UdpTsInputConfiguration
Field |
Type |
Repeated |
Description |
id |
|||
source_name |
|||
host |
|||
port |
|||
stats_sampling |
|||
interface |
|||
timeout |
|||
rtp_decapsulate |
|||
cta708_charset |
|||
debug_dump_source |
When specified (nonempty) dump the raw source input for debug purposes to the given directory |
||
start_threshold |
Optionally drop data until input is 'stable' this will reduce latency and memory use in nearly all cases |
||
chaos_drop_packets |
|||
program_filter |
Optionally filter a single program from MPTS at decode time |
||
time_domain |
|||
health_sampling |
Per-input health sampling (connection_health events). Default on. |
||
allowed_source_addresses |
✅ |
R16: only accept datagrams from these source IPs (exact match, v4 or v6). Empty = accept from anywhere (existing behaviour). The ONLY admission control a connection-less input can have; rejected sources are logged (once per source) and counted, never delivered. |
|
report_ts_errors |
Whether to attach per-frame TR 101 290 (TS error) metrics to decoded frames. Only the qualityMonitoring node reads them; carrying them makes every frame’s quality-metrics map a per-frame allocation that is copied on each processing hop. Unset = attach (the historical behaviour). Set false on many-stream deployments that do no quality monitoring. |
||
report_ts_tables |
As SrtInputConfiguration.report_ts_tables / report_gop_structure: the per-table and per-GOP event streams are only emitted when asked for (the SDK asks iff a callback is registered). Unset = emitted, as before. |
||
report_gop_structure |