(21 June 2007 4:00 pm)

SINGAPORE: With more delivery platforms being made available for content distribution, securing it is becoming important. Attempts at content protection in this digital era have been sporadic at best in places like India, China, and Australia. In order to bring the various stakeholders together, a Center For Content Protection dedicated to the Asia Pacific region will launch in November.

The Motion Picture Association (MPA) is a partner of the initiative. They are now talking to different players including governments to come on-board. The aim is to serve as a central place for information on protection of digital content. The Motion Picture Association of America (MPAA) director, Technology Initiatives for the Asia Pacific Isa Seow made this announcement at a morning’s session at BroadcastAsia which looked at how content can be protected in a digital era.

One way to achieve the first goal is to set up a Broadcast Flag for free TV. This is a solution that America is looking at to stop free TV from going online in an unauthorised manner which is what is happening with sites like Youtube. This will help preserve the value of the content both for the owner and for the advertiser.

“The Broadcast Flag implements protection upon reception. It allows for clear transmission of digital TV signals and then invokes protection at the point where it is first received. The Broadcast Flag uses a combination of technology measures or regulatory obligation,” said Seow.

Another challenge that the center faces is to plug the Analogue Hole. This refers to digital content being transferred to an analogue format and then being reconverted back to digital. Digital hardware devices have analog output capabilities which allows for compatibility with older analogue devices. This allows for the conversion and represents a huge hole in measures being taken for content protection. In fact it nullifies attempts at content protections.

This results in an inconsistent consumer experience, said Seow. It allows and creates an un-level playing field between makers of analogue and digital products. So the center together with the various stakeholders will have to come up with effective laws and implement security measures. The center will also organise seminars and training sessions on content protection. It will have mailing lists through which it will build up a repository of information. It will also provide small grants for the development of content technology solutions. One of the benefits for stakeholders who become a member of the center is that they will participate in R&D and drafting guidelines.

Seow also dwelt on the different content protection measures available for different devices. For instance for the DVD you have the 40 bit scrambling encryption system. There is also the digital Transmission Content Protection (DTCP) system. This allows for content to be transferred from one device in the home to another.

There is the Open Mobile Alliance which works on things like fingerprinting applications to ensure security of content on the mobile. Of course, for Pay TV you have Cas. Casbaa provides guidelines on how content should be protected and what upgrades are available.

Governments across Asia need to be more proactive which will help self regulation come from within the industry, said Seow. For instance hardware makers should ensure that their devices do not play pirated products.

There was also a session on digital watermarking.
Cinea president Rob Schumann noted that forensic watermarking is at the heart of a DRM solution. It puts identifiers in place so that anonymous distribution and use can be tracked. Companies like The Academy of Motion Picture Arts And Sciences in the US use this for their screeners, which are mailed to members. Forensic watermarking also lets you know when one’s DRM system is broken.

The technology of Running Marks, Cinea's patented watermarking technology, was designed specifically to place forensic data in copyrighted material from the source, through the distribution network and across a broad range of consumer playback devices; from PCs to set-top boxes to portable media player. It looks to counter the most sophisticated piracy techniques while totally preserving image quality.

A forensic video watermarking (serialisation watermarking) process marks video with the intent to later recover those marks. The objective is to deter piracy by enabling content owners to track pirated content back to its source. The three steps to adding a forensic watermark are:

Determine where to place the mark (Image Analysis). Determine what information will be in the mark Actually insert the mark

Most forensic watermark solutions treat these steps as one operation, which leads to several problems. First, processing the image to determine where to insert the mark is very computer intensive, requiring a trade off between cost and quality.

Second, such monolithic solutions typically rely on a global secret that, if exposed, renders the mark vulnerable to counter measures. Third, it may be problematic to renew the marking method after compromise because doing so would require updating all fielded devices.

Cinea's technology differs from other solutions on the market in a number of critical ways, Seow said.

An ultra-lite inserter enables serialsation watermarks to be embedded in nearly any consumer video device. As a result, Running Marks can apply device-specific marking to individual video copies during playback, enabling piracy tracking back to the source.

A fully renewable system provides for security algorithms to be renewed on the back-end, with no need for field upgrades, affording cost savings when millions of devices have already been deployed.

The blind watermark recovery system of Running Marks requires no information about the original source content, channel of content distribution, or player. This simplifies the recovery operation for system operators and content owners.

It allows for compressed domain watermarking that affords an additional level of security by allowing the technology to reside within the secure envelope of the digital rights management or conditional access (DRM/CA) provider.

Running Marks has multiple insertion capabilities that allow watermarks to be embedded across the entire distribution and playback chain, helping content owners to identify sources of leakage or points of vulnerability

How does the system work? The original content is analysed to locate suitable video regions as place holders for message insertion. Locations are chosen so that messages will be undetectable to the human eye. For each location, two alternate content representations, "mark images," are created and stored for later use.

Image Analysis can be done in any number of locations.
In a video distribution system the image analysis can be done at the head-end; done once and stored for VOD content and done in real-time for broadcast content. In a fixed media or electronic delivery application, the image analysis can be done in a post production environment thus preserving creative control and optimizing the marks to the content.

Although mark placement image analysis is the most CPU-intensive function, it only needs to be done once in order to support multiple message creations and insertions as content moves through its normal distribution cycle.

Message Creation: User messages can be 64-bits long containing, for example, playback device id, playback date and time, and more according to the specific application requirements. This message data is then encoded to create a robust and redundant message stream.

Insertion: The ultra-light inserter can be integrated at any stage of a multi-node content delivery network, for example: at the head-end, at an interim point, or in the final consumer playback device such as a set-top box, portable media player, PC, or mobile phone. The run-time encoded user message is generated and used to determine which data is placed in each message location.

The insertion process requires only a simple data copy operation and can be implemented with virtually no time or compute overhead. This is critical because it enables insertion in consumer devices with limited computer resources. The marks are inserted into the compressed stream, while the content is still within the security envelope of the DRM/CA, before the content is decoded for playback.