What is Demuxing?
Demuxing separates a media container or transport stream into component video, audio, subtitle, and metadata streams. It changes stream organization without necessarily decoding the compressed content.
How Demuxing works
A multiplexed container interleaves packets belonging to several logical tracks and records how those packets should be timed and interpreted. The demuxer reads container structures, selects tracks, and delivers compressed access units plus side information to the appropriate decoders. This is distinct from decoding, which turns those units into audio samples, video frames, or subtitle events. Demuxing sits between byte acquisition and codec processing in playback, probing, extraction, and remuxing workflows.
Key facts
- 1Packet timestamps use a container or track time base and may distinguish decode order from presentation order; discarding that information can desynchronize tracks or reorder video incorrectly.
- 2Extracting a compressed track can be bit-for-bit lossless, but its standalone form may need codec initialization data that was previously stored in container headers rather than in every packet.
- 3Damaged indexes, truncated boxes, or invalid packet sizes can prevent seeking or parsing even when much of the encoded essence is intact, so demux failure does not necessarily mean codec failure.
When Demuxing matters
Demux media when extracting tracks, inspecting stream properties, or supplying separate decoders. The chosen demuxer must understand the container and preserve timing information, or playback can lose synchronization.