One of the most significant trends in the evolving television experience is multiplatform delivery — the distribution of live and on-demand content to connected devices such as PCs, tablets, smartphones, games consoles and connected TVs, extending the reach of traditional television viewing.
As the many screens in our lives become potential outlets for video content, broadcasters and other content owners are faced with significant challenges. How do you deliver content to such a diverse group of devices in a secure, cost-effective and operationally efficient manner while also providing the quality of experience that viewers will expect and your brand reputation requires?
A single source asset may need to be transformed into multiple versions to match each destination device type. Each platform may have different metadata specifications, UI/UX capabilities, monetization capabilities, content protection schemes, and different rules on how content or applications are published. This article will explore each of these areas and offer practical advice on how to embrace the opportunity of multiplatform delivery and avoid the pitfalls.
Content preparation
In the linear broadcast world, content preparation is relatively straightforward: The asset is sourced, ingested, stored and ultimately played out. (I am simplifying the issues associated with file-based interoperability and workflow here.) The video may be upscaled, downscaled and otherwise transformed, but the destination is relatively fixed and consistent. One input, one output.
Multiplatform distribution changes this model radically. A single source asset may be delivered to many different devices, each having its own specific characteristics in terms of screen size, aspect ratio, video codecs and container formats, hardware processing capabilities, and so on. There is a one-to-many relationship between the source and destination asset, with the number of outputs increasing as consumption devices proliferate.
For broadcasters and other premium content owners, the desired threshold for video quality on each device is high. The most common viewer complaints regarding online video relate to quality of video and playback performance, and as screen sizes increase (such as on connected TVs) these issues get amplified.
Every form factor (in fact, every device) will have its own ideal video encoding profile that balances the minimum bit rate with maximum output quality, but it is probably unrealistic to prepare video assets down to individual device renditions. Instead, a judgement call needs to be made as part of the content preparation strategy for each content provider. Choose which form factors and individual devices really matter and optimize for those carefully.
One area that is beyond the control of the broadcaster in an OTT distribution model is the viewers' available bandwidth. One thing is certain, though: This will vary significantly depending upon how they connect to the Internet, so modern online video systems must adapt to the changing performance of viewer connectivity. A number of technical systems have become available to address this problem, known collectively as adaptive bit rate (ABR) systems. These include products/specifications from Apple, Adobe, Microsoft and others. Any broadcaster intending to deliver high-quality online video should plan to use ABR as a core part of its offering.
The content delivery activities described in this section require an array of encoding and transcoding processes, and the associated workflow activities to manage these tasks efficiently and cost effectively. When building or procuring such services, it is also important to ensure that assets can be turned around quickly as viewers have increasing expectations about the availability of their favorite shows online relative to transmission on linear TV.
Content protection
As consumer broadband performance has improved over the years, the content industry has become increasingly concerned about illegal video downloading and distribution. The early encryption and DRM implementations have come under scrutiny, and a new generation of content protection systems are now being deployed. These typically provide asset-level encryption and require license servers to be deployed to service client requests.
At a minimum, this requires additional processing of each video asset to perform encryption and packaging as well as the deployment of highly secure and scalable license servers.
Leading DRM systems for online video include Adobe's Flash Access, Microsoft's PlayReady and Google's Widevine products. The choice of DRM is not entirely at the content provider's discretion; support for specific DRM systems varies across viewer devices. As broadcasters extend their reach across multiple platforms, they will likely need to support multiple DRM systems too.
DRM is not the only important component of a content protection policy. Watermarking is increasingly used to allow content to be traced, identified and even recognized by secondary devices such as tablets “listening” to a broadcast on the TV using the built-in microphone (using audio watermarking techniques).
Think carefully about how content can be appropriately protected while allowing maximum viewer reach. Content protection should support your commercial objectives and monetization strategy. In addition to securely distributing assets to devices, entitlement policies need to allow flexibility in how viewers consume assets across their smartphones, tablets, PCs and connected TVs. A common model that is evolving allows a single account to share content with a group of named users across a set of specific devices tied to a rights locker. UltraViolet is one such example, but there are many more appearing in the marketplace.
Content publication
Content publication is quite a broad term in relation to online video, but in this context it refers to the process of describing, storing, organizing, categorizing and publishing a content catalog to a viewer device. Publication is a combination of processes, metadata and platforms that ultimately present the viewer with available programs to view.
The publication workflow typically begins with metadata that describes an asset technically and editorially. This metadata may originate from a number of sources but must be normalized into a consistent structure (using a taxonomy initially defined by the broadcaster/content owner) and stored in some sort of dataset. The metadata must identify where the asset is; its technical characteristics; any restrictions on its use, such as date-bound availability windows; how it relates to other content; and descriptive information that viewers will see presented on their devices, such as show details, cast/crew biographies etc.
Depending upon the needs of the broadcaster, this metadata might need to be enriched with more information and associated with related content (such as images, video trailers, links) before being published. Depending upon the destination device and the platforms that serve it, a final step may be required before the content is published. This involves transforming the metadata to match the specifications required by the final publication platform.
When a large number of destination devices are supported, it is not unusual to have multiple transformations performed on the source metadata to enable content publication on each end device. This is an area that is both complex and lacks standardization and interoperability today.
Where possible, it is important to decouple the publication platform from the viewer application serving content on the device. This allows for the greatest level of creative freedom over the viewer experience, but for some platforms this is not always possible and the experience is therefore either predetermined and uniform, or restricted in how far it can be altered.
When building or procuring your own publication and content management platform, think carefully about device diversity, support for search and recommendation, monetization capabilities, and data analytics. As the future is far from clear, flexibility is as important as functionality.
Publication can also refer to the distribution of apps as well as video assets. Again, this is an area of significant complexity today, and the process of app submission, approval and listing within a specific app store is nonstandardized and highly variable.
Content distribution
Delivering content to the connected device is a critical part of multiplatform delivery and one that has seen technological developments in recent years. Early experiments with peer-to-peer technologies have been replaced with unicast IP delivery using the services of a CDN. Global providers of these services, such as Akamai, have built international delivery networks that combine high-speed connectivity, content caching and other advanced features to allow content providers to scale using a volume-driven pricing model. The rollout and take up of CDNs is one of the key enabling services of today's online video environment.
Looking ahead, as more content is delivered live rather than on-demand, we can expect to see increasing use of IP multicast technologies providing a similar distribution model to broadcast. Such technologies are widely used in private networks today and will increasingly play a role in public networks in the future.
Ensure that your distribution plans keep track of advances in this area. As in all forms of video delivery, distribution can be a costly part of the end-to-end service.
Viewer experience
Last, but certainly not least, is the viewer experience of multiplatform delivery. Broadcasters and content owners create a different relationship with their audiences when they are able to more directly interact with them on connected devices. The audience evolves from viewer to consumer, a subtle but important distinction where the relationship becomes more measurable, more engaging and tied to the individual rather than a viewing demographic.
The ways in which the broadcaster engages with the consumer are dependent, to some extent, on the device being used to consume. A mouse-driven PC offers a different experience compared to a touch-driven tablet, a remote control-driven connected TV or a controller-driven game console. Each device has UI differences but also expected operational behavior, and the objective should be to consistently represent the broadcast or content brand across all devices but in a form that is sympathetic to, and optimized for, the type of device being used.
It is not enough to simply get content onto these new platforms; the content and/or the experience must be adapted for the device. This requires a careful blend of technology and creativity. The boffins must work with the luvvies to create a compelling outcome for both content owner and viewer.
We can expect to see a much more integrated multiplatform television experience over the next few years — one where broadcast television and online video feel entirely complementary and not derivative or competitive. (Within the context of a single broadcaster or content domain, of course, intraprovider competition will only increase going forward.) This relies heavily on the viewer experience.
Summary
Multiplatform distribution is an important part of the television experience today and in the future. It can be a complex set of activities due to the sheer variety of end platforms and devices available. The dynamic nature of the device market increases these challenges as new devices/platforms appear, evolve rapidly and sometimes disappear unexpectedly.
Broadcasters and content providers need to track these changes closely, invest in new skills such as application development and software more generally and build processes, architectures and infrastructures to manage the technology, people and partnerships necessary to deliver a more compelling experience to an ever fickle audience.
Steve Plunkett is director technology and innovation at Red Bee Media.
