Piracy is an ever-evolving problem for content creators and owners. Whether because of the cost of programming or the geographic limitations on availability, more viewers than ever are tuning in to illegally-delivered content. The Alliance for Creativity and Entertainment (ACE), made up of 30 leading content creators and media companies, estimates that there were 5.4 billion downloads of pirated films and television shows in 2016.

When content is distributed or downloaded illegally, its quality is usually degraded. As higher resolutions become the norm, 4K can be degraded to HD. This means it remains at a high enough viewing quality, creating even more of a market for illegally-shared content.

This all adds up to much more than just a headache for those in the industry. It costs content owners and distributors a lot of money, so it is a problem they are eager to address – and quickly. After all, for a media company, content is their most important asset. It’s why people use their service.

So, if content is king, how do content owners stop someone stealing the crown jewels?

Creating a unique stream

Watermarking assets isn’t a new approach to protecting content. But with the number of pirates surfing the internet, the amount of content being distributed using IP and the multiple versions of each programme, it’s hard to implement in such a way that it’s easy to track the pirates.

And watermarking content through online delivery is traditionally very difficult to do. The very point of a CDN – which is used to scale TV delivery – is to make delivery more efficient by copying content. To protect content through delivery, broadcasters and filmmakers need to be able to forensically track the source of illegally shared content.

One method that has gained momentum is manifest-based watermarking. Online TV content isn’t streamed as one long video file, but as a collection of three-to-five second segments. By creating two versions of each segment and marking them A and B, content distributors can create a unique string of sections. While this does create a unique watermark that survives CDN delivery, content must be stored twice on each server, increasing operating costs. This watermark also takes a long time to identify, and it is possible to obscure it by splicing together the two versions.

A new kind of watermarking

The goal is to put the spotlight back on content thieves and catch them in the act. By watermarking from the edge of a dedicated TV CDN, a bitstream-based watermarking solution inserts a unique watermark directly into the bitstream, on the fly.

The best approach to make sure content survives the CDN is to manipulate the bitstream for each individual viewer. The process only requires a few byte changes, performed by specialised servers on the edge of the network as they stream content. Just before delivery to the end user, the system identifies just a few pixels within the picture’s frame and changes them in a way that’s completely unnoticeable to the viewer. For example, black might be changed to very dark grey. This pixel identification creates a unique code for each streamed version of the video and is easily detectable by the client. Within five seconds, a pirated stream can be identified and tracked to the specific user.

This bitstream-based watermarking, compared to manifest-based watermarking, can be embedded much faster, at a rate of several bits per second. It’s also much harder to cheat this watermark, so pirates can be identified much faster.

We’ll quickly see an increase in premium online content being protected using this kind of watermarking application when it’s delivered online via a CDN. Protecting content in this way gives broadcasters and content owners the forensic tools they need to track down anyone who illegally streams or copies their content.