One encoder at the source can serve destinations with different latency and reliability needs, while Edge Apps let integrations change without adding another field computer.
One input. Several delivery paths.
Match each output to its destination
The encoding measurement is one part of the end-to-end latency budget.
The engineering view
Videon measures less than 40 ms through video processing and encoding. That is the device contribution to latency, not the delay a viewer experiences. Transport, recovery, decoding and display all add time.
In Videon’s WHIP reference workflow, an appropriately configured end-to-end path achieves less than 200 ms glass to glass. Both figures depend on their measurement endpoints and test conditions, which Videon can provide for a specific workflow.
One live input can produce separate bitrate and resolution profiles for different destinations. A WHIP output can prioritize responsiveness while SRT or Zixi carries a contribution feed with more recovery, RTMP reaches a platform endpoint, and a local recording runs on the same device.
Operational result: one capture point can serve workflows with different latency and reliability requirements, reducing separate encoders and field setup where capacity permits.
Where low latency changes the workflow
The right latency target depends on what the viewer or operator needs to do. These examples show where a fast encoder path and independent outputs can make a practical difference.
01 / Live betting
For in-play betting and trading views, a delayed video feed can make odds, data and pictures feel out of sync. A low-latency output can feed the viewing path while a separate, more buffered contribution path prioritizes recovery.
End-to-end timing must be measured alongside the data feed, distribution platform and player.
02 / Multiviewers
Control rooms and remote operators need timely views of live sources to spot faults, choose a feed and respond. Independently configured profiles can supply a low-delay monitoring path while full-quality outputs continue to their destinations.
Source synchronization and the multiviewer’s own buffers remain part of the latency budget.
03 / Live production
Directors, graphics operators and producers need a current picture when making on-air decisions. A low-latency path supports coordination, while resilient SRT or Zixi contribution and recording can run in parallel from the same input.
Production control should be tested across the complete return and contribution paths.
04 / Interactive experiences
Live auctions, remote interviews, coaching and interactive fan experiences become harder when the audience responds to an old frame. WHIP can support a responsive path where the receiver and playback workflow are configured for low delay.
The application, network and display determine the actual glass-to-glass result.
01 / Latency as a system
Glass-to-glass latency includes each stage between the camera and the viewer. An encoder claim describes only one part of that path.
| Stage | What adds time |
|---|---|
| 01 · Capture and ingest | Frame acquisition, synchronization and input buffering. |
| 02 · Video processing | Scaling, color processing, metadata handling and frame movement. |
| 03 · Encoding | Codec, GOP structure, B-frames, rate control and quality settings. |
| 04 · Packaging and transport | Muxing, protocol buffers, encryption, retransmission and FEC. |
| 05 · Network | Distance, congestion, packet loss, path changes and bandwidth. |
| 06 · Playback | Receiver jitter buffer, decode, render and display refresh. |
| Output family | Typical engineering range | Primary tradeoff |
|---|---|---|
| WHIP | Less than 200 ms glass to glass | Interactive latency |
| RTMP | Approximately 400 ms to 2.5 s | Broad platform compatibility |
| SRT or Zixi | Approximately 500 ms to 3+ s | Recovery and network resilience |
| Low Latency DASH | Approximately 2 to 3 s | HTTP scale and distribution |
| HLS | Approximately 6 to 30 s | Ubiquitous playback |
Ranges depend on configuration and workflow; they are not protocol guarantees.
02 / Platform architecture
The native media path handles time-sensitive ingest, encoding and output. Containerized applications use approved device interfaces for functions that can be developed and deployed independently.
LiveEdge® Compute can run custom Docker images from registries such as Docker Hub or Amazon ECR. LiveEdge Cloud manages configuration, environment variables, ports, restart policies, storage, approved device mappings, status and remote logs across devices.
Operational result: teams can update an integration across sites and troubleshoot it without shipping a computer to every venue.
Containers do not replace the native real-time encoder. An application must have access to the media or metadata interface it needs, fit within device resources and leave encoding performance intact. Other encoders support containers; the useful comparison is which interfaces, fleet controls and isolation each platform provides.
03 / Extending the edge
Choose native code for deterministic timing and hardware access; choose an Edge App when exposed streams or metadata are sufficient and independent updates are valuable. Some workflows need both.
| Implementation | Best when | Examples |
|---|---|---|
| Native media pipeline | Timing, hardware access and deterministic performance are essential. | Core encoding, captions, VANC and tightly integrated outputs. |
| Containerized Edge App | The workflow can consume exposed media or metadata interfaces and benefits from independent iteration. | Protocol adaptation, validation, forwarding, automation and partner integration. |
| Hybrid implementation | A native pipeline provides media while an app supplies control, orchestration or specialized transport logic. | Emerging transports, data-driven SCTE-35 and cloud-specific workflows. |
Integrated output for reliable contribution over unmanaged networks.
A container can validate an incoming MPEG-TS stream, buffer network bursts and forward it through a RIST implementation. Redundancy can extend to packet-level merge approaches such as ST 2022-7. This is an active R&D effort, not a generally available production feature.
Media over QUIC is being developed for scalable low-latency ingest and distribution. A programmable edge provides an environment for prototyping as specifications and ecosystem support mature.
04 / Operational value
The business case depends on the deployment: fewer field components, consistent remote configuration and the ability to introduce a transport or integration without replacing every encoder.
Match each output’s codec, buffer and recovery settings to its destination. Qualify container resource use under the full MBR load, then test packet loss, jitter, bandwidth and failover before fleet deployment.
Use one managed device to encode, record and deliver multiple feeds where the tested workload allows. Push application configuration centrally and inspect logs remotely. This shortens site work and limits truck rolls when integrations change.
Technical references
Latency ranges reflect Videon engineering measurements and representative workflow configurations, September 2026.
Ready to encode at the edge
Whether you need a WHIP interactive path, resilient contribution, simultaneous renditions, or Edge Apps that change without another field computer—we’ll help you design and validate the workflow.
Videon
60 Decibel Rd, Suite 107
State College, PA 16801
