Near-live means the source is genuinely live but the viewer is held a fixed distance behind it on purpose. Sports rights deals often require a delay of 30 seconds so betting partners can settle markets. UGC platforms add a 5–60 second delay so moderation can flag and cut a stream before it goes wide. Religious or political broadcasts may sit on a 5-minute review buffer.
Technically the pipeline looks like live — a contribution feed, an origin, an HLS or DASH manifest — but a deliberate delay node sits between origin and packager. That node holds segments in a ring buffer, possibly runs ML classifiers on them, and releases them only after the configured latency. Done right, viewers see a glitch-free stream that is simply a fixed amount behind real time; done wrong, you see catch-up glitches or audio/video drift.
Near-live also covers situations where the natural delay of the chain happens to land in the 10–30 second range — older HLS with 6-second segments and a 3-segment hold-back, satellite contribution feeds that take a hop through a teleport, or international relays. The label matters because operations teams treat near-live differently from sub-2-second live: ABR ladders can be wider, error recovery has more time, and you can use larger CMAF chunks for better compression.
