Content redistribution has become a major threat for all content creating platforms. Pirates can obtain content from a VoD service, an internet live streaming platform, or a set-top box and redistribute it on the internet to a huge number of consumers around the world. This is an even bigger threat in terms of revenue for operators with exclusive content rights such as for live sports events. Subscriber-level watermarking solutions combined with large scale content monitoring techniques are effective anti-piracy measures that can be used by operators to protect their content and revenue streams from theft going beyond just DRM protected content.
Session-based subscriber watermarking uses unique watermarking information applied per playback session to track the end-user of the content (device ID, user ID. device type, etc.). Subscriber mark product types can broadly be divided into the following:
- Bitstream modification (server + client solutions): In this technique, selected areas of a picture are modified while maintaining the quality of the video. The potential blocks that are to be manipulated are first identified on the server-side and this information along with the content is sent in a metadata file. CDN edge processing or the client-side then executes the changes to give the unique watermark. Although it is robust, this method suffers from computing overhead which makes it unsuitable for live content.
- A/B Variant Video Watermarking (server side solution): This is a two-step process primarily aimed at the OTT and VoD segments. Two version of the content (A and B) are created which are completely similar except for the watermark. The asset is then segmented into chunks and the playouts have unique patterns of As and Bs. The chunks are interleaved (either client-side or by CDN edge processing) to provide the unique watermarked content file.
- Client-side Watermarking: Here, a graphical overlay is composited onto the video stream in the client device. This can be visible or invisible depending on the requirements of the content owner. The watermark is inserted in the set-top box or the OTT client application and implemented as a device firmware or client SDK. This type of watermarking is suitable for live streaming due to less detection time and playback delays.