Forum Moderators: not2easy
The client now wants to go with Digimarc, an option which we have previously discussed, but let idle. I've run through the WebmasterWorld search on the subject, and still have a couple questions.
First, what is involved in defeating a watermark? I've read several comments that this can be done, but haven't seen any insight as to an expected level of defense. Does it take an expert? Or can any fool do it?
Second, does anyone have experience with an average required image size? Many of the images involved will be product shots that definitely won't meet some of the minimum size requirements that I've heard bandied about. (What do watermarks do to file size?)
This client has all his images custom shot, no matter how minor, and while it's not Ansel Adams, no competitor can touch him. We do monitor for thieving competitors, and it's not a surprise that we caught one. The problem is that the next guy might do a better job of doctoring, just good enough that we can't risk an accusation.
What I've been looking for is a way to encode the image using something like stegenography (sp?) then set up a method for searching for the images with my code embeded in them.
This way the image is clean when displayed, but still identifiable as mine. From what I understand even if you change the format of the image the code still remains (depending on which method you use) because it's the pixel values that are modified not the file atributes.
I have not been able to find a robot that will do this for me, but so far I have a script that will crawl and I have a perl script that will encode and decode. All I have to do now is find a way to put them together.
The deathly silence to this thread says something itself. (Though I haven't decided what.)
For now I'm on hold, and hoping that our known watchfulness of the market place is something of a deterrent in itself.
First, the watermark will be applied with ImageMagick. I want to cut down on production time having to use photoshop to insert watermark on every image, thus applying the watermark on server side.
Problem is that a clever user can possibly find the image folder and download the image sans watermark. Solution is to put a folder above the www (public_html) folder, which means that they cannot access the folder directly. If that isn't possible an alternative would be to use SSI to rename the image before posting.
Does this make sense? What do you think?
I also like the ImageMagick idea and I'm playing around with that right now. So far the IPTC headders seem like the best way to identify an image IF you can get them back out, and IF someone didn't think to check them and remove them with PhotoShop.
If you are truly worried about image theft and don't mind paying then I would say go ahead with Digimarc.
I do like the sequence of logic behind your thinking. If it stands up to application, I like the sound of it a lot. Even if it can still be defeated, it seems that the level of sophistication required will eliminate virtually anyone that I would be concerned about. Most of these images don't have any value beyond their narrow market. There should quickly come a point where they aren't worth stealing. Thankfully, the majority of our competitors are decent people.
If you carry this through, I'd appreciate your sharing some stickies with me. Maybe I could sail along on your coat tails, and learn me something too.
I'm not interested in preventing people from downloading images and believe if they really want them they are going to acquire them. My focus is to help out a photographer friend who has a site with a massive number of images. He'd like to find a way to get people to pay for personal use. (Commercial use really isn't a problem because 99.9% of the companies are legit and would rather pay a fee than be taken to court.)
The idea is when they purchase the photo for a couple of dollars they can download the image without the watermark.
As I researched I found that ImageMagick has several advantages over GD when it comes to quality and flexibility. The only hitch I ran into was that my server does not have IM installed for PHP, which means that I had to learn some Perl (oh joy!).
If you are good at programming the documentation for IM probably is okay. For someone like me who does not enjoy programming the documentation leaves a lot to be desired. The best resource I've found are two books by Shawn Wallace and are published by O'Reilly.
The earlier version of Shawn's book, "Programming Web Graphics" is written more for the web designer, while the updated version has more Perl scripting and aimed more at the programmer. Either one will help in getting the job done.