PES packing of audio frames (where applicable). Lower values reduce audio latency, but increase bandwidth due to encapsulation overhead. Minimum 1, applicable maximum bounded by the byte size of a PES packet. If omitted the default is 16. The packing is an upper bound, not the cadence: audio is also released whenever its oldest frame is pcr_delay_ms (default 500) old, so at the default a 48 kHz AAC stream goes in ~341 ms (16-frame) bursts and a 32 kHz one in ~500 ms bursts.

cta708_charset

Cta708Charset

Charset used when emitting CTA-708 P16 (16-bit) captions, for producing East-Asian captions in a legacy DBCS rather than Unicode. Defaults to Unicode.

flush_partial_packets

bool

When true, the egress packet queue stops holding a sub-packet tail while it waits for the next chunk to fill a 7 x 188 = 1316-byte payload: at every chunk boundary it sends whatever whole TS packets it has, as a shorter payload. Removes up to 1316-bytes-worth of delay between the muxer and the socket, which on a low-bitrate audio-only stream is tens of milliseconds; costs the extra per-payload headers.

Simple encoder only — rejected with the broadcast encoder, whose whole point is a constant rate the muxer has already paced. No effect on the file output, which does not queue.

timecode

AncillaryTimecodeSettings

When set, every video stream gets a SMPTE 2038 stream beside it carrying its ATC (SMPTE ST 12-2), one packet per frame.

timecode_pid

OptionalInt

The PID of that SMPTE 2038 stream; by default the first free PID after its video’s.