A video wall controller is usually quoted as a box: inputs on one side, display outputs on the other, and a licence for every layout you might want later. ADHOOPU builds the same function as a distributed layer — decoders drive the panels, an on-premise platform holds the layouts, and the wall grows by adding endpoints instead of replacing the chassis.
When an integrator asks a video wall controller manufacturer for a quote, four separate jobs are usually bundled into one product. Splitting them makes the cost and the risk visible.
Every source — operator workstation, NVR, presentation PC, camera, settop box — has to enter the system as a stream. In a chassis controller this is an input card. In a distributed design it is an encoder at the source, one per signal, no head-end frame.
Which stream occupies which rectangle of which panel, with bezel compensation, rotation for portrait panels and overlap for projection blends. This is the maths most chassis controllers charge a licence for.
A single image spread across nine panels must arrive at the same time everywhere, or the moving edge tears between screens. Genlock and frame-accurate start are what separate a wall controller from nine independent decoders.
Presets, scheduled layouts, KVM for the operator position, and a documented API so the wall can be driven from Crestron, Extron, Q-SYS or a customer SCADA system rather than a dedicated touch panel.
Both approaches put a picture on the wall. They differ in what happens when the wall changes, the room moves, or a board fails three years after handover.
| Decision | Dedicated video wall controller (chassis) | Distributed control (video wall over IP) |
|---|---|---|
| Signal path | Every source cable terminates on an input card at the chassis; a dedicated output cable runs to each panel group. | Sources enter as encoders on the existing LAN; one decoder per panel or per group sits behind the wall. |
| Scaling | Card slots are the ceiling. A 4×4 wall plus six more inputs consumes the frame, and the next wall needs another frame. | The switch is the ceiling. Adding panels means adding decoders on the same VLAN, not a new chassis. |
| Distance | Limited by the output interface; long runs need extenders or fibre transceivers per output. | The network already spans the building. Fibre uplinks between floors carry the wall and the rest of the AV traffic together. |
| Failure mode | Controller down = whole wall down. Spares are chassis-level and often discontinued early. | One decoder down = one panel down. The rest of the wall keeps running; the spare is a stocked endpoint. |
| Expansion after handover | New input types usually mean a new card generation. | New source = one encoder. New layout = a preset in the management layer. |
| Typical fit | Single room, fixed signal set, no IT involvement, short project life. | Multi-room, phased rollout, campus or command centre where IT already owns the network. |
Layout logic lives on the QG-1169 platform, not inside the displays and not inside a dedicated controller. Decoders receive their crop and position from the platform and render it locally, which keeps the wall working at the same latency as a single point-to-point link.
Whole-wall, quad split, picture-in-picture, crawler strip along the bottom row. Presets are stored with names operators recognise, and can be recalled from a panel, a browser or an HTTP call.
Pixel loss behind each bezel is compensated per display so a face is not cut in half at the seam. Portrait rotation is set per decoder, useful for menu boards and station displays.
Time-based layout switching for control rooms that run different walls on day and night shifts, or for retail and lobby walls driven by opening hours.
A stream can be dragged from one panel group to another without re-cabling. Windows stretch across groups for a single wide canvas.
USB 2.0 KVM travels the same link as the video, so the operator controlling the wall sits at the wall instead of in a separate equipment room.
TCP/UDP, HTTP and a documented command set. The wall accepts commands from third-party control systems and reports endpoint status back to them.
A useful rule for early budgeting: one decoder per panel for full-canvas walls, one decoder per group where each group always shows the same source. Bandwidth follows the stream count, not the panel count.
| Wall size | Decoders | Encoders (typical) | Notes from real projects |
|---|---|---|---|
| 2×2 (4 panels) | 4, or 1 if the wall always shows one stretched image | 4–8 | Mostly meeting rooms and briefing centres. Single switch is enough; keep the wall VLAN separate from office traffic. |
| 3×3 (9 panels) | 9 for full canvas, 3 if three columns each show one source | 8–16 | Traffic and security operations. Synchronised start matters here — nine panels showing a moving map will show the seam otherwise. |
| 4×4 (16 panels) | 16 | 12–24 | Command centres. Plan a 10GbE uplink between the wall switch and the core; do not share a 1GbE uplink with workstation traffic. |
| 8×8 (64 panels) | 64 | 24–48 | Control rooms and public lobbies. Multicast design and IGMP snooping are non-negotiable; document the VLAN plan before the first cable is pulled. |
| 120+ panels | 120+ | 40+ | Multi-building estates. Split into zones with routed multicast between them rather than one flat L2 domain. |
Two endpoint families cover wall projects. Both are manufactured in our own facility, so firmware behaviour and spare parts do not depend on a third-party chassis roadmap.
4K@60 4:4:4 decoding, H.264/H.265, HDMI output to the panel, windowing and crop applied locally. One unit per panel for full-canvas walls; one unit per group for fixed layouts.
HDMI plus audio in, stream out over 1GbE. Point-to-point, point-to-multipoint and many-to-many routing, with KVM passthrough when the wall doubles as an operator position.
The five functions in one box: encoding, decoding, storage, transcoding and control. Holds layouts, presets, schedules and endpoint firmware for the whole wall.
Dual-role endpoints (FBSLTR002 / FBSLTR003) are available where a position needs to both send and receive — typical in control rooms where an operator pushes a workstation to the wall and pulls a camera feed back to the desk.
Any source on any screen, KVM at the operator desk, and a wall that keeps running when one decoder fails. Layout presets mapped to incident levels rather than memorised button sequences.
Camera mosaic on the main wall, single-camera zoom on a side group, crawler for incident text. Long cable runs replaced by the existing fibre between floors.
Evidence display across a bench wall, document camera and remote witness on separate windows, recording triggered from the same control layer.
OR camera, PACS and endoscope feeds distributed to a lecture wall and to overflow rooms, with no re-cabling between procedures.
Menu boards and brand walls driven on a schedule, portrait panels rotated per decoder, content changes pushed from head office over the network.
Lecture capture to overflow rooms, signage walls in atrium spaces, all on the campus 1GbE backbone the IT team already maintains.
Eight questions that separate a wall which survives five years from one that gets value-engineered out at commissioning.
Per-panel licences look small on a 4-panel quote and become the largest line on a 64-panel one. Ask for the number at final panel count, not at tender.
In a chassis design, the answer is usually "the wall stops". Confirm the spare lead time in writing.
Ask for the mechanism, not the marketing word. Genlock, a shared clock, or buffered start — each behaves differently on a moving map.
H.264 at 20 Mbps and H.265 at 12 Mbps look different on a 1.8 mm LED wall. Get the two files and compare on site.
If yes, multicast design, IGMP snooping and QoS are your responsibility. Confirm who signs off the switch configuration.
Ask for the API document before purchase. If there is no document, integration becomes a project risk, not a feature.
Wall controllers live 7–10 years. Confirm that firmware and configuration backups are exportable and that older units stay supported.
A brand that resells an OEM chassis cannot change firmware behaviour for you. A manufacturer can. This matters most in year four.
Panels, sources, rooms and whether KVM is required. We reply within 24 hours with an endpoint count, a switch port plan and the layout-preset structure for the wall.
Every major platform we replace, compared at family and model level with migration notes.
Open the hub ›How to pair endpoints, what bandwidth each stream needs and where latency comes from.
Read the guide ›The system layer around the wall: management, recording, transcoding and firmware control.
Read the guide ›Appliance-managed, integrated and software-only architectures compared on total cost of ownership.
Open the matrix ›Encoders, decoders, KVM and centralised management for networked AV distribution.
View the platform ›Eight dimensions worth comparing before a platform is specified.
Open the guide ›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