Posted: August 14, 2012
For years people have said that watermarks are the domain of the hobbyist or paranoid semi-pro. But the rampant online spreading of photos with no way of knowing who took the original has opened my eyes to the value of a well-designed (and placed) watermark.
The real world examples are numerous, I'll share one from last week: I've been trying to license a handful of images that I've found on facebook pages that had no credit or copyright information. I contacted the owner of the facebook page to ask him for credit info for the pictures in question (I knew he wouldn't have it, but wanted to make a point), and his response was that he usually doesn't know because he "gets them from other sites with no info".
So I began my hunt to try and find the original creator of the image. Google's similar image search is helpful, but if a photo has gone viral, Google will find countless similars but they're all just tumblr reblogs and pinterest repins.
As it is obvious that the cat is out of the proverbial bag when it comes to unauthorized sharing of images online, I think the best thing you can do is create a watermark or graphic that will travel with the photo as it hops around from twitter to tumblr to facebook to pinterest and back again.
Of course, there are people who are working hard to make it easier to remove the watermark. A quick Google search for "how to remove a watermark" yielded over 45,000 results.
What are some other ways that you can leave a digital fingerprint? Any great examples of watermarks that are well-designed?
01. Marc
August 15, 2012As an anti-theft device for photos, watermarks still make infinitely more sense than thinking one’s photos are safe on one’s site if right click is disabled.
Foolproof? No.
Smarter? Yes.
Less annoying? Definitely!
02. Ed
August 15, 2012What about the embedded metadata – is that stripped from properties? Also, Digimarc – a pixel used as an embedded digital watermark rather than something over the image – might be a good solution.
03. Nicole
August 15, 2012Definitely been thinking about this a lot lately. I have an image on Tumblr with about a thousand reblogs and then who knows how many times it’s been reblogged and pinned and posted from there. Every time I see it pop up somewhere it burns me that I didn’t add something to link it back to me. Now it’s out there and maybe it will never return. I also saw one of those great images from Theron Humphrey of Maddie used as one of those dumb text memes on Facebook the other day. Felt bummed to see it used that way and even if there was a watermark it would have maybe been cut out. Le sigh.
04. Jan
August 17, 2012I don’t think visual watermarks are the answer. The problem statement above is not about copyright protection, but about the ability of any layperson to easily and reliably identify the rights holder for whatever reason.
Visual watermarks may be a deterrent to some (has been debated at nauseum), but they also interfere/deface the most valuable asset we have – our creative product.
The problem of identifying the rights holder is a data & business problem, not a visual & creative problem, and thus the solution should be too.
Luckily all the images in question, and the vast majority of all images today are in digital form. And any digital image exists in the form of a file, which has headers and various form of meta data attached. Whether the layperson actually has the file directly, or views it via the proxy of a webserver/browser doesn’t matter, the information is retrieved from a file that could easily store all this information and more.
And in fact, those files usually do contain a lot of meta data already. There are EXIF and IPTC standards for storing technical, workflow, and rights data for image files in JPG, TIFF, and most other image file formats. Lightroom, Photoshop, and most image editing software have extensive functionality to properly populate and manipulate this data through efficient presets. And most image viewers and even operating systems have built-in support for reading these meta data.
The only two missing pieces are integrity of the data stream from owner to consumer – making sure that existing meta data are not stripped during transmission and storage, and the ability of reading meta data of images displayed in a browser.
Unfortunately most online platforms, such as Facebook, post process image files for their platforms and in the process strip meta data. And after downloading an image from Facebook (and other similar platforms) and inspecting the meta data, the information is lost. Also, just viewing an image in a browser doesn’t make it easy to see the meta data.
But these are very solvable problems. With a little bit of industry awareness and lobbying online platforms can be improved to retain critical rights information. Similarly browser developers and webserver developers could easily enhance the software to make it easy to query rights information for images displayed on any website if that information is present in the images meta data.
There are digital watermarks (digimarc is well known, and comes pre-installed in every copy of Photoshop). But they’re not cheap – a one year subscription for a small business runs $499 for watermarking 500 images. And the software to read the digital watermark is not universally installed.
Again the problem statement was that any lay person looking to possibly license an image could easily establish and contact the proper rights owner. That means the mechanism to do so has to be low friction (i.e. free, and w/o complicated software/plugin installs) and ideally universal and often pre-installed. Think Adobe Reader.
That leaves operating systems and web browsers as an ideal common denominator for solving this data problem.
Digital watermarks are great if you need expert testimony in a copyright infringement case, but that’s a different scenario.
05. Tony Novak-Clifford
August 20, 2012No matter how rigorous your workflow is when it comes to embedding meta-data in your image files, many social networking sites (Facebook & Google Blogger just to name a couple) automatically strip that data when images are uploaded to their site. Watermarking images is hardly foolproof when it comes to image protection, however it is the most cost-effective & efficient method readily available to most of us.
06. Patrick Krolis
August 20, 2012I think watermarks are the way to go and make a lot of business sense too. If I upload an image to the internet, I assume that it is going to be copied, blogged, tumblred and pinned on pinterest, shared on facebook and the likes.
The sole reason why a photographer would allow that to happen is to increase the chance for the images to be discovered, and one day be asked for a formal license to use the image commercially, or to be offered an assignment because the images impress somebody.
adding a watermark simply a recognizable and google friendly name will allow for the photographer to be found more easily.
My 2 cents
07. Kyle G.
August 24, 2012If you put a photo on the internet, it will get stolen. Period, end of story.
I think social media sites like Facebook and Tumblr strip photos of metadata and change file names upon upload so they can cover their ass if they ever get sued by a photographer. They can blame the users and wipe their hands of the matter.
Also, it terrifies me that Pinterest allows users to embed pinned photos into other websites.
08. Mark E Tisdale
September 2, 2012I didn’t include any sort of watermark for years but recently came around to including a small border with my name in it. I consider it a calling card more than a deterrent. There’s only so much you can do short of not uploading online.
People who are determined to remove the content creator from the picture (so to speak) will crop or otherwise remove any watermark that’s the least bit subtle. And if it’s not subtle, that’s all anyone sees. When the watermark goes straight across the subject of the image in 45 point font, you might as well not have bothered putting it online in the first place…
For me, the border was for the honest people. It doesn’t cover the image and if they share it forward, hypothetically someone else can still locate me.
That said, I wish Google et al would modify the reverse image search to make it easier to find the FIRST occurrence of an image. Not just for the task of tracking the artist down but because that is ultimately the site that should be given preference. If we upload an image to our sites and then it gets tumbled or pinned or whatever, those are the copies, they shouldn’t appear above the original in search results. I know there are probably instances where the content appears somewhere other than the creator’s domain first, but I still think giving preference to the originating site would help more often than hurt.