A substream (or secondary stream) is a second, lower-resolution copy of a camera's video that the camera encodes alongside its full-quality "main stream". The camera publishes both at once: the main stream (say 4K at 8 Mbps) for recording and full-screen viewing, and the substream (say D1 or 720p at a few hundred kbps) for everything that does not need full detail.
Substreams exist to solve a hard limit: decoding video is expensive. A 32-tile video wall showing 32 main 4K streams would overwhelm the client's CPU/GPU and the network, dropping frames. By feeding the grid with substreams instead, the same wall stays smooth, and the main stream is fetched only when an operator expands a single camera. Server-side analytics often run on the substream too, since detection rarely needs full resolution. Many cameras now offer a third stream so recording, viewing, and analytics can each get a tailored copy.
The pitfall is leaving clients pulling main streams everywhere — the single most common cause of a laggy, frame-dropping video wall. The fix is configuration, not hardware: bind multi-view layouts and mobile clients to substreams, and reserve the main stream for recording and the expanded single view.
