Home › AV over IP › AV over IP Encoder Decoder

AV over IP Encoder Decoder — How to Pair Endpoints, Size Bandwidth and Budget Latency

Most AVoIP problems at commissioning trace back to the pairing decision made at tender: the wrong codec profile, a decoder undersized for the wall, or an encoder feeding more simultaneous viewers than its stream design allows. This page covers the pairing rules we apply on real projects, with the numbers behind them.

What the Pair Actually Does

An AV over IP encoder turns an HDMI signal into a network stream. A decoder turns that stream back into HDMI. Everything that matters happens in between: how the stream is compressed, how many receivers can subscribe to it, and how much delay the whole chain adds.

Encode

HDMI in, scaled and compressed to H.264 or H.265, wrapped for IP transport. One encoder per source, regardless of how many screens show it.

Transport

Multicast on a managed switch with IGMP snooping, or unicast for point-to-point. The switch, not the endpoint, decides how far the system scales.

Decode

Stream in, decompressed and output as HDMI. Windowing, crop and rotation are applied here, which is why the decoder matters for video walls.

Control

Routing, presets and status live in the management layer. Good designs keep this separate from the media path so a control change never interrupts a live screen.

The Two Endpoints

Both units are 1GbE, Linux-based, and covered by a 3-year warranty. Firmware is ours, which is why we can change behaviour for a project rather than wait for a chipset vendor.

SpecificationQG-FBST52K — encoderQG-FBSR52K — decoder
RoleHDMI → network streamNetwork stream → HDMI
ResolutionUp to 4K@60 4:4:4 in, streamed at project profileUp to 4K@60 4:4:4 out
CodecH.264 / H.265, profile set per projectH.264 / H.265 decode
Network1GbE RJ45; fibre SFP variant on request1GbE RJ45; fibre SFP variant on request
Routing modesPoint-to-point, point-to-multipoint, many-to-manySingle source, or windowed multi-source for walls
KVMUSB 2.0 passthrough for keyboard, mouse, touchUSB 2.0 host emulation at the operator position
AudioEmbedded HDMI plus analogue inEmbedded HDMI plus analogue out, breakout on request
ManagementCentralised from QG-1169: status, presets, firmwareSame, plus per-panel crop, rotate and bezel settings
VariantQG-FBST5K (1GbE, alternative model)QG-FBSR5K (1GbE, alternative model)

Pairing Rules That Avoid Rework

Rule 1 — one encoder per source, not per screen

A camera shown on twelve displays is still one encoder and one multicast group. Buying twelve encoders is the most common over-spec on AVoIP quotes. Count sources and destinations separately before pricing anything.

Rule 2 — match the decoder to the wall, not to the encoder

If a decoder must show four windows on one panel, it carries four decode sessions. Spec the wall decoder for the window count, then choose the encoder for image quality.

Rule 3 — the codec is a project decision

H.265 halves the bitrate of H.264 at similar quality, which matters on congested uplinks. H.264 has wider interoperability with third-party recorders and older decoders. Pick per project, not per catalogue.

Rule 4 — use dual-role endpoints where the cable count hurts

Operator desks that both send a workstation and receive a feed need two units unless the endpoint is dual-role. FBSLTR002 / FBSLTR003 do both in one box, which halves rack space and switch ports.

Rule 5 — keep control traffic off the media VLAN

Presets, firmware and status polling belong on a management VLAN. Mixing them with multicast media is how a harmless firmware push becomes a visible glitch on the wall.

Rule 6 — spare one of each per twenty

Endpoint cost is small next to a site visit. A stocked spare encoder and decoder per twenty units turns a failure into a swap instead of a return-to-base.

Bandwidth: The Numbers Integrators Ask For

Bitrate is a design choice, not a fixed specification. The table shows the ranges we use when sizing a network, at 1080p60 and 4K30/4K60, for typical content.

ProfileResolutionTypical bitrateStreams per 1GbE portWhere it fits
H.264 high1080p608–16 Mbps~40–60Legacy interoperability, mixed estates, recording to third-party NVR.
H.264 high4K3020–30 Mbps~20–25Signage and presentation where motion is limited.
H.265 main4K6012–20 Mbps~30–40Default for new builds. Best balance of quality and uplink load.
Low-latency1080p6020–40 Mbps~15–25KVM positions and live camera work where delay is visible.
Visually lossless (project)4K60 4:4:480–200 Mbps~4–8Radiology, colour-critical review, high-end court evidence. Needs 10GbE uplinks.

Port counts assume a dedicated media VLAN with no other traffic. Always size the uplink for the worst-case simultaneous view, then add 30% headroom for control, recording and bursts.

Latency: Where the Milliseconds Go

Encode stage

Frame capture and compression, typically 15–40 ms depending on profile and GOP structure. Shorter GOPs reduce delay at the cost of bitrate.

Network stage

Switching and buffering, usually under 5 ms on a well-configured LAN. Jitter buffers added by the decoder to survive congestion are the hidden cost — hence QoS.

Decode stage

Decompression plus display sync, 10–30 ms. Displays themselves add 10–30 ms of processing that no AVoIP vendor controls.

Practical outcomes: a camera-to-wall chain in a control room lands around 80–120 ms end to end and is invisible to the operator. A KVM desk chain needs under 80 ms to feel direct, which is why KVM endpoints should run the low-latency profile instead of the bandwidth-saving one. Ask any vendor which profile their quoted latency figure was measured on — a number without a profile is marketing.

Choosing the Pair for the Room

Room typeEncoderDecoderWhy
Meeting room, 1–2 screensQG-FBST52K per sourceQG-FBSR52K per displaySimple any-source-any-screen with no matrix cabling; KVM optional for the room PC.
CourtroomQG-FBST52K for evidence PC, document camera, remote witnessQG-FBSR52K at bench, bar table, public galleryIndependent windows per position, no re-cabling between hearings, recording from the same layer.
OR and clinical teachingQG-FBST52K for endoscope, OR camera, PACSQG-FBSR52K in lecture hall and overflow roomsColour fidelity and low latency matter more than bitrate here; run the low-latency profile.
Control room deskDual-role FBSLTR002/003Dual-role FBSLTR002/003Each desk sends its workstation and receives a feed in one unit; halves ports and rack space.
Digital signage wallQG-FBST52K per playerQG-FBSR52K per panel groupH.265 keeps bitrate low across many panels; scheduled layouts handled by the platform.

Frequently Asked Questions

Is an AV over IP encoder decoder the same as an HDMI extender?
An extender carries one signal from A to B over a dedicated link, usually HDBaseT or fibre. An encoder/decoder pair puts the signal on a switched network, so any decoder can take any encoder's stream and one source can feed many displays at once. Extenders are cheaper for a single point-to-point run; they stop being cheaper the moment routing is needed.
Do encoder and decoder have to be the same brand?
Functionally yes in almost every case. Compression profiles, stream headers, control protocols and firmware behaviour are vendor-specific, and the few standards that exist (such as SDVoE or IPMX) are only interoperable when both ends implement the same profile. Mixing brands is a research project, not a procurement shortcut.
Can one encoder feed an unlimited number of decoders?
In theory multicast allows any number of subscribers. In practice the limit is the uplink out of the switch where the encoder sits, and whether IGMP snooping is configured properly. An encoder on a 1GbE port streaming at 16 Mbps will not carry a hundred viewers across a shared uplink without congestion.
What happens if the network drops a packet?
TCP-style reliable transport would stall the picture, so AVoIP uses RTP over UDP with a jitter buffer. A lost packet shows as a brief artefact or a repeated frame, not a frozen screen. Persistent artefacts point at congestion or a multicast group leaking to ports that did not request it — both are switch configuration issues, not endpoint faults.
Can the encoder take audio separately from video?
Yes. Embedded HDMI audio is carried by default, and analogue audio input can be embedded at the encoder so a lectern microphone travels with the presentation. Audio breakaway — video from one source, audio from another — is configured in the management layer.
How is firmware updated across a hundred endpoints?
From the QG-1169 platform, in staged batches, with version reporting per unit. Updating the whole estate in one push is fast and risky; we recommend a canary group of five endpoints, then the rest, so a bad build never reaches the wall.
Are the endpoints PoE powered?
Power options depend on model and are confirmed per project. Where PoE is available it removes a power outlet per endpoint, which matters in ceiling-mounted and behind-panel positions. Where it is not, the units take a local supply.
Do you supply encoders under our own brand?
Yes — enclosure labelling, boot logo, packaging and a project-specific default configuration are available through the OEM/ODM programme. Minimum order quantities and lead times are quoted per project.

Tell Us the Source Count and the Screen Count

We will come back with an encoder/decoder pairing, the bitrate profile, the switch port plan and the latency figure for your application — within 24 hours.

Request an Endpoint Pairing   See the AV over IP Platform

Related Pages

ADHOOPU

Professional AV manufacturer — Integrated AV Platforms, Distributed AV over IP, and AI Edge Computing & Video Analytics. OEM/ODM for system integrators, distributors and AV brands worldwide.

19 years of AV engineering · Shenzhen, China · 3-year warranty

Products

Integrated AV Platform QG-1169 AV over IP Encoders ST52K Encoder SR52K Decoder AI Video Analytics 615 AI Analytics Box

Solutions

Warehouse Safety Industrial Safety Oil & Gas Safety Hospital Safety Campus Security Parking Analytics Traffic Analytics Smart City Analytics Logistics Safety Retail Analytics Airport Analytics Smart Courtroom Surgical Teaching Command Center Education Meeting Room Smart Security Construction Site Factory & Industrial

Company

OEM / ODM About ADHOOPU FAQ Resources Contact
ADHOOPU · Room 1206, Block A, Dachanwan, Jingang Commercial Building, Xixiang Street, Bao'an District, Shenzhen, China
WhatsApp +86 138 2352 5531 · Email jimmy@adhoopu.com · Web https://www.adhoopu.com
© 2026 ADHOOPU. Integrated AV Platform · AV over IP Encoder / Decoder · AI Video Analytics · Professional AV OEM Manufacturer.