Title: Scalable, live transcoding with support for adaptive streaming and failover
Assignee: Google
Publication Number: US20130272374
Search for Prior art using: Google Prior Art Finder
It seems like there is prior art for the basic idea, which more or less amounts to:
Find a keyframe and/or scene change within a given timestamp range, and split at this point.
Encode all N chunks in parallel with slightly overlapping bits, and at different qualities.
Recombine these chunks (either concat, or via HLS or DASH).
Claim 1. A system, comprising:
A memory that stores computer executable components;
A processor that executes the following computer executable components stored in the memory:
A segment identification component that identifies segment boundaries within an interval of input streams for a plurality of transcoding pipelines;
A transcoding component that transcodes the input streams of the plurality of transcoding pipelines; and
A segmentation component that segments the input streams into equivalent chunks based on the segment boundaries identified, wherein the input streams include different transcode qualities.
Anyone even slightly familiar with digital video could think of this, and I have indeed implemented something like this myself (perhaps I can submit prior art?).
There are also things like:
http://omion.dyndns.org/x264farm/x264farm.html
which long predate this.
Looking for specific Prior Art references below predating 4/11/12 which can be provided to the USPTO through Ask Patents.
prior-art-request
