see some of them pretty clearly. They generally show up as very low or very high frequency noise distributed more or less uniformly around the origin. Then there's edge detection; most computer-based photo manipulation creates or erodes edges and a basic edge detection algorithm will show the problem to most human observers.

      As mentioned by others, a low quality original can make it much harder to detect manipulation.
      • Re:Question (Score:3, Interesting)

        by Anonymous Coward

        I am a photographer more than I am a scientist. To what degree would the digital darkroom techniques that I routinely practice be considered fraudulent by scientific journals?

        1. Use of a little unsharp masking: this is the digital equivalent of tweaking the focus of the enlarger in a chemical darkroom-- a practice that I believe was very common in the days of chemical microphotography
        2. Use of the histogram controls and gamma adjustments: the digital equivalent of choosing films, papers, chemistries, and pro
        • IANAS but I'll answer your questions as best I can...

          Use of a little unsharp masking: this is the digital equivalent of tweaking the focus of the enlarger in a chemical darkroom-- a practice that I believe was very common in the days of chemical microphotography

          It is the digital equivalent - but unfortunately, in this case, the digital equivalent of an analogue process introduces severe aliasing artifacts, and can actually create or destroy spurious signals with repeated use - most scientists frown heavily
      • They generally show up as very low or very high frequency noise distributed more or less uniformly around the origin.

        What exactly do you mean by "high freq. noise ...distributed around the origin?" (Origin = 0hz, right)? Amplitude plots of images in the freq domain will always be symmetric since the 'input' is 'real'.

        I thought most un-altered images would have strong low-frequency components (as the image is likely to have some "average" intensity that is non-zero). Perhaps the lack of high-frequency n
    • The last two lines of the article:

      "How were many of these scientists caught? They submitted layered Adobe Photoshop files that showed exactly what they had done."

      Layered Photoshop files contain a history of the file.

  • Sad really. (Score:3, Funny)

    by NeoThermic ( 732100 ) on Tuesday January 24, 2006 @12:21PM (#14548862) Homepage Journal
    Ctrl+E or Layer -> Merge Layers is so difficult these days...

    NeoThermic
    • Re:Sad really. (Score:4, Insightful)

      by Yvanhoe ( 564877 ) on Tuesday January 24, 2006 @01:04PM (#14549327) Journal
      I do believe (and hope) that most of the photo-manipulators were simply unaware of the journal's guideline. It is in fact a gray area of the scientific publication ethics. For example if I take a phtography of a rock for a geology publication, I am surely allowed to tweak contrasts with photoshop in order to show clearly the structure of the rock. But I suppose that the exact same manipulation would be unethical if I were to hide details that could serve as a counter-example of my thesis. While in doubts, it is probably better to send a photoshop file to the journal, showing all the (hopefully) minor photomanipulations made to the image. If the journal thinks that something is abnormal, one can discuss about it, most of the time a polite conversation can probably solve these issues.
      • Re:Sad really. (Score:4, Insightful)

        by Otter ( 3800 ) on Tuesday January 24, 2006 @01:22PM (#14549528) Journal
        For example if I take a phtography of a rock for a geology publication, I am surely allowed to tweak contrasts with photoshop in order to show clearly the structure of the rock. But I suppose that the exact same manipulation would be unethical if I were to hide details that could serve as a counter-example of my thesis.

        Agreed. It sounds like, though, that most of the incidents here were more like if you included your foot in the bottom of the picture of the rock and edited it out. The editors said that only a very small subset of the violations were deliberate attempts at fraud.

  • Didn't have to register or nuthin'. I even went to the main page and checked other articles... all viewable. Glad to see they've abandoned their all or nothing policies. Though I use adblock, I left all ads intact. Nice to see that the NY Times have finally listened to the masses, and removed the requirement to register, when you view most content.
  • by mal0rd ( 323126 ) on Tuesday January 24, 2006 @12:30PM (#14548960)
    If they submitted multi-layed photoshop files, most of them probably were not concerned with getting caught. So they must not have thought what they were doing was unethical. And having the journal come up with some guidelines and a review process for images is hardly going to make a difference.

    The problem is the attitude. If they think that modifying these images isn't unetherical, then how about the data? And how will you ever catch those people? It's just a sad state of affairs in this scientific community.
    • How will they catch these people? You've missed the point - it's not the job of the publication process to catch people cheating on their data, it's the job of the scientific community - if a result is irreproducible, then sooner of later people start asking questions. Journals aren't equipped to do full in-depth analysis of papers, and referees are only judging the interest and relevence of a paper. This image analysis is of interest to the journals because it is relatively straight-forward and quick. As
    • If they submitted multi-layed photoshop files, most of them probably were not concerned with getting caught.

      Yeah, and if the US gov't publicly released PDFs of documents with the "secret stuff" censored out by merely drawing an opaque black box over it, leaving the actual classified text still present in the PDF, then they probably were not actually concerned with security.

      Never underestimate people's capacity for computer ignorance.

    • If they think that modifying these images isn't unetherical, then how about the data?

      Probably depends on how the images are modified. If I take a series of images and combine them to form one large image, I don't see a problem with that. It allows people to see the full picture at once. If I take them and erase one thing or add something that wasn't there to begin with, that is something else entirely. It all depends on the modifications.
  • Not layered images (Score:5, Informative)

    by dmoore ( 2449 ) on Tuesday January 24, 2006 @12:35PM (#14549022)
    Unless I'm mistaken, the article never states that the scientists submitted layered Photoshop images that revealed their misdeeds. I find it very unlikely that a journal would ever accept an image in photoshop format -- they usually want press-ready formats like PDF, EPS, or JPG.

    From the article, it sounds like the editors just fool around with brightness and contrast of submitted images, and that often reveals the discontinuities from an edit. However, the specifics are not in the article, so don't jump to conclusions.
    • However, the specifics are not in the article, so don't jump to conclusions.

      You are aware you're posting on slashdot, right?
    • They don't. I believe they only take CMYK TIFF's, in common with most other journals. That is unless I have got it mixed up with other cell biology journal. Which is common practice for electronic submission for journals. That's why most of my users have to use Photoshop as there are very few 16bit CMYK capable image manipulation packages in existance.

  • Photoshopping germs to look better for journals couldn't be any worse than photoshopping models to look better for magazine covers.
  • Glad this came up (Score:5, Interesting)

    by smooth wombat ( 796938 ) on Tuesday January 24, 2006 @01:01PM (#14549299) Journal
    Not because it exposed people who were fast and loose with their photos but because it brings up the whole issue of digital photo manipulation for entries. I recently submitted two photos for a local photo contest which were done from film.

    When it came time to see the submitted photos I took my parents along so they could see how my prints stacked up to the others.

    Of the roughly 30 prints that were submitted there were at least two I was sure had been manipulated and possibly one more. The one case was blatant. The submitter had done a poor cut and paste of a wood duck. The other was the merging of two photos which produced a very nice looking picture.

    As more and more people use digital cameras and then PhotoShop (or other programs) the 'art' of photography goes away since the original photo can so easily be manipulated. Unlike in traditional photography where the negative or slide is the original and any manipulation of that original can be easily seen.

    Personally I would like to see photo contests have two separate categories. One for film cameras and one for digital with the understanding that the digital photo may have been manipulated in a way beyond what traditional photos can be. Like the article I'm not talking about enhancing contrast since that can be done with different paper or chemical process for film photos but rather the addition or subtraction of wholesale items.

    Yes, adding and removing objects from a film photo can be done but it is more labor-intensive and harder to pull off than with a photo manipulation program.

    Granted, if people would just be honest about what they submit we wouldn't have to have this discussion but the same could be said about laws or notices to not do something.
    • One for film cameras and one for digital

      There won't be film cameras anymore in the future, I think. Nikon already went all digital, IIRC.

      • Just because manufacturers no longer make film cameras (well, most don't) doesn't mean that film won't still be used.

        Olympus hasn't made a film camera in roughly a decade but myself and others still use them.

        Maybe in the distant future, something like 200 years, film cameras will no longer be used but for the forseeable future film will still be with us. Not to mention slides which give better color saturation than any film can.
        • What will kill film cameras will be the end of processing chemicals. As makers phase out thier production or the cost rises too high less and less film will be shot.

          I bet it's 10 years and then the cost will become too high for 95% of film photographers.
          • Re:Glad this came up (Score:3, Interesting)

            by drinkypoo ( 153816 )
            I suspect that black and white will hang around for a long time because it's pretty inexpensive stuff. Color, on the other hand, is on the way out. It's not that there's no reason to use color film - you still can't get the resolution out of any CCD that you can get out of large-format film. But, that will change!

            Now, IANAP(hotographer) and the furthest I ever got with it was one B&W class at a community college, but it seems to me that the color quality is actually a secondary issue. It's well-known

    • I know this is probably far out, but I would like to see some kind of "signature of authenticity" feature as standard part of digital cameras. Some way that, with the signature file, a copy of the picture taken with the camera, and the serial number of the camera, either you or the camera company could determine if the image had been altered or not.

      We can already do this with simple email and public key encryption. ("signed messages" that let you see the message). I guess it's just a matter of the "general
      • I shoot mostly with a Canon EOS 20D digital SLR (the 1.6x sensor crop and smaller pixel pitch it great for extending your telephoto shots). Anyway, there is a "custom function" (i.e. a user-configurable parameter) in the camera (and all high-end Canon dSLRs) that enables "image verification". It basically digitally signs the original image, and then with the $700 Canon DVK-E2 Data Verification Kit [bhphotovideo.com], you can verify the signature. Apparently, this is used a lot in the insurance fields.

        Now, when it'll come t
    • For quite a while, we've had three categories. Black and white, color, and "digitally manipulated" (i.e. doing creative things with Photoshop, etc). Strangely, there are very few entries in the manipulated category, which I suppose shows that we've got mainly traditionalists in our group, but hey.

      I entered a time-composite of over about 90 minutes at dawn. The left edge was half an hour before sunrise, the right edge was an hour after sunrise, and by dissolving through about 30 frames, I created an even
    • by Xtifr ( 1323 ) on Tuesday January 24, 2006 @03:53PM (#14550874) Homepage
      > As more and more people use digital cameras and then PhotoShop (or other programs) the 'art' of photography goes away

      Ok, so it was a little before my time, but I seem to remember hearing that the whole idea that photography could even be an art form was rejected at first, especially by painters. Frankly, while I'm willing to concede that photography most certainly can be art, it seems to me that digital image manipulation provides at least as much, and possibly far more, room for artistic expression.

      Seems to me like there's a bit of poetic justice here. (Is poetic justice a form of art?) I bet there's a bunch of dead painters who would (if they could) be rolling around laughing at the irony of a photographer complaining about people who take advantage of technology to make their art "too easily". :)
      • Seems to me like there's a bit of poetic justice here. (Is poetic justice a form of art?) I bet there's a bunch of dead painters who would (if they could) be rolling around laughing at the irony of a photographer complaining about people who take advantage of technology to make their art "too easily". :)

        Sigh, I never have mod points when I need them...Insightful!

      • > Ok, so it was a little before my time, but I seem to remember hearing that the whole idea that photography could even be an art form was rejected at first, especially by painters.

        Yeah, until they started seeing pictures of naked ladies. Then most of them ran out and bought a camera.

    • The "art of photography" actually has a long history of manipulation. Photos of ghosts -- multiple exposures in which one of them had a person in white or bright clothes (or a sheet) -- were the first active deceptions. In some ways, it's far easier for a novice to pull this sort of trick than a digital overlap. With a ghost image, it's right there in the negative, after all.

      But the "art" takes a whack from every significant step forward in technology. As operating systems got easier to use, noobs got more
    • "the 'art' of photography goes away since the original photo can so easily be manipulated. Unlike in traditional photography where the negative or slide is the original and any manipulation of that original can be easily seen."

      A good part of the "art' of photography has always been in the processing of the print! For instance a lot of Wedding photographers still use film because they have labs that can process the film just right. All you need to do is take some professional negatives and have them develope
    • Personally I would like to see photo contests have two separate categories. One for film cameras and one for digital with the understanding that the digital photo may have been manipulated...

      That's actually a really bad way of distinguishing whether a photo has been manipulated. Digital photographs can be presented 'as is' (as much as digital can be), and film photos can still be manipulated. A lot of things can be done in the darkroom. And then there's people like myself, who have 'gone back' to film

    • Yes, adding and removing objects from a film photo can be done but it is more labor-intensive and harder to pull off than with a photo manipulation program.

      Umm, no. You *ARE* aware that modern photo-printers, atleast those that normal mortals use generally work by *scanning* the film, and then *printing* the resulting file the same way they would print a file coming from a digital camera ?

      The fact that one file comes from a CCD, and the other file from a scan of a negative makes no difference for the d

  • they meant the brush tool.

    ...I'll get me coat.
    • as insurance stops covering more and more procedures, the X-rays get touched up rather than the operation gets scheduled. when peer review and editor review finds these things, they ought to alternate-page the publications... left side, the author's stuff, right page, the evidence discovered that, ahhhh, suggests a lack of evidentiary demonstration and perspicacity on the part of the authors. (because none dare call it fraud and weaselness.)
  • by wsherman ( 154283 ) * on Tuesday January 24, 2006 @01:07PM (#14549360)
    From the summary:
    ...violating the journal's guidelines. Examples included ... use of the contrast tool to hide data...

    The actual article is a bit more nuanced:

    To prohibit such manipulations, Dr. Rossner and Dr. Mellman published guidelines saying, in effect, that nothing should be done to any part of an illustration that did not affect all other parts equally. In other words, it is all right to adjust the brightness or color balance of the whole photo, but not to obscure, move or introduce an element.

    If a researcher manipulates only part of an image then the researcher is implicitly admitting that there was something in the image that they chose to ignore. On the other hand, if a researcher changes the contrast of the whole image to make it easier to see the patterns they are drawing their conclusions from, then they can always claim they really didn't notice the other stuff. Essentially, researchers have to avoid doing things that prove that they were deliberately dishonest.

    • On the other hand, if a researcher changes the contrast of the whole image to make it easier to see the patterns they are drawing their conclusions from, then they can always claim they really didn't notice the other stuff. Essentially, researchers have to avoid doing things that prove that they were deliberately dishonest.

      Don't discount the possibility that there was nothing else interesting to see either. Scientists are not all dishonest!

      Disclaimer: I am a scientist.
    • I work in professional medical advertising and those are the rules we have to live by - we can adjust some color and lighting in case studies to compensate for time, lighting, the particular camera and other variances but that's all. The medical reviewers where I work are draconian on this matter for good reason.

      On the other hand, I can see why a researcher might do this, given the fame, pressure to produce and the dollar amounts involved. This is similar to the Korean DNA/cloning scandal - He fudged hi
    • If a researcher manipulates only part of an image then the researcher is implicitly admitting that there was something in the image that they chose to ignore. On the other hand, if a researcher changes the contrast of the whole image to make it easier to see the patterns they are drawing their conclusions from, then they can always claim they really didn't notice the other stuff. Essentially, researchers have to avoid doing things that prove that they were deliberately dishonest.

      Having worked in a field whe
  • Moral of the Story (Score:3, Insightful)

    by Kelson ( 129150 ) * on Tuesday January 24, 2006 @01:26PM (#14549567) Homepage Journal
    Always flatten your images before submitting them.

    (OK, not really, but you know some people of less-than-sterling ethics are going to walk away with that instead of the real lesson, i.e. don't fudge your data.)
  • I think a possible solution is to have an invisible watermark like what could be done steganographically. The photo industry could repsond by making cameras that encode a stego watermark on some or all of their cameras. A journal could require that 1. submissions require stego watermarks. and 2. Annotations be provided in a separate layer that can be overlaid at press time.

    Unfortunately, I can think of ways to get around this like rephotographing a manipulated image so I don't think you could stop a dete
    • Unfortunately it also raises privacy issues. While you're talking about implementing the quasi-steganographic approach for a limited subset of camera models (I'm assuming here we're talking about the CCDs used to photograph off microscope feeds) I'm sure it wouldn't be too long until the watermarks started appearing in consumer camera models.

      A little less tinfoil-ish consideration is that any watermarking done to the image is necessarily a manipulation of the image, and doing it on the hardware level pre

  • Rubbish (Score:4, Insightful)

    by bvwj ( 473084 ) on Tuesday January 24, 2006 @01:50PM (#14549803)
    All photographs are manipulations. None capture the exact data represented by the obect in real life.

    Before photographing the item it lit, framed, lensed, exposed, and captured at a point in time. Plenty of room for manipulation.

    So what's the difference between physical manipulation before the capture and digital manipulation after?

    There is only honesty and dishonesty. Manipulation is a given.

    • The difference is that it's an entirely different kind of skill and one that should in all fairness be judged differently. Besides, most photographs worth judging aren't set compositions outside of where the photographer stood, which way they were facing, and what lens, aperture, film, and shutter speed they used. They don't have control over lighting aside from what time they take the photo.

      The real manipulation in such photographs comes at print time. I'm not sure how that works with color - I didn't ta

  • by davidwr ( 791652 ) on Tuesday January 24, 2006 @02:07PM (#14549933) Homepage Journal
    Altered photos are a good thing IF they are clearly labeled as "photo-illustration," "enhanced photograph," "composite photograph," or the like, with the original photos made available to the publisher and the peer reviewers or better yet to the general readership.

    When studying a scientific photo, I don't want to be distracted by less-interesting bits of information - I want my attention drawn to the point the author is trying to make.

    You only have a problem if the reader thinks he is looking at one thing when in fact he is looking at something else.
    • I've included somewhat complex images in a scientific paper before. I called out specific portions of the image by circling, labeling it, or both.

      This is a scientific paper -- it should include unmanipulated data to let others come to their own conclusions. Cropping is one thing, since people reading the paper will know that there will be data outside of the field of view. Manipulating within an image is an entirely different proposition.
      • That's all fine and well, but the degree of asininery that goes on with peer-reviewed publications can be just unbelieveable sometimes. I submitted a paper that was rejected from 4 different journals because I drew conclusions that was directly opposed to the current thinking (some really obscure molecular cell biology; nothing earth-shattering). Nevermind that the previous data was putting chicken peptides into frog cells, whereas I put human peptides into human cells. Apparently, my conclusions were wrong
    • Yes, I can see your point. There are some cases where an image may be hard to see, and some degree of manipulation may make things much clearer. Personally, I'd have no problem with it if it was clearly labelled, and there was a URL to a website where I could directly get the straight raw images. I understand that in some cases it may be infeasible to print a full explanation with all source images in an article for print, but in this day and age there is no reason to not make all information readily ava
  • Old News (Score:2, Interesting)

    by Arandir ( 19206 )
    New news in the scientific community, but very old news in the news community. You can never believe any picture you see in the news media. Here's a recent example: http://michellemalkin.com/archives/003780.htm [michellemalkin.com] . The media (both big and small) have been repeated caught photoshopping their images. One of the great advantages of the blogosphere is that this sort of stuff gets found out very quickly.
    • One of the great advantages of the blogosphere is that this sort of stuff gets found out very quickly.

      And one of the great disadvantages to the blogosphere is that it's called the blogosphere.
    • You can never believe any picture you see in the news media.

      And sometimes it's more blatant than Photoshopping. An Agence France-Presse journalist used a fisheye lens [take-a-pen.org] to exaggerate damage done to a terrorist's house by the IDF. Here's my JE about it. [slashdot.org]
  • by Goldsmith ( 561202 ) on Tuesday January 24, 2006 @03:28PM (#14550674)
    In microscopes I have used (of the electron and scanning probe variety), often automatic image processing is used when you save an image. We've had to be very careful to make sure that we have disabled anything we can, and in some cases, have had to scrap software and write our own to make sure we weren't losing data.

    Really, the problem is that there isn't room in a scientific paper to put dozens of images proving your point. When you're given 3 pages, you select the best image or two that you have. Ultimately, an editor or reviewer can ask for supplemental information if there is any doubt in the result. Perhaps the problem is that reviewers aren't asking to see (or people are not providing) the supplimental data which a good scientist uses to determine truth.
  • ...for the need to have media like film. It's alot harder to manipulate a piece of film than it is an image file. Sad to see film seems to be on the way out.
  • Sharpening up images is one thing if it's needed to highlite a point (but in a scientific journal, there needs to be text explaining the manipulation).

    However, if there are major manipulations to an image that border fraud, one has to wonder how many of the numbers are pretend.

    That really was bigfoot in the photo.
  • by Stan Vassilev ( 939229 ) on Tuesday January 24, 2006 @04:48PM (#14551364)
    Today Adobe releases the 10-th release of its popular program "Photoshop". The hit feature in this release is that the software maker has forbidden all sort of image manipulation, answering concerns from the scientific community:

    "There were numerous reports about photo manipulation in manuscripts from the scientific community. Few years back, when the government asked us to forbid opening images with scanned banknotes in them due to possibility for money conterfeiting, we responded and implemented the appropriate algorithms to comply. With this release, we're just taking the next step."

    Among the features unaffected in Photoshop 10 remain zooming in/out and panning.

    According to Adobe, Photoshop 10 will be available to purchase at retailers world-wide early next month.
  • Comment removed based on user account deletion

Get hold of portable property. -- Charles Dickens, "Great Expectations"

Working...