In 2018, the CNBC All-American Economic Survey found that nearly 60% of Americans subscribe to at least one type of streaming video service.* That's up from around one-half of Americans just a few years ago. To meet rising demand, broadcasters are increasingly taking steps to make their sizeable content libraries available online.
However, when a library includes tens of thousands of hours of content, encoding it can be an enormous challenge. Traditional hardware solutions don't scale well, and they also lack the flexibility to adapt to inevitable changes in encoding standards. As consumers continue to demand faster, clearer and more reliable online video, the formulas required to provide it will necessarily change, too.
A cloud-based software solution offers clear advantages – but what should broadcasters look for? Here's our guide to encoding a large content library through the cloud.
To prepare video for an SVOD service, it doesn't just need to be encoded in multiple formats required by different platforms and players. It also must be outfitted with metadata that's searchable by recommendation engines, whose sensitive machine learning algorithms often carry extremely specific guidelines. The video also needs to be flexible, able to stream in both standard and high-definition formats so customers with varying internet speeds can still enjoy it without buffering. Furthermore, each video needs to include digital rights management (DRM) information that specifies who can access it, in which regions, for what length of time, and other customer-specific criteria.
A good cloud encoding platform should be able to complete these tasks with one click – and keep your content secure at the same time. For example, at Verizon Digital Media Services, we've designed a system that effectively automates encoding in the cloud. Our Slicer software encodes a file once at the highest desired bitrate. Then, it "slices" that file into encrypted segments that are then uploaded to the cloud, where they are encoded at all other desired bitrates. We call this model: encode once, run everywhere. It makes encoding video content incredibly easy.
There are many moving parts to encode a broadcast video library effectively, and that's before technological advancements move the goalposts yet again. That's why our video streaming service streamlines large-batch encoding projects into one simple digital workflow.
What makes encoding a large content library with us extremely easy is a tool called Slicebot. Each Slicebot is effectively an instance of Slicer that watches a folder. When a broadcaster adds video assets to that folder, the Slicebot gets to work. It encodes the video using the Slicer technology while pulling all the metadata it can from the file.
Since the Slicebot tool lives and works in the cloud, it also scales massively. Broadcasters with particularly large libraries can have as many Slicebots working in parallel as their CPU performance will allow. Users can also run a script that tells them when the encoding process is falling behind, cueing the system to spin out another instance of Slicebot automatically. It also includes configurable notification hooks that alert you (or your content management system) when assets are ready for playback.
Of course, using more compute increases costs as well as speed, there is always a tradeoff. But compared to other options, a cloud-based, automated solution is nearly always going to deliver greater value.
To get the same effect with a hardware solution, a broadcaster would need to purchase and configure multiple video encoders to run in parallel. After thousands of hours of video are formatted and uploaded, most of that expensive equipment will sit around unused. If the broadcaster needs to suddenly offer content in a new format, perhaps in 4K quality, they will need to buy new equipment to keep up. By contrast, with a cloud-based tool, keeping up with change is as simple as upgrading your software.
Encoding at scale is never going to be a simple task because of changes in standards and the massive need for compute power, but with automatic, cloud-based solutions, it can become a much easier one.
*Liesman, Steve, "Nearly 60% of Americans are streaming and most with Netflix: CNBC survey." CNBC, cnbc.com/2018/03/29/nearly-60-percent-of-americans-are-streaming-and-most-with-netflix-cnbc-survey.html