CONNECTING THE CLOUD FOR LIVE EVENT BROADCASTING by Richard Heitmann
The media world is changing. Consumers today expect to access more content than ever before, from wherever they are, and on the devices most convenient to them. This trend has challenged traditional media companies to produce content in faster timeframes and in a wider range of formats to meet demand, putting a considerable strain on their fixed-capacity data centers, IP networks and satellite distribution systems. To alleviate these problems while improving the quality and availability of their programming, broadcasters are increasingly shifting their workflows to the cloud to keep pace with the growing demand.
Delivering high-quality programming to support the huge number of diverse devices in the market is not a simple task. Varying performance levels and screen resolutions of consumer devices necessitate an extensive amount of video processing to optimize the original content for each device platform.
Live and near-live streaming of broadcast quality video content have also required expensive and specially provisioned infrastructure. Traditional distribution systems use expensive live satellite feeds from the streaming source or dedicated terrestrial networks with high quality of service to ensure low latency and low packet loss rates so as to not degrade play-out quality. This cost is burdensome in any media application or service, but it becomes especially impractical in live and second screen experiences for events that occur one time, such as sporting events, movie premieres, concerts, operas and other one time events that can’t justify the investment in the dedicated infrastructure for direct distribution, or amortize CDN costs over long periods of viewing.
The cloud is the ideal candidate for meeting the variable and growing performance requirements for fast and secure transfer, processing and delivery of content. As demand rises, cloud infrastructure enables real-time provisioning of as many virtual machine instances as necessary to handle the increased transfer demands of video from the venue and the associated higher transcoding workloads required to transform the content and prepare it for delivery. When demand decreases, compute capacity can be scaled back to avoid over-provisioned scenarios, ensuring the most cost-effective use of the underlying infrastructure. By only paying for the infrastructure they truly need, production teams are empowered to process their workloads faster than ever before while eliminating large upfront investments in expensive IT equipment that would typically remain idle for a large portion of the day.
A great example is a recent global sporting tournament in Brazil. During this time, 34 cameras covered each stadium and produced around 60 hours of in-game footage from the 90 minutes of game time plus an additional 30 hours to cover live pre-match and post-match footage. This resulted in a massive amount of data that needed to be moved over the WAN and into the cloud using Aspera On Demand.
To satisfy consumer demand for content anytime, anywhere, and from any device, the organizers partnered with broadcasters to ensure all video footage was available and optimized for second screen devices such as laptops, phones, and tablets. Video had to be processed for streaming and uploaded in near real-time to the cloud where it could be transcoded for each target device platform as soon as game coverage began. In addition, six live streams of produced content had to be delivered concurrently alongside the direct camera streams. This totalled 40 streams of high
BROADCAST BEAT MAGAZINE
IBC Issue September 2015
43