Secure your content without impacting the user experience with JIT watermarking

The global Media & Entertainment (M&E) industry faces massive losses because of online piracy. Content like movies, web series, TV shows, games, and music is leaked by bad actors for illegal gains jeopardizing the legitimate businesses of filmmakers and studio executives. Piracy is not anymore limited to the lone wolf in a theatre with camcorder or the nerd behind a desktop with screen capture software. It’s a multi-billion dollar illegal enterprise.

According to the U.S. Chamber of Commerce, online piracy was responsible for at least $29.2 billion in lost domestic revenues, 230,000 in lost American jobs, and $47.5 billion in reduced GDP in 2020. Leaking content before its premiere is the sole focus of the online piracy world, dealing the biggest blow to film studios, television stations and OTT platforms waiting to benefit from the monetization, be it theatre box office, broadcast primetime or streaming premiere.

It is this high vulnerability of content, that has made the need for strengthening content security an obsession for the M&E industry. Today, most pre-release content that needs to go through content operations like reviews, screening reviews, localization etc. need to leave the organization boundary with a sufficiently high risk of content leakage. Given the large revenues at stake, it is natural that strong security controls need to be put in place, to protect such high-value content.

Most content owners follow a multi-layered protection strategy covering all stages of the content journey – content at rest (encryption), content in motion (DRM), content in use (DRM), and if all the above fail, traceability (ability to identify the point of compromise and the thief) through watermarking technologies. But watermarking is not the sole panacea for online piracy. It has its limitations, primary being it's just a deterrent.

To start with, M&E enterprises have to invest in strong deterrence mechanisms to deter people from leaking such high-value content. The studios and content owners typically rely on invisible watermarks (commonly referred to as forensic watermarks) as there exists no technology that transcodes content in real-time to apply the visual user watermark without impairing the user experience. A forensic watermark has the ability to trace back to the original source of the leak (device) due to the forensically embedded information. But they cannot deter such leaks from happening in the first place.

To work around such user experience challenges, client-side technologies that provide a visual overlay on the player and offer the feeling of a watermark are generally employed. However, client-side visual watermarking is not fool-proof as it can be removed by hacking the player code with advanced JavaScript skills. There is a library out there that burns the user details at JavaScript level via browser DOM elements.

A combination of the two is generally considered a strong deterrent for criminals engaging in piracy activities.

Hence, in-stream visual watermarking which is user specific yet does not impact the user experience is the solution. PFT’s technology innovation has led to the creation of a viable deterrent to content leak which does not mar the user experience – JIT (Just-In-Time) Watermarking – this patented technology delivers a visual overlay on the HTML element so a hacker cannot change or remove it on the client side. Like forensic watermarking, it has absolute traceability too.

PFT’s patented solution comes with two major innovations: (1) the ability to apply user-specific burnt-in watermarks, real-time and (2) make the experience of a “Just-In-Time-Watermarked” stream as seamless as that of a CLEAN stream. At its core, this invention treats every streaming-segment request from a user’s video player, as an atomic segment that will need to be transcoded by a GPU-farm and handed back in real-time to the player. All this, in conformance with the needs of the standard streaming protocols including HLS/DASH, among others. Consequently,

• Any segment requested, across any bit rate, is served with a watermark and
• Any high-speed trick-play request (up to 3X) is served elegantly, since it relies on high-speed transcodes through the GPU farm.

JIT Watermarking an HLS or DASH video segment results in it permanently bearing a visible watermark that has information to identify the user allowed to view it. A visible watermark does not prevent content copying, but it serves as a deterrent to malicious users that intend to copy it via screen recording software and handheld cameras.

When the player requests a content segment, the reverse proxy receives the request and,

• Downloads the segment from WOWZA
• Stores the segment in TMPFS
• Transcodes the segment using FFMPEG running on a GPU
• Performs specific steps for encrypting and adjusting atom structures based on protocol
• Serves the watermarked segment’s bytes back to the player

The steps in stage 4 differ based on the streaming standard used, such as HLS or DASH as they have different file formats, and atom & box structures.

In JIT Watermarking of DASH segments, a few additional steps are involved due to the difference in file formats. DASH segments are of ISOBMFF specifications while HLS segments are of Transport Stream specifications. The box structures in these formats are different too.

Secure your pre-release content against bad actors with JIT Watermarking. With Cloud costs continuing to drop and GPUs becoming even more affordable, it is only a matter of time before the industry decides to provide streaming URLs to their content partners and vendors, instead of creating copies of content, individually watermarked for their use.