Forensic Watermarking 101
| fəˈrensik ˈwɒtərˈmɑːrkiŋ |
Forensic watermarking is a process by which a unique, invisible serial number can be added to video or audio content. The watermark is designed to remain with the content, regardless of how it might be transcoded, resized, downscaled or otherwise altered for distribution.
Supporting the Entire Premium Content Lifecycle
Digital CinemaAs described above, NexGuard is the leading supplier of audio and video forensic watermarking technology to the Digital Cinema market. Our software is deployed on over 100,000 movie screens worldwide.
OTT StreamingWhen content is streamed to a customer-owned device such as a tablet, PC or mobile phone, NexGuard Streaming adds a unique session-based identifier to the stream, making it possible to trace content that has been captured or re-streamed back to the original session.
Pre-ReleaseThis phase encompasses work done to prepare a title for theatrical, broadcast, PayTV and OTT delivery. It includes files created for editing, localization, SFX, screeners and mezzanine files.
Network ID provides Live SDI & Satellite watermarking capabilities for broadcasters & playout centres, to know the distribution path used to pirate your channel.
For premium live and VOD content delivered to Set-Tops, SmartTVs and other video players, NexGuard technology enables each device to embed a unique watermark for any video shown. This is the only proven means of tracing illicit re-distribution back to a specific account.
Content Protection Beyond DRM
Traditional access control works up to the point where the customer starts watching the content. From that point, all bets are off.
As a means of ensuring that programming reaches the correct user at the correct time and on the correct device, CAS and DRM systems are an ideal solution, offering scalability and security benefits that are well-known. However, once playback begins, these technologies offer little-to-no protection against the capture and redistribution of the content.
Forensic Watermarking for video and audio content allows the creator or service operator to embed a unique serial number in the content as it is playing. This payload remains completely imperceptible to the consumer while at the same time being resistant to nearly any transformation of the video in size, format or resolution.
Key Component for Delivery of High-Value Content
As Content Distribution to the Home Evolves to Include UHD Content, Hollywood Studios Are Taking a Stronger Stand on Security
End-to-End System Best Practices Forensic Watermarking
- The system shall have the ability to securely forensically mark video at the server and/or client to recover information necessary to address breaches.
- The watermarking shall be robust against corruption of the forensic information.
- The watermark shall be inserted on the server or on the client such that the valid insertion is guaranteed even if the device and its secrets are compromised.
A NexGuard Forensic Watermark Is Imperceptible and Robust
Unlike the type of watermark you find on currency or a fancy piece of paper, forensic watermarks are not intended to be seen by the human eye. Indeed, if you’ve gone to the movies in the past 10 years, you have not seen our watermark, but it’s been there, just in case someone decides to make their own copy of the film you are watching.
A forensic watermark is only valuable if it can be found in suspected pirate samples. This means that we need to be able to extract the watermark payload from video that might have been captured by camcorder and then cropped, rotated, inverted, transcoded, downscaled, upscaled, or subjected to any number of other transformations. This is what is referred to as robustness.
What is Actually In the Forensic Watermark?
It’s probably simplest to think of the watermark as a serial number in the form of an integer. Depending on the technology used, the number of unique payloads available for a single instance of NexGuard ranges into the hundreds of millions.
NexGuard provides a service for extracting the watermark from suspected pirate content. This detection process is quick and extremely accurate. For more information on detection, please contact us.
How Are Watermarks Embedded in Pre-Release Content?
Featured Use Cases
What About Session-Based Watermarking?
Remember how we said a few paragraphs ago that DRM didn’t provide optimal protection once the consumer starts watching the content? This is where the forensic watermark shines. By adding a unique identifier to each stream being watched at home or on a PC or mobile device, the content becomes traceable without impacting the viewing experience.
In SmartTVs and Set-Top Boxes, the watermarking technology is built directly into the device and the watermark is applied as the video is played. Regardless of whether the content is captured from the screen, or through a digital or analog output, the watermark is present in the video.
For live and file-based streaming solutions, NexGuard Streaming is able to embed a unique watermark in every stream. This product works with all popular DRM encryption methods as well as adaptive bitrate solutions such as HLS, MPEG-DASH and SmoothStreaming.
”By integrating Civolution’s forensic watermarking solution, we comply with the latest Hollywood requirements for content protection so that we can bring the best possible cinematic experience into the home. The entire integration process of the Civolution NexGuard technology was completed in only a few weeks and fit seamlessly with our HEVC/DASH platform.”Jordi Miró
Featured Use Case: Premium VOD
Device-based or network-side forensic watermarking for every premium video transaction, on any platform.
Frequently Asked Questions
What's the difference between fingerprinting and watermarking?
Fingerprinting is a method of content recognition which works by calculating a unique signature for video or audio content so that you can recognize what the content is. Forensic watermarking makes imperceptible changes to the file itself and is used to identify that specific version of the file.
Think of it this way: Fingerprinting can help you recognize a video file as being “Citizen Kane”. Watermarking would tell you that it’s the copy of “Citizen Kane” that was sent to the producer on Wednesday.
Can't a pirate just cut out the part that's watermarked?
Not really. The watermark is hidden throughout the content. Removing enough content to make the watermark irretrievable would also make the video unwatchable.
What if the video is copied by using a video camera pointed at the screen?
We call this “camcording”. From the beginning, we have offered watermarking that is robust against that. In fact, that’s the primary way that movies are pirated in the cinema, so that requirement is part of the Digital Cinema Initiative’s specifications.
What if the video is cropped?
Again, cropping is a common modification of pirate video. Our algorithms are robust even when a significant portion of the image has been cropped out.
What about transcoding? Scaling? Flopping the image?
Yes, yes, and yes. There are a whole slew of typical attacks that don’t impact our ability to extract the watermark.
Can you watermark still images? What about documents?
From the beginning, we have been focused on watermarking premium video and audio. There are other companies that offer solutions for stills and documents.
Can I add more than one watermark?
Our Pre-Release products are designed to support multiple watermarks (up to three). In addition, Pre-Release watermarks will not conflict with watermarks added with our PayTV or Streaming products.