By Linda Noe
By inserting watermarking IDs or codes into frames of premium video assets, forensic watermarking is gaining popularity as a technique to prevent content leakage and monitor leaked streaming video. The interest in this technology has been fueled by the massive expansion of OTT platforms, particularly during the Covid-19 pandemic, even when it has been around ever since digital signals were developed.
A forensic watermark, often called a digital watermark, is a code or group of characters inserted in a digital document, video, audio, image, or programme that allows the content producer and authorized user to be uniquely identified. In a dynamic scenario where a connected user makes multiple queries for digital signals, like in the case of streaming videos, video watermarks are generated instantaneously so that information about the user can also be included in the watermarking data.
The two most popular types of forensic video watermarking approaches are manifest-based or A/B watermarking and bitstream-based watermarking. The pixels in the transcoder are modified in the A/B watermarking approach, and the content is pre-processed to detect the changeable pixels and modify them without impacting the quality of the host video. Depending on the content owner’s preferences, pre-processing can be done using a CLI preprocessor, preprocessing libraries, or a SaaS packaging service, all of which can be provided by a multi-DRM SaaS vendor. After that, the pre-processed data is passed into the transcoder, which modifies pixels.
The assets are divided into blocks, resulting in a distinct pattern of As and Bs in the playouts. In the original video frames, different values (A/B or 0/1) are added. Once this is done, the output contains two sets (A/B) of the encoded video. When the client makes a query for playback, two content copies are combined in such a way so as to provide a unique manifest for each subscription session carrying the information about the user and the session. Finally, the packager creates a unique manifest by combining the video portions from each of the two copies of the DRM protected content. The watermark thus obtained on this one-of-a-kind manifest can then be used to track down the original user if unlawful activity is detected.
In a world full of hackers who want to break the security of premium content, the A/B watermarking method has the reputation of being secure and difficult to crack, making it an effective solution against for DRM protected content.
