In my last post I looked at ‘Forensic Watermarking’ which protects video content along with multi-DRM.
The ‘forensic watermarking’ can be classified as follows according to the subject of insertion, application field, and implementation method.
For the content of this article, I referred to ‘Forensic Watermarking Implementation Considerations for Streaming Media’ proposed by Streaming Video Alliance. The Streaming Video Alliance is a group of companies from the video streaming industry researching related technologies and standards.
Forensic watermarks can be categorized as follows depending on who is inserting a watermark in the content:
- Operator Mark (Distributor Mark): Inserted by content owner such as Hollywood studio. This type of solutions insert and track information about content distribution channels or content service providers.
- Session Mark (Subscriber Mark): Inserted by content service provider. This type of solutions insert and track end-user information of content services. (User ID, device ID, IP, time stamp, etc.)
Forensic watermarking for video content is being applied to a variety of areas, including:
- Pre-release: Pre-release content is distributed in the form of files or discs to internal or external stakeholders of the studio before release. Studios apply a static watermark similar to ‘Operator Mark’.
- Digital Cinema: Movie theaters apply watermark to digital cinema contents to track illegal recording. Theater and screen time information are inserted as a watermark.
- OTT VOD service: Mainly applied to premium contents of online movie service(e.g. Netflix, Hulu, etc). ‘Subscriber Mark’ is applied to track the user in case of illegal leakage.
- OTT Live Service: ‘Subscriber Mark’ can be applied to live/linear contents such as sport event in real time. Due to the characteristics of live content, it is necessary to reduce the playback delay and to detect watermark quickly.
Types of Subscriber Mark
The ‘Session Mark (Subscriber Mark)’ solution, which inserts unique information for each user in real time, is categorized as follows according to the implementation method.
- Client side: Inserts watermark from set-top box or OTT client. Implemented in the form of device firmware or client SDK.
- A/B variant (server side): Prepares two watermarked versions of a video by pre-processing on the server. Combines the two versions at playback time to apply watermark information to the stream. (Mainly on CDN edges)
- Bitstream modification: A combination of server and client methods. The server pre-processes the video to identify modification areas. The the watermark is inserted in the CDN edge or client using the information transferred in the metadata.
Among the above watermarking implementations, client-side and server-side watermarking have the following pros and cons.
- Pros: No need to change the server infrastructure of the service backend. It is suitable for live service because playback delay and detection time is short.
- Cons: Integration with client device or player is necessary. Because the watermarking logic can be exposed to clients, a hardware-based security (TEE) environment or additional application security is required.
- Pros: No player integration is required, so any player or device can be used. Watermarking logic is not exposed to the client. Ideal for multi-platform OTT VOD services that need to support a variety of clients.
- Cons: Needs changes in server infrastructure. Storage and transport costs increased due to the two variants of content. It may cause longer playback delay and requires longer video for detection than client type solutions.
Due to the advantages and disadvantages of each method, server side watermarking is mainly suitable for OTT VOD service. And client side watermarking is widely used for live services such as live streaming or IPTV.
As we have seen, forensic watermarking is used for many purposes and in many areas of the media and entertainment industry.
If you’re interested in protecting the copyright of multimedia content and revenue of media services, you need to understand the differences between various watermarking solutions to find one suitable for you.
To quickly and easily apply forensic watermarking and multi-DRM, which are essential for premium content security, it is recommended that you use a professional company that can provide both solutions as a unified service.
‘PallyCon Forensic Watermarking’ is a cloud-based forensic watermarking SaaS service provided by INKA Entworks, a content and application security technology company. It is integrated with ‘PallyCon Multi DRM’ service in a unified content workflow.
- #1 Forensic Watermarking Implementation Considerations for Streaming Media: https://www.streamingvideoalliance.org/download/27623/