The LED signal chain is the path video data takes from the playback source to the individual pixels on an LED video wall, and it decides whether content arrives frame-accurate and colour-correct or arrives late, torn and washed out. When we are called out to diagnose a misbehaving wall, the fault usually sits somewhere in that chain, not in the panels themselves. This guide walks the whole route from source to pixel, explains what each stage actually does, and shows where specifiers should focus attention and budget.
What is an LED signal chain?
The LED signal chain is the complete path video data travels from the playback source to the individual pixels on an LED video wall. It runs source → processor → distribution cabling → receiving cards → driver ICs. Each stage transforms the data, and overall image quality is capped by the weakest stage, which is why the chain must be specified as a system rather than as separate parts.
Key takeaways
- Work from the LED raster first, and derive it from the module grid. An LED video wall built from eight columns and eight rows of 600 × 337.5mm cabinets at 2.5mm pitch is exactly 1920 × 1080 pixels, not “roughly Full HD”.
- The chain has five main stages: source, processor, distribution, receiving cards and driver ICs. A weakness at any stage caps everything downstream.
- The processor is the single most consequential purchase in the chain. It handles scaling, colour management, frame-rate conversion and bit depth. Panels can only display what the processor hands them.
- Port capacity is a fixed budget spent three ways: pixel count, frame rate and bit depth. Raise one and the others shrink.
- Ring-wired data loops mean a single cable failure drops zero cabinets, and a hot-standby processor covers the launch-day worst case.
- Gigabit copper is rated to 100 metres, but marginal runs fail as flicker and dropped cabinets long before outright blackout. Fewer conversions and cleaner runs mean fewer faults.
At a glance: the LED signal chain
The LED signal chain has five stages: source, processor, distribution, receiving cards and driver ICs. Data flows from the source through the processor, out over Ethernet distribution to a receiving card in each cabinet, and finally to the driver ICs that turn data into light. Image quality is capped by the weakest of the five.
| Stage | What it does | Typical hardware | Common failure symptom |
|---|---|---|---|
| Source | Plays or generates content | Media player, PC, camera feed, broadcast router | Wrong resolution or frame rate at input |
| Processor | Scales, colour-manages and maps content to the wall | Brompton SX40, Novastar all-in-one or sending card | Soft scaling, banding, colour drift |
| Distribution | Carries pixel data to the panels | Gigabit Ethernet, fibre for long runs | Dropped panels, flicker on marginal cables |
| Receiving card | Decodes data for its section of panels | Card inside each cabinet | A single cabinet showing garbage or black |
| Driver ICs & pixels | Turn data into light, thousands of times per second | Driver chips on each module | Dead lines, ghosting, blotchy low-brightness tones |
Start with the raster: the wall’s native canvas
Every LED signal chain design begins with the wall’s native pixel count, and the correct way to calculate it is from the module grid: modules wide × pixels per module, and modules high × pixels per module. Dividing physical dimensions by pixel pitch looks like a shortcut, but it silently breaks on non-integer module counts and non-square modules, and it is how tenders end up quoting resolutions the cabinets cannot actually build.
Take a common fine-pitch format: 600 × 337.5mm cabinets at 2.5mm pitch, each carrying 240 × 135 pixels. Eight columns and eight rows give a 4.8m × 2.7m wall of exactly 1920 × 1080 pixels — a native 16:9 Full HD canvas. A standard 1080p source maps one-to-one with no scaling, which keeps the whole chain simple.
Now compare a 5120 × 1440 pixel wall, roughly 7.68m × 2.16m at 1.5mm pitch. That canvas is wider than UHD but shorter than a 16:9 4K frame, so a normal 3840 × 2160 source will not fill it one-to-one. You either author content at 5120 × 1440, scale from another format, crop, or accept unused area. “Make it 4K” is not a content brief for a wall like this; the brief should state the exact pixel canvas, frame rate and safe areas.
Content authored to the native raster plays back 1:1 with no scaling, and scaling always costs sharpness. When specifying fixed-install LED video walls, settle the pitch before buying control hardware, because pitch changes the raster, the processor load and sometimes the content workflow. Our pixel pitch guide covers how pitch, viewing distance and cabinet layout determine that native canvas.
Stage 1: the source
The source is the first stage of the LED signal chain: the media player, playback PC, camera chain, presentation switcher or broadcast feed that generates content. The single most important specification decision here is matching the source output to the wall’s native pixel canvas and holding the entire chain to one frame rate. A 50Hz broadcast feed into a chain configured for 60Hz forces frame-rate conversion somewhere, and conversion introduces judder or latency. In the UK, 50Hz often makes sense where video production and lighting are involved; 60Hz is common for IT and signage sources. Decide at design time and hold every device to it.
Most fixed-install sources hand off over HDMI, DisplayPort or 12G-SDI. HDMI is practical in meeting rooms and signage but awkward over distance without proper extension, and EDID handling matters: if the source cannot read the display chain correctly, a laptop may output 1920 × 1080 to a 3840 × 1080 wall. DisplayPort suits PCs and media servers driving large canvases at higher bit depth, and SDI locks cleanly into broadcast infrastructure. Bandwidth is the practical trap. 4K at 60Hz with 10-bit 4:4:4 colour exceeds the 18Gbps ceiling of HDMI 2.0 and needs an HDMI 2.1 path end to end, so check every cable, extender and switcher in the run before assuming the signal survives it.
Stage 2: the LED video wall processor (Brompton and Novastar)
The processor takes the source signal and turns it into panel-ready pixel data. It scales content to the wall’s canvas, applies colour management, converts frame rates where unavoidable, sets bit depth, and maps regions of the input to physical positions on the wall. Every visual property of the finished LED video wall passes through it.
Two names dominate this stage in professional work: Brompton and Novastar.
Brompton’s Tessera SX40 is the reference point at the top end. It processes a 4K canvas, supports high bit-depth workflows, and pairs with Tessera receiving cards for colour management and camera work. For broadcast studios, virtual production volumes and premium corporate installs — the sort of projects where we specify DFC Series fixed-install panels — Brompton processing is usually the right call.
Novastar covers a wider span, from all-in-one controllers such as the VX1000 up to 4K-class units like the MX40 Pro, and its sending-card architecture drives a large share of the world’s installed fixed LED. Novastar’s technical pages document the product families. For mid-range fixed installs on DX Series panels, Novastar gives strong results at a sensible cost.
The critical maths at this stage is port capacity. A single gigabit output port carries a fixed data budget, spent three ways: pixel count, frame rate and bit depth. Raise any one and pixel capacity per port falls. A processor that handles a wall comfortably at 8-bit 60Hz may need extra ports, or a bigger unit, the moment the brief moves to 10-bit or higher refresh. Our Novastar processor calculator does this arithmetic for you: enter the wall dimensions and pitch and it returns the pixel load and the class of processor that covers it.
The trap is comparing processors by one headline pixel count. That number changes with frame rate, input type, colour settings, redundancy mode and output configuration. Under-specifying the processor is the design error we see most often in tenders: the panels get the attention because they are the visible line item, and the processor gets treated as an accessory. It is the opposite. The processor is the ceiling on what the panels can ever show.
Planning a fixed-install LED video wall? Start with our LED video walls hub to see the DFC and DX fixed-install systems these chains are designed around.
Stage 3: distribution
From the processor, pixel data travels to the panels over Ethernet: gigabit copper for most fixed installs, fibre where runs are long or the environment is electrically noisy. Each processor port feeds a chain of cabinets, and data loops through from cabinet to cabinet. Three design decisions live at this stage.
Cable topology and mapping. Cabinets are daisy-chained per port, and the chain order must match the processor’s mapping. A tidy, documented topology — labelled runs, clear start and end points per output, a saved configuration file the service team can recover — makes commissioning quick and fault-finding possible. Treating mapping as a software task to sort on site costs real time if the cabling went in without a data plan. On fixed walls we want processor location, cabinet order and data routes agreed before containment closes, and that matters even more on DVO Series outdoor permanent systems, where routes and access are harder to change later.
Redundancy. Professional receiving cards accept data from both ends of a loop. Wire the chain as a ring, with the processor feeding one end and looping back into the other, and a single cable failure mid-chain drops zero cabinets: data simply arrives from the other direction. On broadcast and mission-critical installs we go further with a backup processor on hot standby. Decide what failure is acceptable. For a corporate atrium wall, a spare player and a documented processor file may be enough; for a live broadcast backdrop, duplicate sources and a tested changeover plan. A redundant system nobody has tested is only a drawing.
Cable quality and length. Gigabit over copper is rated to 100 metres, but marginal terminations and cheap cable show up as flicker and dropped cabinets long before outright failure. On permanent installs we test and certify every run.
Stage 4: receiving cards
Inside each cabinet sits a receiving card: a small board that takes the incoming Ethernet stream, extracts the pixel data belonging to its cabinet, and drives the modules connected to it. Receiving cards are where processor ecosystems become real. Brompton Tessera cards only work with Brompton processors, Novastar cards with Novastar sending devices. That pairing, made at specification time, locks the LED video wall into one ecosystem for its service life, which is why we treat processor selection as a whole-life decision rather than a line-item comparison.
Receiving cards also hold the calibration data: per-pixel brightness and colour corrections measured at manufacture, stored on the card and applied in real time. When a wall shows a visibly different cabinet after a swap-out, the usual cause is calibration data that was never uploaded to the replacement card, not a bad panel.
Stage 5: driver ICs and pixels
The final stage is the driver ICs on each module: chips that switch individual LEDs on and off thousands of times per second. Pulse-width modulation controls brightness. The LED is either fully on or fully off, and perceived brightness comes from the ratio of on-time to off-time within each refresh cycle.
This is where refresh rate is actually generated. A wall spec quoting 3840Hz refresh is describing the driver-stage PWM cycle, and it matters for two audiences: cameras, which see banding on low-refresh walls, and viewers close to fine-pitch product, who can perceive artefacts during fast motion. Driver IC quality also governs low-brightness performance. Cheaper drivers lose greyscale precision when a wall is dimmed for a dark studio or an evening event, producing blotchy near-black tones. On premium fixed-install work we specify panels whose driver stage holds greyscale at the low brightness levels studios actually run.
On a properly configured chain, the whole journey from source frame to photons typically completes in one to four frames end to end — roughly 16 to 66 milliseconds at 60Hz. For a retail wall nobody will notice. For a virtual production volume where camera and wall must stay in lockstep, those frames are the thing you design around, and Brompton Technology publishes latency figures for the Tessera range for exactly that reason. It is one more reason processor choice leads the specification.
From the field: when the LED signal chain fails
The chain failure I still bring up in project meetings was a corporate lobby wall where the client’s integrator had specified beautiful fine-pitch panels and then fed them from a processor that was flat out of headroom. The wall looked soft and nobody could work out why; the panels tested perfectly. My first question on site was “what’s the pixel load per port?” and the answer explained everything: the processor was scaling a 4K source down to fit its own capacity, then the wall was displaying that compromised image at native pitch. We swapped the processing for hardware with genuine headroom, kept every panel, and the same wall looked like a different product.
Since then my rule is simple: decide the canvas from the module grid, run the processor maths before anyone falls in love with a pitch number, keep the conversions down, label the data runs, and save the processor file somewhere the service team can actually find it. It is the check we now run on every fixed-install quote before a panel is ordered.
LED signal chain: frequently asked questions
What is an LED signal chain?
The LED signal chain is the complete path video data travels from the playback source to the individual pixels on an LED video wall. It runs source → processor → distribution cabling → receiving cards → driver ICs. Each stage transforms the data, and overall image quality is capped by the weakest stage, which is why the chain is specified as a system rather than as separate parts.
Why does pixel pitch affect the signal chain?
Pixel pitch sets the wall’s native pixel count. A smaller pitch gives more pixels over the same physical area, which increases processor load and may change the source format. A 4.8m-wide LED video wall at 2.5mm pitch is 1920 pixels wide; the same width at 1.5mm pitch is 3200 pixels wide — a very different processing job.
How many panels can one processor port drive?
It depends on the data budget. A gigabit port moves a fixed amount of data per second, and more pixels, deeper colour or a faster refresh all eat into it. A fine-pitch wall at 10-bit colour consumes far more per square metre than a coarse-pitch wall at 8-bit. Use a processor calculator to convert wall dimensions and pitch into pixel load before selecting hardware.
Is 4K always the right source format for an LED video wall?
No. Many LED video walls are not 3840 × 2160. A wide wall might be 5120 × 1440, and a portrait wall taller than it is wide. The right source format is the one that matches the wall’s native canvas, or feeds the processor with a deliberate, controlled scaling plan rather than a last-minute fix on site.
What causes one cabinet on an LED video wall to go black?
A single dark cabinet usually points to the receiving card or the data loop at that position: a failed card, a marginal Ethernet termination, or a break in the daisy chain. If the chain is wired as a redundant ring, data arrives from the other direction and a single cable fault drops nothing. Panels themselves are rarely the cause of a whole-cabinet blackout.
Can I mix Brompton and Novastar equipment in one chain?
No. Receiving cards only work with processing from the same ecosystem: Brompton Tessera cards pair with Brompton processors, Novastar cards with Novastar sending devices. The choice is made when the panels are specified, because the cards live inside the cabinets, and it determines your processing options for the wall’s entire service life.
What causes flicker or banding on camera?
Camera issues can come from refresh rate, scan behaviour, shutter speed, dimming level, source frame rate or processor configuration, and the fix is rarely one setting. For camera-facing walls, check source frame rate, processor mode, brightness level and camera shutter together, then store the working configuration as a preset so it can be recalled.
Does the signal chain add latency?
Yes. Every processing stage adds some delay, and a well-configured chain typically totals one to four frames end to end, roughly 16 to 66 milliseconds at 60Hz. For signage and corporate walls this is imperceptible. For broadcast, live camera magnification and virtual production it must be measured and managed, which is why those sectors favour processors with low, consistent, published latency figures.
Conclusion
An LED video wall specification that starts with panels and treats everything else as accessories gets the priorities backwards. The panels display exactly what the chain delivers — no more. Derive the raster from the module grid and get it confirmed in writing, author the source to that canvas, size the processor with real headroom, wire the distribution as a redundant ring, and confirm the receiving cards carry the calibration data, and the pixels will do their job without drama. That is the LED signal chain, end to end, and it is where the performance of the wall is actually decided.
If you are planning a wall and want the LED signal chain designed properly from day one, use our LED screen configurator to turn a physical size into a full pixel specification, or call us on +44 (0)203 489 9878 or get in touch — we will run the processor maths, the port counts and the redundancy plan before a single panel is ordered.
