Copyright proof for digital images

Discussion in 'Digital Cameras' started by Alan Browne, Jan 4, 2004.

  1. Alan Browne

    Alan Browne Guest

    One very sure way to prove copyright of an image originally captured on
    film is possession of the negative or positive film itself.

    How so with digital? EXIF's can be edited easilly enough. For example,
    if I rip off an image from Bret Douglas, like the full detail hawk image
    he posted recently on pbase, I could use a binary editor to change the
    date and whatever other little details may please me and then claim it
    as my own.

    If there were embedded codes in the image data itself (a la Digimarc),
    will editing in photoshop lose/damage the watermark?, playing with
    brightness, contrast, USM... etc.

    So how would Bret prove that it was his image? Let's avoid the obvious
    like resemblence to his existing work on that poor hamburger-obese hawk
    (although that would likely convince a court, there could be other shots
    where he would not have comparable evidence ... eg: at a football game,
    I could show lots of my own images that resembled his).

    Alan Browne, Jan 4, 2004
    1. Advertisements

  2. Alan Browne

    Nick Zentena Guest

    Show it was published some place first? Other then that isn't it on a
    balance an issue of who is more believable?

    Nick Zentena, Jan 4, 2004
    1. Advertisements

  3. Bluntly, there is no way. The same risks exist when you send out high
    res digital files, or slides, or prints. It is possible record digital
    files to film if you wanted to. Digimarc can be removed.

    The best protection is two-fold. One, never take just one image. This
    gives you dupe slides (good in case an editor loses yours, happens all too
    frequently), and provides additional similar images that can establish your
    proprietorship in court, if necessary.

    Two, register the images with the copyright office. This can be done
    in bulk for just one fee (at a time), and is the strongest protection you
    have. With it, court cases are a lot stronger and can provide additional
    recompense (such as court fees and so on).

    Date, time, and location can help, in extreme cases. Sun angles can
    be plotted and determine if the pic really was taken on that day, should
    definitive landscape details and shadows exist.

    But it is a serious question, and I'm waiting for the day when
    digital images are called into question in court. This isn't the risk it
    seems at first, since all evidence photography is treated the same as any
    other evidence, requiring responsible trained parties, handling, storage,
    and specific documentation ("chain of custody"). Thus in most cases,
    manipulated digital images presented in court would require the complicity
    of the police or feds. There is also some suspicion in my mind that this
    has already happened...

    - Al.
    Al Denelsbeck, Jan 4, 2004
  4. Alan Browne

    Chris Brown Guest

    This is another advantage of shooting raw mode - only ever give JPEGs,
    TIFFs, etc out and keep the raw files yourself.

    OK, with some work, a JPEG could be reencoded as a CRW, NEF, etc., but
    faking such a file to the required standard would be hard work, and if the
    fake is taken from a JPEG, the real original raw file should have shadow
    detail, etc. that isn't present in the fake.
    Chris Brown, Jan 4, 2004
  5. Alan Browne

    PTRAVEL Guest

    Of course, the best proof of ownership of copyright is a copyright
    registration. Failing that, proving ownership is the same as any other kind
    of legal proof -- witness testimony, circumstantial evidence, etc. There's
    nothing magic about digital intangible property -- it's handled just like
    anything else.
    PTRAVEL, Jan 4, 2004
  6. Alan Browne

    Mark Roberts Guest

    This brings up an interesting point: Isn't it possible for software to
    convert a TIFF image back into a RAW file? I don't think any such
    software exists... yet. I'll bet it's only a matter of time until
    someone does creates it, though.
    Mark Roberts, Jan 4, 2004
  7. Alan Browne

    Ken Weitzel Guest

    Provided the future value of the pic could perhaps be high
    enough one day to be worth an investment of a dollar or two...

    How 'bout just printing a real good copy of the original,
    putting it in a well sealed evelope and mailing it to yourself
    via registered mail?

    Then don't open it, just put it away as is. At least that
    way you'd have clear evidence of the date you first possessed

    Take care.

    Ken Weitzel, Jan 4, 2004
  8. Alan Browne

    Alan Browne Guest

    All digi's could have GPS receivers and the data could include the time
    (GPS time is quite accurate), location, sat emphemerides and pseudo
    ranges (to validate the location data), etc. Wouldn't work indoors and
    some other locations, however... GPS receivers do suck batteries, alas.
    Alan Browne, Jan 4, 2004
  9. Alan Browne

    Canopus Guest

    This was exactly what I was going to post, in fact post a batch of your work
    to yourself before you ever place them on the web. Better still, if you
    have a solicitor then post them to him with a note on the envelope that the
    letter is from you and not to open it until authorized by you. Remember,
    everything you create is automatically copyrighted, it's just a case of
    proving you did it first.

    Canopus, Jan 4, 2004
  10. There are several things that come to mind with photographs of birds,
    people, or whatever. One way is with all the other images you made of that
    same bird. The thief would have only the one, and you as the photographer
    would have a half dozen or several dozen as the case may be.

    Another way is that the bird or person is in some distant place; you have
    tickets, receipts, and other proof that you were at that place at the time
    the photo was taken. Presumably, the thief has only the photo and no way to
    prove how or when he took it. Your testimony in court, if it comes to that,
    should easily persuade a judge or jury that you took the photo, especially
    if you have similar images made that day and receipts; even without
    receipts, similar images and your testimony about the circumstances will be
    very persuasive when contrasted with the testimony of the thief who has no
    idea when and where he took the photo, no similars, no receipts or other

    If it's a bird, the bird can't testify, but if it's a person, that person
    may remember that you took the photo. If the bird is captive in some zoo or
    other, it may be that in addition to receipts, you have people who remember
    you. Louise and I generally give people our cards when we photograph them,
    since we publish photos on our Web site. The people get to see the photos
    online, and we offer to email best quality files if they wish. If we take
    pictures of a restaurant, hotel, or whatever, we get permission first and
    give our card to the owner or manager who gave permission.

    If we have landscapes or scenes or animal photos, we have tons of similars
    in our files, either slides or digital. I would be very comfortable going
    into court to testify that a particular image is mine, even if it was
    originally digital.

    We also generally put a copyright notice in each online image. The thief
    may display it with our copyright notice intact, or he may crop it out or
    smudge it. In the latter event, the image will be noticeably changed from
    the original, and the crop or smudge will be right where the copyright
    notice is. The thief will then have to explain why that is, in addition to
    explaining how he came to be in possession of such a photo.

    By the way, copyright exists independently of the film or other medium. It
    may well be that I come into legal possession of film without holding title
    to the copyright in the work. Conversely, a thief may steal my
    transparencies, but he will not then have copyright in the images. I
    understand your context in the assertion that proving copyright is easy
    when you have the film, but possession of the film does not _necessarily_
    mean that the possessor holds the copyright as well. Possession is one step
    in a chain of proof.
    Phil Stripling, Jan 4, 2004
  11. Alan Browne

    Tony Spadaro Guest

    I never put an entire frame on the web -- so if it is stolen I have the
    full frame version to prove that I took the picture. Since I'm not into
    being published I never send anything to magazines etc - but people who do
    are always taking a bit of a risk (with slides at least) as there is no way
    to withhold part of the image - I suppose having several other shots on the
    roll of film might be helpful though.
    Tony Spadaro, Jan 4, 2004
  12. TWIAVBP, but let's assume U.S. law applies.

    If Bret's trying to prove ownership, as opposed to a prosecutor, the
    standard of proof is only preponderence of the evidence. So all he
    has to show is that it's more likely that he owns the shot than not.

    There are lots of ways to accomplish this. Bret could produce a
    series of shots of the same subject matter, or have witnesses testify
    they saw Bret taking the shot. He could show prior publication. He
    could point out unique features of the background to establish where
    the shot was taken, and then place himself at that location through
    testimony. He could also introduce evidence to attack your claim,
    such as lack of ownership of equipment, lack of physical evidence, or
    different style. Either of you could question the other on the stand
    under oath, and the chances are the 5th amendment would not apply.

    No doubt I'm missing a couple of obvious ways; I'm a little rusty at
    this sort of thing. But this sort of determination of fact is exactly
    what the civil legal system is set up to do.
    Michael Benveniste, Jan 4, 2004
  13. Alan Browne

    Alan Browne Guest

    I think I stated above "let's avoid the obvious like resemblance..."
    Agreed. OTOH, there is nothing like having the phusical evidence
    (slide, negative) to say, "see it's mine!"
    Alan Browne, Jan 4, 2004
  14. You could always bring a print to a notary public and have them notarize
    it....the date they stamp on the back would be proof that it was yours.....
    William Graham, Jan 4, 2004
  15. Alan Browne

    Jon Bell Guest

    Or burn a bunch of originals on a CD and mail it to yourself as described.
    Jon Bell, Jan 4, 2004
  16. Alan Browne

    PTRAVEL Guest

    All that would be is evidence of creation, and would be weighed like any
    other evidence. A copyright registration creates a _legal_ presumption of
    ownership and validity, meaning the burden of proof shifts to anyone
    challenging ownership, i.e. if you sue someone for infringing your
    copyright, _they_ would bear the burden of proving either independent or
    earlier creation.

    Copyright registration is cheap and easy -- it's a single form and, if I
    recall correctly, about $20.
    PTRAVEL, Jan 5, 2004
  17. Alan Browne

    PTRAVEL Guest

    One thing everyone seems to be forgetting: independent creation, i.e.
    without copying, is non-infringing. I was in Pisa last month and took a
    number of shots of the Leaning Tower. Because of the way the site is set
    up, there are very few locations from which you can get a really good
    picture of the entire tower. I guarantee you that there are tens of
    thousands of photos that are darn near identical to mine. They don't
    infringe my shot, nor does my shot infringe theirs.
    PTRAVEL, Jan 5, 2004
  18. Yes, but since there are so many of them, they aren't worth anything,
    William Graham, Jan 5, 2004
  19. Alan Browne

    Chris Brown Guest

    It would be possivle to manufacture a, say, CRW from a TIFF, but I don't
    think it would help, the fake would be easy to spot. Making a usable image
    from a raw file is lossy - it involves demosaicing, colourspace conversion,
    white balancing, converting from linear to logarithmic space, and almost
    certainly sharpening. The published TIFF would most likely be 8 bits as

    ost or all of these processes cannot be reversed in such a way as to
    get the precise input data back, so while you can certainly concieve of a
    TIFF2CRW type program, the CRW it produced would not stand up to scrutiny
    against the original CRW. The shadow detail would be missing, attempts to
    reverse the sharpening would lose detail on edges, you'd probably get
    posterisation when you reverted to camera colourspace and white balance, and
    remosaicing the image in such a way as to get anything like vaugely
    convincing data out is going to be "fun".

    In short, converting a display-ready TIFF back to CRW, or other raw formats,
    is going to result in a CRW that, when converted back to TIFF or JPEG, will
    give results which would probably best be described as "crappy" when
    compared to the genuine CRW.

    Furthermore, taking the specific case of CRW, there are more pixels in the
    CRW file than are in the converted files (unless you use dcraw), so the
    converted image, even if "uncropped" is missing some pixels (and therefore a
    small bit of the scene) that the original file contained.

    You also have to know *exactly* how the camera creates its thumbnails from
    the sensor data if you want to do a really convincing job, bit for bit,
    otherwise analysis of the file by someone who does know will show it to be a
    fraud, even before you look at the question of image quality.

    So in summary, I think we're safe on this front - a fake raw file might look
    convincing at a glance, provided it wasn't compared in detail with the
    original, but I doubt it would stand up to any kind of close scrutiny.
    The software sort of exists already - it's sitting in the cameras in the
    shape of firmware. You could do a quick and dirty job by ripping it from the
    camera, and then just feeding it the appropriate red, green and blue data
    from the TIFF pixels, and discarding the two values for each pixel that
    don't count. Problem with doing this is that the output will end up having
    only a fraction of the resolution of the original, because mosaic sensors
    don't actually work the way "George Preddy" thinks they do. ;-)
    Chris Brown, Jan 5, 2004
  20. Alan Browne

    Guest Guest

    In order to be de-constructable, the TIFF would need to have a
    bit-depth three times larger than the source RAW file. So I doubt
    it'll happen. But then, some folks *do* counterfeit one-dollar bills

    Guest, Jan 5, 2004
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.