Synopsis
With video streaming on the rise, video player solutions face a growing need for an efficient and scalable approach which is adaptable to variable network conditions. Viewers are becoming increasingly demanding when it comes to streaming online video, as latencies and zapping times are expected to be as low as in classic (or even analogue) technologies.
In this talk, the current low latency streaming implementations are reviewed, including LL-DASH, LL-HLS and HESP. A deeper understanding of how they perform for different criteria is explained, such as latency, bandwidth, zapping time, scalability, etc., as well as identifying the strengths, limitations and best practices.
Furthermore, the discussion explores how these approaches can be further improved. Proposed improvements include: the concept of two complementary streams with different encoding parameters, the initialization stream and the continuation stream. The discussion navigates how the combination of these two streams allows for instantaneous viewing and how this leads to sub second latencies and zapping times as well as lower bandwidth consumption without compromising on scalability and network adaptability. The results of this new approach will be presented based on conducted field trials.
