SMPTE 2019: How Cloud Technology is Changing the Live Transcoding Landscape
Appliance-based encoding is still prevalent today for HD AVC and UHD HEVC. But it begs the question: what will happen when the industry moves to more complex codecs such as AV1 and VVC, and to higher resolutions such as 8K? New codecs such as AV1 and MPEG VVC will enable broadcasters and service providers to deliver better experiences, including 4K, 8K streaming, and 8K VR, but brings with it increased encoding complexities. Since adaptive streaming requires multi-resolution synchronized encoding, the delivery of next-generation video becomes even more challenging. In fact, for live applications, the amount of complexity will be unmanageable with current CPU or GPU on-premises technologies.
This presentation will look at various codecs (i.e., AVC, HEVC, AV1 and VVC) at different resolutions (from SD to UHD-2) and compare their encoding complexities. It will also examine the limitations of on-premises server based live encoding and explain how cloud computing can solve these issues. The presentation will explore at length the various benefits of cloud based encoding, including details about how the cloud enables operators to gain business agility, in terms of being able to quickly turn on and turn off services and efficiently scale up to support limited event or sports channels that might only last for a few days. With cloud-based encoding services, video content and service providers can launch new services without requiring a pre-existing engineering team or datacenter, both of which can be expensive. Using the cloud is also advantageous for channels that would not necessarily have a clear ROI if fixed infrastructure were to be used.
In addition, this presentation will also highlight the concept of elastic encoding, where resources are adjusted in real time based on the complexity of the content. This technique applies quite well in a cloud environment and is already showing promising results.
Speaker:
Thierry Fautier (Harmonic)