Ndi
NDI workflows are excellent for high-quality low-latency production inside controlled networks, but they require clear boundaries when moving to cloud and internet delivery. This guide covers practical NDI architecture decisions. For this workflow, Player & embed is the most direct fit.
What it means
NDI is an IP video transport ecosystem commonly used in studio and LAN environments. Production use requires transport planning, conversion boundaries, and monitoring for packet behavior.
Decision guide
- Define NDI scope: local production only or cloud-connected workflow.
- Choose conversion points between NDI and streaming protocols.
- Set bandwidth and network segmentation policy.
- Define operator monitoring and fallback outputs.
- Automate orchestration with API where possible.
Useful mapping: Ingest and route, Video platform API, Calls and webinars.
Latency budget and architecture budget
- NDI source capture
- Conversion/transcode boundary
- Distribution path
- Playback/render path
Reference: SRT to NDI cloud pattern and SRT low latency.
Practical recipes
Recipe 1
- Keep NDI traffic in controlled network segments.
- Use protocol gateway for internet delivery.
- Track packet loss and jitter at boundaries.
Recipe 2
- Create dual outputs: operator low-delay and audience stable path.
- Use fallback stream for failure scenarios.
- Document failover procedures.
Recipe 3
- Automate route creation via API.
- Attach monitoring to each route.
- Audit config drift regularly.
Practical configuration targets
- Dedicated network path for NDI transport.
- Keyframe cadence 1-2s on converted outputs.
- Stable ABR ladder on distribution side.
Limitations and trade-offs
- NDI over unmanaged internet is fragile.
- Conversion increases complexity and cost.
- Cloud scaling needs strict orchestration discipline.
Common mistakes and fixes
- Mistake: direct WAN NDI without safeguards. Fix: gateway protocol boundary.
- Mistake: no route observability. Fix: per-route metrics and alerts.
- Mistake: manual operations only. Fix: API-driven route lifecycle.
Rollout checklist
- Network segmentation validated.
- Gateway and fallback tested.
- Monitoring and runbooks active.
Example architectures
Studio NDI core with SRT/RTMP gateway to cloud distribution gives a practical balance of quality and reliability.
Troubleshooting quick wins
- Unstable feed: inspect boundary packet loss.
- Latency spikes: check conversion queue and buffer policy.
- Output failures: fail over to backup route and isolate source path.
Next step
Continue with NDI cloud conversion, streaming video architecture, and Ingest and route.
