According to most accounts, PayTV and Over-the-Top (OTT) video content providers are locked in a winner-take-all struggle for the eyeballs — and wallets — of consumers. While conflict makes for a good story, and there is definitely competitive tension between these two camps, the reality is that a majority of video consumers still find value in both PayTV and OTT services. A June 2015 survey, for example, found that nearly 70% of Netflix users also subscribe to a PayTV service.
The explanation for this statistic is simple: most video consumers still enjoy a mixture of live/linear and on demand content. Currently, the only way to satisfy this twin craving is to subscribe to multiple suppliers. But that means dealing with multiple interfaces and a fragmented viewing experience.
The potential to unite the immediacy of live/linear and the convenience and breadth of OTT programming in a single, unified experience is what has segments of the media industry so excited about cloud DVR (cDVR), which is quickly emerging as an enabling technology for a new and personalized video consumption experience. A second-generation of cDVR solutions is now making it possible for content owners and aggregators to combine the best elements of PayTV packages and OTT offerings into a single service.
An Abundance of Bytes
The major challenge of offering a comprehensive cDVR service is massive storage requirements. Though steady breakthroughs in compression technology mitigate this situation to some degree, cost issues related to the storage of video content will only grow more acute as higher resolutions, such as 4K/UHD and, eventually, 8K, gain prominence.
But that’s only the tip of the storage iceberg. To offer a cloud-based video delivery service that allows subscribers to consume all of their video from any device or any location, service providers must support multiple bit rates and format options for each piece of content. That means having potentially dozens of versions of the same program always at the ready. On top of that, current copyright policies require most content distributors to store a separate copy of programming for each subscriber.
The costs and complexities associated with these massive storage requirements have prevented content aggregators from delivering a fully personalized and affordable new television experience. Service providers have had to make sacrifices, including functionality, affordability or the breadth of their content catalogue, to offset storage expenses.
But those days are coming to a close. Recent breakthroughs in storage optimization are enabling service providers to overcome these obstacles and introduce ground-breaking and cost-effective video services that unify the best of PayTV and OTT services into a seamless viewing experience.
Three of the most important are summarized here:
1. Just-in-Time Packaging
Content streamed over the Internet must be compatible with a growing number of playback formats, including HLS, HSS, HDS and Dash. Storing each recorded asset in multiple formats is a huge multiplier of needed storage. Just-in-Time Packaging (JITP) allows service providers to store content segments in a single mezzanine format, such as DASH-TS or HLSv4. JITP works by re-packaging the mezzanine ABR fragments to the clients’ requested delivery format at the time the content is requested. Using JITP also allows application of session-based encryption, where and when it’s required.
Deploying JITP capabilities, of course, introduces a trade-off between the cost of storage and computational resources required for on-the-fly packaging. Download the white paper Itorage Orchestration for Unlimited Cloud-DVR Capacity for precise information about balancing computational resources and storage costs. As a rule of thumb, though, content aggregators and distributors can realize overall savings if they support even just two packaging formats.
2. Just-in-Time Transcoding
Similar to packaging, transcoding technology has reached a point where it is often more cost-effective to store only the highest bitrate profile in the ABR table and then transcode to other profiles at the time of the client request.
As with JITP, Just-In-Time Transcoding (JITT) offsets reductions in storage with additional computational cost. In the case of cDVR services, storage savings vs computational cost optimization can be achieved by applying JITT only to assets with a certain streaming probability. Recent cost analysis shows that limiting JITT resources to 20% of your streams can save about 54% of total storage and about 50% of the total infrastructure cost of storage and JITT servers. Reduction in total storage also reduces space and operating expenses related to power consumption and hard disk drive replacements.
3. Streaming Probability
Arguably, the most important component of any storage optimization strategy is the ability to accurately project the frequency with which content will be accessed. Streaming probability, calculated for each asset, provides the cDVR recording system with the information to determine the number of profiles to be stored or to be streamed on the fly. Streaming probability can also be used to determine where to store an asset. For example, if the streaming probability of a shared copy asset is very high, it is best to store it in edge servers. If it’s very low, archiving it to tape, where storage cost is low, becomes a possibility.
Mastering the art — and the science — of asset lifecycle management requires extensive knowledge of caching algorithms and other factors that indicate how likely a content asset will be viewed over a defined period of time. To complicate things even more, the rules for shared copy deployments are completely different from private copy deployments. In either case, it’s impossible for storage optimization purposes to overestimate the value of knowing, for example, that an asset that isn’t viewed within 72 hours of being recorded has a roughly 20% chance of being watched on the fourth day.