Watermarking FAQ from Digimarc

What is a watermark?

A Digimarc watermark is a special message embedded in an image -- a photo, video or other visual content. Digimarc's software embeds these messages by making subtle changes to the brightness of the pixels in an image. Unlike a printed watermark, the Digimarc watermark is imperceptible to the human eye. While Digimarc watermarks are imperceptible, they also need to be robust. That's why one of our patented innovations is the ability to make Digimarc watermarks "perceptually adaptive." They conform to the unique pattern of each image, placing more energy in areas with rich detail and less energy in flat areas. This process ensures that your watermarks remain strong without being easily detectable visually. The watermark can be read by a computer or other software application. It's easy to create and detect Digimarc watermarks since our technology is already bundled in millions of copies of image editing and asset management applications. Because Digimarc watermarks are carried by the image's pixels, they are independent of the file's format. They can survive normal image edits, file format transformations, copying, printing and scanning. Images that have been watermarked look untouched. This is because the watermark is buried in the image's pixels; it's not a separate file header, wrapper or tag. At high magnification, the code appears as a slight change in the image's grain or texture.

What's in a watermark?

Digimarc watermarks can contain the following: Digimarc ID, Distributor ID, copyright year, image/transaction ID number, and a set of image attributes. Image attributes include Restricted Use, Adult Content and Do Not Copy. The technology embeds a Digimarc ID into your image which links to the web destination of your choice or a contact profile in Digimarc's on-line database registry. We only embed a small amount of data for two reasons. First, using only an ID number keeps the embedded message very small so it can be repeated many times throughout the image, making it more robust. Second, you can change the destination or contact information linked to your ID without re-watermarking your images. That ensures your watermarked images will always be current, giving you the flexibility to change marketing programs, web destinations, or even your company name without having to re-process your images.

Why Watermark?

There are three primary reasons to watermark images: copyright communication, enhanced asset management and increased e-commerce opportunities. The keys to enforcing your image copyrights on the Web are communicating your ownership, and being able to track your images and how they're used. Digimarc digital watermarks are a great way to accomplish both goals. They embed information in your images that links to you or your company's identity and can be tracked across the public Web. Digital watermarks are also easy to create and read. There are no visible bylines or tags to interfere with image quality, however if would like to add visible watermarks you can use our SDK to embed your logo file visibly on each image. Each Digimarc-watermarked image carries an ID embedded directly into the pixels of the image. That ID leads to anyplace on the Web-your home page or another destination you choose. By following this link, viewers learn that you are the owner of the image. You then have the opportunity to offer them product information, a licensing agreement for the image or simply an easy means to contact you. The result is that you communicate your copyrights, retain attribution of the image as its creator and have the opportunity to generate incremental revenue and brand recognition through your images. Digimarc-watermarked images also deter infringement. Whenever a watermarked image is opened or scanned into an application enabled with a Digimarc reader, the user is reminded--via a copyright symbol in the image's title bar--that your image is copyrighted. The watermarked images can also be tracked across the most highly trafficked public areas of the World Wide Web. Our MarcSpider(r) service scans the indexible Web identifying your images and their copies and lets you determine whether the images are being used appropriately and with permission. (Please note that our searches are limited in scope by the indexing capabilities of search engines.)

What images can you watermark?

The kind of images that you can watermark depends on the Digimarc embedding tool you are using. If you are using the Digimarc Plug-ins bundled with many popular image editing and asset management applications (http://www.digimarc.com/support/csidtutor.shtml), you can watermark any of the file formats supported by the host image editing application. The Digimarc Plug-ins currently ship with Adobe, Corel, Micrografx, Cerious, and Jasc products. The Digimarc technology is also offered by web and server-based solution providers such as www.creativepro.com, www.workstation.com, and www.xat.com. If you don't intend to apply any compression or other image edits, the image can be as small as 100x100 pixels. However, if you believe your image might be compressed, cropped, rotated, scaled or otherwise modified, we recommend a minimum size of 256x256 pixels. There is no upper limit on the size of an image that can be watermarked. Watermarking your images does not add data to the image file, since it only makes very small changes to the luminance of pixels. However, watermarking can change the size of a compressed image slightly, but again the size of uncompressed images doesn't increase. Non-lossy compression such as Compact Pro, Stuff-It and the .ZIP format do not affect survival of the watermark because no data is lost. Lossy compression methods such as JPEG actually remove some image data, which generally degrades image quality. This can have varying effects on survival of the watermark. Generally if you favor image quality over image size when selecting the degree of compression, you'll get the most favorable results. In order to watermark a JPEG image, it must be uncompressed and then re-compressed after embedding. In general, you'll get the best results by starting with an uncompressed image. The Digimarc watermarking process is color space independent, so you can embed a watermark into RGB, CMYK, grayscale and LAB images.

Can a watermark be removed?

Digimarc does not provide tools or software to remove its watermarks. While it is theoretically possible for someone to engineer a Digimarc removal system, the process is asymmetric. By this we mean that removing a Digimarc watermark without discernibly altering the image quality is a difficult process. Because watermarks cannot be removed, we suggest that the image creator/owner retain an original copy of the unwatermarked image. Digimarc's Readers can read the watermark after scale changes of as much as .6x to 2x of the original image size. A Digimarc watermark can also survive any amount of image rotation.

What is MarcSpider?

MarcSpider (http://www.digimarc.com/imaging/prspider.shtml) is an image tracking service that crawls the most highly trafficked public areas of the World Wide Web, searching for watermarked images and reporting back details about when and where the images are found. This service allows content developers like Web designers, photographers, stock photography agencies and publishers of entertainment, sports and news images to track their work as it travels the Web. MarcSpider is a great service for tracking your images in marketing programs, monitoring the use of images, and identifying potential image usage infringements. Similar to our Watermarking Subscriptions, the annual MarcSpider subscriptions are volume based, which this includes unlimited access to your MarcSpider report. MarcSpider reports are available online any time of day and are updated continuously so you'll always see the latest image finds. You can view your reports one month at a time or see a complete report history.

How do I watermark my images?

There are a number of ways you can watermark your images. The first step is to determine which tool you would like to use to embed your watermark. Digimarc works with a number of image editing and asset management tools to offer watermarking solutions. The Digimarc watermarking technology is integrated into leading image editing and asset management tools such as Adobe, Corel, Micrografx, Cerious, Jasc, Creativepro.com, Workstation.com, and Xat.com. (http://www.digimarc.com/support/csidtutor.shtml). Our Plug-ins, bundled with Adobe, Corel, Micrografx, Cerious, and Jasc provide you with the technology to embed and read the presence of Digimarc watermarks in images. Digimarc watermarking embeds a message within your images that the human eye cannot see, but that a computer can read. This message makes your images "smart" so they can announce your copyright and track your images on the Web. Watermarking an image is easy. First, open your image in any of the image editing or asset management applications we are integrated with. For best results, perform all image edits prior to watermarking your image. Once you've gotten your image in final form (color corrected, resized, etc.), save a copy of the unwatermarked image. To watermark your image, select Digimarc Embed Watermark from either the Filter or Image menu. Enter your unique Digimarc Creator ID to personalize the software, select the desired image attributes for your image (Restricted Use, Do Not Copy, and Adult Content) and set the watermark durability. Then, click "OK." Your image is now a watermarked. The Digimarc technology is also available through and by web-server based solution providers such as Creativepro.com, Workstation.com, and Xat.com. For instructions for each of these, please visit our partner's website.

How can I read a watermarked image?

The image editing and asset management applications (http://www.digimarc.com/support/csidtutor.shtml) are capable not only of watermarking images, but also of detecting and reading them. When an image is opened or scanned into one of these applications, the Digimarc Plug-ins' auto-detection software quickly scans the image for the presence of a watermark. If one is present, the application displays a copyright symbol ((c)) in the title bar of the image window, providing an instant visual cue that it is a watermarked image. Automatic detection generally takes between 1/2 and 1 second regardless of image size. The passive detection and proactive notification are key features of Digimarc watermarking. To read a watermark, open your image in any of the image editing or asset management applications we are integrated with. Select Digimarc Read Watermark from either the Filter or Image menu. The Digimarc ID is shown immediately. ReadMarc, Digimarc's stand-alone application to detect and read a watermark, is available free for Windows and Macintosh platforms. download the latest version of ReadMarc

How do I get a Watermarking Subscription and Digimarc ID?

You can register for a Digimarc ID here. We offer volume-based Watermarking and MarcSpider Subscriptions. Here is a link to the page on our website that outlines our subscription tiers http://www.digimarc.com/imaging/pridsignup.shtml and links to our online registration area. Subscription volume tiers are based on the total number of watermarked images you have in distribution, or the number of watermarked images in your collection plus an estimate of the number of images you plan to watermark over the next year.

When should I watermark an image?

As you prepare an image for its final use, you may take it through a number of different transformations. You may save the image in multiple resolutions, and you may perform a number of the edits such as color correction, cropping, rotating, scaling, etc. Watermarking should be one of the very last processes you apply to your image, except for compression. The recommended workflow sequence for watermarking, then, is as follows:

  1. Make all necessary modifications to your image until it has the desired final appearance.
  2. Embed the watermark.
  3. If the final image should be compressed, compress by converting to JPEG or GIF format.
  4. If the image is intended for printed output, perform the color separation now.
  5. Read the watermark and use the signal strength meter to verify that the image contains a watermark of sufficient strength for your purposes.
  6. Publish the image.

| resources |