It should be appreciated that even though FIG. 4A and FIG. 4B illustrates only the 2-dimensional data space (i.e. bit rate and quality data space) for clarity purposes, the described techniques can be applied to any n-dimensional space to satisfy various constraints superimposed by any combination of bit rate, quality, encoding runtime, network bandwidth, playback devices limitations, end-user distribution and geographic constraints, etc.
FIG. 3B illustrates a flowchart with additional details of analyzing the results of the trial encodings (308, FIG. 3A) as part of generating an event-adaptive encoding ladder for live content. In particular, FIG. 3B illustrates that analyzing the results of the encoded segments and generating an event-adapted encoding ladder (308) may include a cross-sample joint optimization process.
In a class of implementations, analyzing the results of the encoded segments (308) may include pruning results (320) from encoding a plurality of segments with a plurality of encoding profiles (e.g., removing results with excessive runtimes, removing results having a quality level below a threshold cutoff, etc.).
