LED Video Basics

Introduction

A projector could be wrong and still be usable. It would be dim, or soft, or slightly the wrong shape, and the show went on. An LED wall is not like that. It is either specified correctly or it is visibly wrong in a way a client can see from the back of the room and describe in a complaint.

The good news is that almost every one of those decisions is arithmetic made weeks before the truck arrives, and arithmetic can be learned.

Why this book has specifications in it

Ask why a wall looks wrong on camera and the honest answer is usually a number: the refresh rate was too low, the pitch was too coarse for the viewing distance, the brightness was left at full, the data ran past what one port could carry. None of that is a matter of taste.

So this book gives you the figures and, more importantly, the reasoning that produces them. Panel specifications change every year. The arithmetic does not.

What you need before this book

You should know what a video signal is and how it gets from a source to a display. If that is shaky, read AV Fundamentals 101 first. This book assumes you can trace a signal and are ready to learn what the boxes in the middle are doing.

How to Use This Book

Each chapter follows the same shape, and the shape matters more than it might look. You read the ideas, then you see them play out in a real situation, then you get the procedure, then you are shown what goes wrong and how to recognise it early. Reading only the parts that look new is the fastest way to arrive on a job holding half of something.

1. Read the chapter body first. This is the understanding layer, and it is written to be read straight through.
2. Read On the job. The same material as a real situation on a real show, which is where it stops being abstract.
3. Work through How to do it. These are procedures you can follow on a job this week.
4. Study What goes wrong. Experienced technicians are not people who never make mistakes; they are people who recognise mistakes early.
5. Do the exercise before reading the answer key. Recognition feels like learning and is not.
6. Take the checklist with you. Photograph it on your phone. Use it until you no longer need it.

Keep a notebook

One habit is worth starting today, before the first chapter: keep a small notebook and write the technical details in it as you meet them. LED work is arithmetic more than it is opinion — pixel pitch, the viewing distance that pitch implies, panel counts across and down, what one processor can actually drive, brightness, refresh rate. These are the numbers that decide whether a wall is right weeks before anybody switches it on.

Write it by hand rather than photographing this page. Copying a number puts it on paper; rewriting it in your own words is what puts it in your head, and deciding how to word it is most of the understanding.

Then use it twice. Read a few pages on the way to a job, which takes five minutes and turns things you half-know into things you know. And carry it on the job, because the number you cannot remember is always the one somebody is waiting for.

A note on the words

LED vocabulary is inconsistent and you should expect that rather than be caught by it. The thing you hang on the wall is a panel to one manufacturer, a tile to another and a cabinet to a third, and this book uses panel and tile as the same word because the trade does. Pitch is quoted as a number in millimetres or as a code on a spec sheet. The box that drives the wall is a processor to some people and a sender or a scaler to others, depending on how much of the job it is doing.

None of that is worth arguing about. What matters is that you can say what a box does, because a person who can describe the function will be understood anywhere, and a person who only knows one brand's word for it will not.

A note on safety

LED walls are heavy, and most of them end up in the air. Suspended loads are a life-safety matter: what is above an audience is not a place to find out you were nearly right. The panels also carry mains power to every column, and a wall is assembled by people working at height while it is going up.

This book gives you awareness, not authorisation. Rigging and electrical work are trades with their own tickets, and knowing why a thing is done is not the same as being certified to do it. If you are ever unsure whether something is yours to touch, the answer is to ask.

Where this book sits

This book assumes you have read Video Basics, or that you already know what is in it.

Experience counts instead. If you have time on video crews where you already know what a signal chain and a processor are, start here.

If none of that is true yet, read Video Basics first.

What LED Video Is

An LED wall is not a screen in the way a projector and a surface are a screen. It is a large number of small light sources arranged in a grid, assembled from modular cabinets, fed with power and data, and told what to do by a processor. Understanding it as a construction rather than as a display explains nearly everything about how it behaves.

Direct-view screens made of panels.

LED screens

LED video screens are direct-view displays made of LED panels (tiles) — modules covered in tiny LEDs that, assembled together, form a large screen that emits the image directly (the LEDs themselves make the picture). Unlike a projector (which throws light onto a surface) or a TV (a single fixed panel), an LED wall is built from many panels into a screen of (almost) any size and shape. LED is now the dominant technology for big screens — concert stages, broadcast backdrops, large events, signage, and creative installations. Understanding LED as direct-view, panel-built screens frames this course. LED screens are built from light-emitting panels.

LED's advantages drive its dominance: it's very bright (works in any light, even daylight), scalable (build any size and shape from panels), seamless (no bezels between panels), and impactful (vivid, high-contrast). This is why LED has taken over big-screen video. Understanding what LED is and why it's everywhere frames the course. LED is the bright, scalable, dominant big-screen technology.

The LED system

An LED video system has a few key parts. The panels (tiles) — the modules that form the screen surface. The processor (LED processor/controller) — the device that takes the video signal and drives the panels, mapping the image across them correctly. And the supporting infrastructure — the rigging/structure that holds the wall, the power that feeds the panels, and the data cabling that carries the image from the processor to the panels. Together these make a working LED wall. Understanding the parts of an LED system is fundamental. The panels, processor, and infrastructure make up an LED wall.

These parts work together: the processor receives the video and sends the right part of the image to each panel over the data cabling, the panels display it, and the rigging and power hold and feed the wall. Understanding this system frames how LED walls are built and run. Knowing the LED system's parts frames the course. An LED wall is panels, a processor, and the infrastructure to support and feed them.

How a pixel makes a colour

Everything a wall can do comes from one arrangement repeated hundreds of thousands of times. Each pixel is three emitters — red, green and blue — sitting close enough together that the eye cannot separate them at any sensible distance. Drive all three at full and you get white; drive red and green and you get yellow; drive none and you get whatever the panel's surface looks like with the power off, which is why black on an LED wall is really dark grey.

That last point matters more than it sounds. A projector makes black by putting no light on a screen that is already lit by the room. An LED wall makes black by switching its own emitters off, and the surface between them is deliberately matt and dark to keep it that way. It is the reason LED holds contrast in a lit room where projection collapses, and the reason a dusty wall looks flat: dust is a pale surface sitting between the emitters.

The three emitters are packaged together, and how they are packaged is a specification worth knowing by name. In SMD — surface-mounted device — the three sit in one small package on the board, which is what almost every indoor panel uses. In COB — chip on board — they are mounted directly and covered with a protective layer, which survives handling better and is appearing on close-pitch panels for exactly that reason. You do not need to choose between them often, but you will see both on a spec sheet and they are not interchangeable when you are matching panels.

Module, cabinet, wall

Three words describe three different things, and using them loosely is how orders go wrong.

A module is the small tile of LEDs, often about 250 by 250 millimetres. It is the thing that carries the pixels and the thing you replace when pixels fail. It has no frame and no connectors of its own worth speaking of; it plugs into what holds it.

A cabinet — the trade also says panel, and this book uses both — is the frame that holds several modules, commonly 500 by 500 millimetres, so four modules to a cabinet. The cabinet is what carries the locks, the handles, the power and data connectors and the receiving card. It is the unit you hire, count, hang and carry, which is why every figure that matters is quoted per cabinet: six to ten kilograms, two hundred to five hundred watts.

A wall is a grid of cabinets. It has no size of its own; it has a count. That is the whole reason the arithmetic in this chapter exists, and it is why "how big is the wall" is a question nobody on a load-in can answer until somebody says how many cabinets across and how many high.

Get the words right and a conversation with a rental house takes a minute. Get them wrong and you order modules when you meant cabinets, which is a difference of about four times the money and none of the hardware you need to hang anything.

The parts, and what each one does

Every LED wall, from a small corporate backdrop to an arena rig, is made of the same chain of parts. Learn the chain and any system becomes readable.

How signal actually reaches a pixel

• A source produces a picture, for example a media server or a laptop.
• A switcher or scaler selects and formats it to the wall's exact canvas size.
• The processor divides that canvas into regions and sends each region down a data port.
• Each data port feeds a chain of cabinets, one after another, in a defined order.
• The receiving card in each cabinet takes its slice of the data and drives its own LEDs.

The two processing ecosystems

Almost every wall you meet runs on one of two processing platforms, and knowing which one you are standing in front of tells you where the controls are before you touch anything. Brompton Technology's platform is Tessera. NovaStar's is its own range of sending devices and the software that drives them. They do the same job — take the show picture, decide what every pixel does, and hold the wall consistent — and they do not do it the same way.

A wall is therefore specified as a pairing rather than as panels alone. ROE Visual panels are processed by Brompton, and only by Brompton, so ordering ROE decides what has to be on the truck. Other manufacturers are more open. Either way, which panels and which processing are one question with two halves, and answering only the first is how a wall arrives that nobody can drive.

Learn one platform properly rather than both badly. The vocabulary carries across — brightness, gain, calibration, frame rate and redundancy are the same ideas on either — and the second is then a matter of finding where the same controls live.

The brands you will actually see

A short list covers most of what turns up on a truck. ROE Visual, Absen, INFiLED and Theatrixx are names worth knowing by sight, and there are a handful more depending on where you work and who your rental house buys from. Panels also arrive under names nobody recognises, and some of those are perfectly good.

Price separates panels, and country of manufacture does not tell you which is which — several of the best manufacturers in the trade are Chinese, and so are several of the worst. Four things separate a good panel from a cheap one on a job, and every one of them is something you can check yourself.

On the job

A client asks for a video wall four metres wide and two and a half metres high for a conference, and wants to know what it will cost and whether it will look good.

You cannot answer either question yet, because nothing has been decided. So you ask three things. How close will the nearest person be sitting? About three metres, front row. Will it be on camera or streamed? Yes, both. Where is the power and how much is available?

Those three answers do almost all the specifying. A three-metre front row means a pitch of roughly 2.6 to 3 mm, because coarser than that and the front row will see the individual dots. Camera means the refresh rate must be high, at least 3,840 Hz, and that the wall will run at reduced brightness. A four by two and a half metre wall at 2.6 mm pitch is about 1,540 by 960 pixels, so a standard 1920 by 1080 source will need scaling to that exact canvas.

Now you can specify, and now you can cost it. The client asked a question about size. The answer was determined by seating, cameras, and power.

How to do it

The order in which an LED wall is specified. Do it in this order and the decisions cascade cleanly; do it in any other order and you will redo it.

1. Establish the physical size available, including height limits and rigging points.
2. Establish the distance from the wall to the nearest viewer. This sets the pixel pitch.
3. Establish whether it will be on camera. This sets the refresh rate and the working brightness.
4. Establish the ambient light. This sets the brightness requirement.
5. Choose the panel, and calculate the wall's true resolution from its width and pitch.
6. Calculate weight and confirm the rigging or ground support can carry it.
7. Calculate power at maximum draw and confirm the supply and breakers.
8. Calculate data ports required and confirm the processor can drive the canvas.
9. Only now decide how show content will be scaled and fed to that canvas.

What goes wrong

Practice it

Answer before checking.

Pixel Pitch & Brightness

Pixel pitch and brightness are the two numbers a client will never ask about and will always judge you on. Get them right and the wall simply looks good. Get either wrong and no amount of content or processing rescues it.

The key specs of LED screens.

Pixel pitch and resolution

A defining LED spec is pixel pitch — the distance between adjacent LEDs/pixels (e.g., 2.6mm, 3.9mm), usually in millimeters. Smaller pixel pitch means the pixels are closer together, giving higher resolution (more detail) and allowing closer viewing (the audience can be nearer without seeing the pixels). Larger pixel pitch (pixels farther apart) is used for bigger screens viewed from farther away (where fine detail isn't needed). Pixel pitch (with the screen size) determines the wall's total resolution. Understanding pixel pitch is central to LED. Pixel pitch sets the resolution and viewing distance.

Choosing the right pixel pitch matters: a fine pitch (small mm) for close viewing (a presentation screen, a broadcast backdrop seen on camera), a coarser pitch for far viewing (a big stadium screen). Pixel pitch is a key spec when selecting and using LED. Understanding pixel pitch frames LED work. The pitch is chosen for the viewing distance and detail needed.

Brightness and quality

LED's standout feature is brightness — LED screens are very bright (far brighter than projection), so they work in any lighting, including bright rooms and daylight, which is why they're used outdoors and in bright venues. Beyond brightness, LED image quality involves colour (good LED has vivid, accurate colour), contrast (LED can be very high-contrast, with deep blacks), and the processing/calibration that makes the wall look uniform and correct. Understanding LED's brightness and quality frames why and how it's used. LED is bright, vivid, and high-contrast, working in any light.

This brightness and quality (plus scalability) is why LED has become the go-to for impactful big-screen video. Good LED, well-configured and calibrated, looks stunning; the technology delivers bright, vivid images at scale. Understanding LED's brightness and quality frames its use. LED's brightness and image quality make it the dominant big-screen choice.

Viewing angle, and why the side seats see a different wall

A wall does not look the same from everywhere in the room, and the specification that tells you so is viewing angle. It is quoted as two numbers, horizontal and vertical, and like a loudspeaker's coverage it is the angle within which the picture still holds up rather than the angle within which it is visible.

What falls off is brightness first and colour second. Move round to the side of most panels and the image dims, and the white point drifts — often towards blue or green, because the red, green and blue emitters in each pixel are not identical packages and do not fade off-axis at the same rate. A wall that is a warm neutral white on axis can be a cool one from the cheap seats, and nobody in the room can tell you why the presenter's shirt changed colour.

This is a seating decision, and it is made long before the wall is hung. If the front row wraps round the stage, the people at the ends are looking at the wall from sixty or seventy degrees off, and no processing fixes what the panel physically does at that angle. The answers are the ones you would expect: choose panels with a wider quoted angle, angle the wall towards the audience rather than square to the room, or accept that the extreme seats get a compromised picture and price the job accordingly.

On camera the same effect appears differently. A camera at an angle to the wall photographs the off-axis colour, and grading cannot separate it from the content, because the wall itself is the light source. On any job where the wall will be on camera, the camera position is part of the wall design.

Moire, and why the camera sees what you do not

A wall can look perfect in the room and come back from the camera covered in soft rippling bands that are in none of the content. That is moire, and it is not a fault in the wall or in the camera. It is what happens when two regular grids overlap.

The wall is a grid of pixels. A camera sensor is a grid of photosites. When one grid is photographed through the other at certain distances and angles, the small difference between their spacings produces a large, slow pattern — the same effect that makes a striped shirt shimmer on television.

Because it comes from the relationship between two grids rather than from either one, the things that change it are geometric. Move the camera and it moves. Change the focal length and it changes. Change the distance between camera and wall and it can disappear entirely. A wall that was clean at the rehearsal can moire in the show because the camera operator found a better position.

The fixes are the ones that break the coincidence. Pull focus slightly off the wall, so the pixel grid is not sharply resolved — this is the commonest answer and it is why a wall used as scenery is rarely shot in critical focus. Change the angle or the distance. Use a finer pitch, so the pixel grid is too small for the sensor to beat against. On a fixed installation, this is a design decision made before purchase, because a pitch chosen for the eye may be wrong for the camera.

Refresh rate is a separate problem that looks similar and is not. Moire is a spatial pattern that sits still; refresh artefacts are horizontal bands that roll or flicker through the frame. If it moves, it is refresh — 1,920 hertz is adequate for the eye and a camera wants 3,840 or more. If it sits, it is moire, and no refresh setting touches it.

Pixel pitch and viewing distance

Pixel pitch is the distance between the centres of adjacent LEDs, measured in millimetres. It sets both how sharp the wall looks and how close a viewer can get before seeing dots instead of an image.

The rule, and its limits

• The working rule is that minimum comfortable viewing distance in metres is approximately equal to the pixel pitch in millimetres.
• For critical or close inspection, use 1.5 to 2 times that distance instead.
• For camera work, treat the camera as the nearest viewer. A camera pushing in on the wall sees it from centimetres away in effect.
• Resolution of the wall equals its width in millimetres divided by the pitch. A 6,000 mm wall at 3.9 mm is about 1,538 pixels wide.

Brightness, refresh, and camera

Brightness is measured in nits, or candela per square metre. Refresh rate is how many times per second the LEDs are cycled, and it is the single specification that decides whether a wall is usable on camera.

Why cameras see what your eye does not

• LEDs do not stay lit; they cycle on and off very fast, and your eye integrates that into steady light.
• A camera shutter samples a short slice of time, so if the refresh is too slow it catches the wall mid-cycle.
• The result is horizontal banding or scan lines rolling through the shot, which no colour correction removes.
• Higher refresh rate means more cycles inside the shutter interval, so the banding disappears.
• Shutter angle and frame rate also matter. If banding persists at high refresh, work with the camera department on shutter settings.

On the job

A product launch. The wall is 6 metres wide, the front row is at 2.5 metres, and there are three cameras including one that will push in tight on the wall during the reveal.

The front row at 2.5 metres suggests a pitch around 2.6 mm. But the camera pushing in changes the calculation entirely, because at the end of that move the camera is effectively a viewer at well under a metre. If you specify for the front row, the reveal shot, which is the single most important image of the event, will show a grid of dots.

So you specify for the camera: 1.9 mm or finer, and a refresh rate of at least 3,840 Hz. Then, on site, you run the wall at around 40 percent brightness so the camera does not clip it, and you shoot a test with the actual cameras at the actual shutter settings before anyone signs anything off.

The client asked for a six metre wall. What they actually bought was a decision about one camera move.

How to do it

Choosing pitch and setting brightness for a specific job.

1. Measure or confirm the distance to the nearest viewer, including any camera that will move close.
2. Take the pitch in millimetres as roughly equal to that distance in metres, then go finer if content includes small text.
3. If cameras are involved, treat the closest camera framing as the nearest viewer.
4. Specify refresh rate of at least 3,840 Hz for any camera work.
5. On site, set brightness for the room rather than leaving it at maximum. Indoors on camera this is usually 30 to 60 percent.
6. Match colour temperature to the lighting design, commonly 6,500 K.
7. Shoot a camera test at the actual shutter settings and look for banding before the show.
8. Record the settings that worked, because the next show in that venue starts from them.

What goes wrong

Practice it

Answer before checking.

Building & Feeding an LED Wall

A wall that is correctly specified can still be impossible to build. Weight, power, and data are the three constraints that decide whether the thing on the drawing can exist in the room, and all three are arithmetic you can do before the truck is loaded.

Assemble it, power it, and send it the image.

Assembling and rigging

An LED wall is built by assembling the panels into the screen — connecting the tiles together (mechanically and electrically) into the desired size and shape, and supporting the structure (rigged/flown from above, ground-stacked, or mounted, depending on the application). The panels must be assembled flat and aligned (so the screen surface is even and seamless) and safely supported (LED walls are heavy, and flown walls are a rigging/safety matter). Assembling and rigging the wall correctly and safely is the physical foundation of an LED setup. Understanding assembly and rigging is central to LED work. The wall is built by assembling and safely supporting the panels.

This is hands-on work (connecting panels, building and rigging the structure) that must be done correctly (flat, aligned, secure) and safely (especially for flown walls — overhead rigging). A well-built wall is even, seamless, and safe; a poorly built one is uneven or unsafe. Understanding assembly and rigging frames LED setup. The professional assembles and rigs the wall correctly and safely.

Power, data, and the processor

An LED wall needs power and data, driven by the processor. Power feeds the panels (LED walls draw significant power, distributed to the panels). Data carries the image from the processor to the panels (the processor sends each panel its part of the image over data cabling). The processor is configured to map the video correctly onto the wall (telling it the wall's layout — how the panels are arranged — so the image displays correctly across them). Connecting power and data and configuring the processor brings the wall to life. Understanding power, data, and the processor is central to LED. The processor, power, and data make the wall display the image.

Getting this right means the wall is properly powered (every panel lit), correctly fed with data (the image reaching every panel), and configured so the processor maps the video onto the wall's exact layout (so a panel in the top-left shows the top-left of the image, etc.). Understanding the power/data/processor setup frames LED work. The professional powers, feeds, and configures the wall so it shows the image correctly.

Cable management, and why it is not tidiness

The cabling on the back of a wall looks like housekeeping and is actually maintenance planning. Every panel takes a power jumper and a data jumper from its neighbour, and every one of those connectors is a place a show can stop. How they are dressed decides what happens when one panel fails at half past six.

The rule is that a cabinet must be removable without disturbing the ones around it. That means a service loop at each connection — enough slack that a panel can come forward on its hinge, or come out entirely, without dragging its neighbours' cables with it. A wall cabled tight looks immaculate and costs you twenty minutes and three more faults the first time anything goes wrong.

Loops are dressed to the frame rather than left hanging, because a hanging loop is a weight on a connector and connectors are the part that fails. On a flown wall it is also a dropped-object risk, which makes it a safety matter rather than a preference: anything above an audience is tied off, including cable.

Label both ends of every run. On a small wall you will remember; on seventy-two cabinets in a dark room with a show in an hour, nobody remembers. The convention that survives contact with a real load-in is to label by position — column and row — rather than by cabinet serial, because the position is what you are looking at when you need to find it.

Outdoor walls, and what makes a panel outdoor

An outdoor panel is not an indoor panel used outside. It is a different product, and the differences are the reason it costs more.

The first is brightness, and the book's figures show the scale of it: 800 to 1,500 nits indoors against 4,500 to 10,000 outdoors. That is not a preference, it is competition with daylight — a wall that is brilliant in a ballroom is a grey rectangle in a car park at two in the afternoon.

The second is ingress protection, quoted as an IP rating. Two digits: the first is solids, the second is water. Outdoor panels are commonly IP65 on the front and a lower rating at the back, because the front faces weather and the back faces a structure. Read both numbers rather than the word "outdoor" on a datasheet, and read them for the panel you are actually hiring.

The third is what nobody quotes: the whole system has to survive the same conditions. Processing, power distribution and data have to be housed, and the housing is part of the specification. A weatherproof wall fed from a rack under a plastic sheet is a weatherproof wall with a single point of failure in the rain.

And wind is a structural question rather than an electrical one. A wall is a sail. Above a certain wind speed a freestanding outdoor wall has to come down, and the number is set by the structure it is built on, not by the panels. That threshold belongs in the plan before the wall goes up, with somebody named who decides.

Weight, power, and what the room can take

These are representative figures for a common 500 by 500 mm indoor cabinet. Your product data will differ, sometimes substantially, but the method of calculating does not.

A worked example

• A wall 6 m wide by 3 m high in 500 mm cabinets is 12 across by 6 high, which is 72 cabinets.
• At 8 kg per cabinet the panels alone weigh about 576 kg, before frames, bumpers, and cabling.
• At 350 W maximum per cabinet, maximum draw is about 25.2 kW.
• Average draw is roughly a third of that, so about 8.4 kW for typical content.
• You size the supply and the breakers for 25.2 kW, and you plan generator load around the average.

Data, and how many ports a wall needs

A processor sends pixels down gigabit ethernet ports, and each port carries a finite number of pixels. Exceeding that limit is a common planning error that appears as part of the wall simply not lighting.

Working it out

• Total pixels equals wall width in pixels multiplied by wall height in pixels.
• Divide by the processor's stated pixels per port to get the minimum number of ports.
• Round up, then add redundancy if the show warrants it.
• Confirm the processor has that many ports before specifying it, not after.

On the job

A touring show carries a wall 10 metres wide by 5 metres high in 500 mm cabinets at 3.9 mm pitch. On paper it is straightforward. In the third venue, a third of the wall goes dark.

Work through it. Twenty cabinets across, ten high, so 200 cabinets. At 128 pixels per cabinet the wall is 2,560 by 1,280 pixels, which is 3,276,800 pixels. At roughly 650,000 pixels per port, that needs at least six data ports, and the processor in use has four.

It worked in the first two venues because a previous crew had quietly configured it to run at reduced colour depth, which increased the pixels each port could carry. Someone in venue three restored the intended colour settings, the ports ran out of capacity, and a third of the wall stopped receiving data.

The wall was never correctly specified. It had been running on a compromise nobody had documented, and the failure appeared three shows later in front of an audience. Do the pixel arithmetic before you specify the processor, and write down any compromise you make.

How to do it

Calculating a build before it is committed to.

1. Calculate cabinet count from wall dimensions divided by cabinet size.
2. Multiply by cabinet weight for total panel weight, then add frames, bumpers, and cable.
3. Have a qualified rigger confirm the venue can carry that load at those points.
4. Multiply cabinet count by maximum power draw, and size supply and breakers to that figure.
5. Calculate average draw at about a third of maximum for generator planning.
6. Calculate total pixels, divide by the processor's pixels per port, and round up for the port count.
7. Confirm the chosen processor has that many ports at the intended colour depth and frame rate.
8. Plan cable runs, keeping copper well within 100 metres and using fiber beyond.
9. Document every compromise made, because the next crew will not know.

What goes wrong

Practice it

Do the arithmetic before checking.

Setting Up & Using LED

The wall is hung, powered, and lit. Now it has to look right, and looking right is a set of deliberate adjustments rather than a default state. Calibration, mapping, and content format are where a technically working wall becomes a good-looking one.

Configure, check, and run the wall.

Configuring and checking

Once built and connected, the LED wall is configured and checked. Configuration means setting up the processor for the wall (its layout, resolution, the input signal) so the video maps correctly onto it, and setting brightness, colour, and calibration so the wall looks uniform and correct. Checking means verifying the whole wall works — every panel lit, the image mapped correctly, no dead panels or errors, the image looking good across the whole surface. Configuring and checking deliver a working, good-looking wall. Understanding configuration and checking is central to LED setup. The wall is configured and checked so the image looks right.

This applies signal and display knowledge to LED: feeding the right signal, configuring the processor and wall, and verifying the result. A well-configured, checked wall displays the content perfectly; an unchecked one may have mapping errors or dead panels. Understanding configuration and checking frames LED setup. The professional configures and checks the wall for a perfect image.

Using LED in production

In use, the LED wall displays the content for the show — whether that's a video feed, IMAG, presentation content, or creative visuals (often fed by a media server or switcher — see the creative-tech courses). The LED wall is the canvas for big-screen content, and using it well means feeding it good content correctly configured for the wall. LED is central to modern productions (concert and event screens, broadcast backdrops, immersive visuals), and understanding the basics is the entry to working with it and to the creative LED and media-server craft. Understanding LED in production frames its place. The LED wall is the big-screen canvas of modern productions.

This connects LED basics to the wider creative-tech world (media servers, LED creative, XR), where LED walls are the display for sophisticated content. The basics here — how LED works, builds, and is configured — are the foundation for all of it. Combining what LED is, pixel pitch and brightness, building and feeding, and configuring and using is LED video basics. The professional understands and can set up an LED wall. This is the foundation of modern big-screen video. LED is the canvas; this course is how it works.

Try this. Look at LED screens around you (a stadium screen, a stage backdrop, a digital billboard, a store display) and consider: it's made of panels, driven by a processor, very bright. Notice how big and bright LED can be — and that it's all built from tiles.

Dead pixels, and what maintenance actually looks like

The most visible failure an LED wall has is also the smallest. A single dead pixel on a black background is invisible; the same pixel on a white one is a hole the client sees from the back of the room and will mention every day of the run.

Pixels fail in three ways and they look different. One emitter of the three dies and the pixel shifts colour — a magenta dot where the green has gone. All three die and it is black. Or the driver behind a group fails and a whole block goes, which reads as a rectangle rather than a dot and is a different repair.

Finding them is a job for a test pattern rather than for content. Full white, full red, full green and full blue in turn, looked at from a few metres, will show you every fault the wall has, and it takes about a minute once the wall is up. Doing it at the end of the build, before the truck leaves, is the difference between a spare panel swapped quietly and a conversation with a client.

Repair is module-level on most modern panels: the small tile of LEDs comes off the front with a magnetic tool and a replacement goes on, without taking the cabinet out of the wall. That is why spares are quoted in modules as well as cabinets, and why a rental order that includes no spares is a rental order that assumes nothing will fail.

The number worth agreeing in advance is how many faults are acceptable. Manufacturers quote a dead-pixel allowance and it is never zero. Knowing the figure before the wall is built is how you avoid arguing about it afterwards.

Commissioning: proving the wall before anybody needs it

A wall is built long before it is trusted, and the hour between those two states is commissioning. It is a sequence, and the order matters because each step depends on the one before it being right.

Power first, and power alone. Bring the wall up with no signal and watch what happens: every cabinet should come alive, and the ones that do not are a power problem rather than a data one. Finding that now, while the room is empty, is the difference between a five-minute fix and a fault you are chasing under content with a client watching.

Then data, one port at a time if the processor lets you. A port that is over its pixel budget shows as part of the wall simply not lighting, and the failure looks identical to a dead cable. Knowing which ports carry which cabinets — because you wrote it down when you planned it — turns that from a mystery into a line on a page.

Then a full-field test: white, red, green, blue, in turn. This is where dead pixels, mismatched cabinets and a module seated badly all announce themselves, and it takes a minute. Then a moving image, because a still frame hides refresh and frame-rate faults completely.

Last, the thing people skip: look at it from where the audience will be. A wall commissioned from two metres away, on a ladder, has been judged from a position no paying person will ever occupy.

What "it looks right" actually means

Somebody will ask whether the wall looks right, and the honest answer needs more than an opinion. Four things are being judged, and they fail differently.

Uniformity is whether the wall is one surface or several. On a flat mid-grey field, cabinets that were calibrated at different times or came from different batches show as rectangles. It is the fault a client spots fastest because the human eye is very good at edges, and it is the one that cannot be fixed on site beyond swapping cabinets.

Colour is whether white is the white everybody agreed on. Typically 6,500 kelvin, and it has to agree with the lighting design and with what any camera is white-balanced to. A wall that is right on its own and wrong against the lighting rig is a decision nobody made rather than a fault.

Brightness is whether the level suits the room and the camera — and on camera that means turning it down, to thirty or sixty percent, not up.

Seams are whether the joins disappear. A physical step between cabinets throws a shadow line under stage light that no processing removes; that is the locks, and it is a build problem, not a picture one.

Judge all four deliberately, in that order, and "it looks right" stops being a feeling and becomes something two people can agree on.

The button on the back of the panel

Almost every cabinet has a small function button on its rear control card, and it is the most useful diagnostic tool on the job because it needs nothing else to work. Press it and the cabinet displays its own test pattern — usually a cycle of flat red, flat green, flat blue and white — generated by the receiving card inside the panel.

What makes it worth knowing is what it PROVES. That pattern comes from inside the cabinet, so it does not travel through your processor, your data cable, your mapping, or your content. If the panel shows a clean test field, everything from the receiving card outwards is healthy and your fault is upstream. If the panel shows the same fault on its own test pattern as it does on the show, the fault is in that panel.

That single distinction saves more time than any other habit in LED work. A dark cabinet during a show could be a dead panel, a dead data run, a port over its pixel budget, or a mapping mistake — four completely different repairs. Press the button and you have eliminated three of them in about five seconds.

The colour cycle earns its place too. A pixel with one dead emitter is invisible on white and obvious on the colour that emitter makes: a pixel with no green is a normal white pixel to a careless glance and a missing dot on a flat green field. Run all three colours and every partial failure declares itself.

Learn where the button is on the panels your rental house stocks, because it is never in the same place twice and it is often behind a cover you have to know about in advance.

When the picture is wrong, what it usually means

A wall that is on but wrong is a more common call than a wall that is dark, and the shape of the wrongness tells you where to look. Five patterns cover most of it.

THE WHOLE IMAGE IS STRETCHED OR SQUASHED, and faces are too wide or too narrow. The content is not the wall's native canvas. A wall of twelve cabinets by six at 160 pixels each is 1,920 by 960, and a file made at 1,920 by 1,080 has to be squashed, cropped or letterboxed to fit. Fix it in the content, not in the processor — scaling to the wrong aspect ratio is why it looks like that in the first place.

A COLUMN OR BLOCK OF CABINETS IS DARK while the rest is correct. That is a data problem: a failed cable, a break in the chain, or a port pushed past its pixel budget. Press the test button on one of the dark cabinets. If it lights, the panel is fine and the data never arrived.

A BLOCK SHOWS THE WRONG PART OF THE IMAGE — the top left corner appearing in the middle of the wall, or a section repeating. That is mapping. The processor's cabinet map does not match the physical build order, which happens whenever a wall is built in a different sequence from the one on the plan.

THE IMAGE TEARS HORIZONTALLY, with the top half offset from the bottom, or the whole picture judders. That is timing: frame rates that do not agree somewhere in the chain, or a missing genlock on a system that needs one.

THE COLOUR IS WRONG ON SOME CABINETS AND NOT OTHERS. That is calibration, and it is a patchwork most visible on a flat field rather than on content. Cabinets from different batches, or calibrated at different times, do not match; the answer is recalibration or swapping cabinets, and neither happens on show day.

Work in that order — content, data, mapping, timing, calibration — because it runs from the cheapest thing to check to the most expensive thing to fix.

Setup values and what they do

These are the controls you will actually touch on site, and what each one is for.

The setup order that avoids rework

• Confirm every cabinet is lit and communicating before adjusting anything.
• Load the correct calibration data for the panel batches actually in the wall.
• Set the cabinet map to match the physical build, and verify with a test pattern that numbers the cabinets.
• Set refresh rate before shooting any camera test.
• Set brightness and colour temperature under show lighting, not work lights.
• Check uniformity across the whole wall with a flat grey field.
• Only then load show content and check framing and scaling.

On the job

A wall is built and lit and the client walks in during setup. The wall is showing content, and it looks patchy: one column noticeably pinker than the rest, and a rectangle in the lower left slightly brighter.

This is not a fault, it is an unfinished setup, but the client does not know that and first impressions are difficult to undo. The pink column is a batch of panels from a different production run with the wrong calibration data loaded. The brighter rectangle is a block of cabinets that did not receive the brightness change because their receiving cards were not communicating when it was sent.

Both are ten-minute fixes. Load the correct batch calibration, confirm every receiving card is communicating, and re-send the settings to the whole wall.

The lesson is about sequence rather than technique: do not put show content on a wall until the wall itself is uniform. Show a flat grey field, fix what you see, and only then let anybody look at it.

How to do it

Commissioning a wall on site.

1. Power up and confirm every cabinet is lit and every receiving card is communicating.
2. Display a numbered test pattern and verify the cabinet map against the physical build.
3. Load the calibration data matching the actual panel batches in this wall.
4. Display a flat white and a flat grey field and look for colour and brightness variation.
5. Correct any variation before going further; re-send settings to the entire wall, not just the affected area.
6. Set refresh rate, then brightness, then colour temperature, then gamma, under show lighting.
7. Shoot a camera test and check for banding and clipping.
8. Load show content and confirm it is scaled to the wall's exact canvas.
9. Record every final setting.

What goes wrong

Practice it

Answer before checking.

LED as the Modern Canvas

LED stopped being a screen some years ago. It is now scenery, lighting, and in virtual production the environment itself. That shift changes who you work with and what you are responsible for far more than it changes the technology.

The foundation of big-screen and creative video.

LED basics mastered

Pulling it together, LED video basics is understanding how LED screens work — direct-view panels assembled into bright, scalable walls, driven by a processor, with pixel pitch and brightness as key specs, built, powered, fed, configured, and checked to display the image. Understanding LED basics makes you able to work with the dominant big-screen technology and is the foundation for the creative LED and media-server world. LED is the canvas of modern big-screen and creative video, and this course has explained how it works. LED basics mastered is understanding the screens that modern shows are built on.

LED is everywhere and growing (it has largely taken over big-screen video and is central to creative and immersive work), so understanding it is increasingly essential AV knowledge. The basics here open the door to LED setup work and to the creative-tech specialties (LED creative, media servers, XR) that use LED as their canvas. Understanding LED basics is foundational modern video knowledge. This course has built that foundation. LED is the modern big-screen canvas, and you now understand it.

Developing LED skills

Developing LED skills means getting hands-on building, feeding, configuring, and running real LED walls (learning the panels, processors, rigging, and setup), deepening your understanding of the technology and signal, and building experience across applications (events, broadcast, creative). From here you can specialize — LED engineering/crew chief (building and running walls), LED creative/content (programming visuals for LED), or media servers (driving LED content). The best LED professionals know the technology deeply, build and configure walls expertly, and deliver flawless screens, built through hands-on work. Deliberately developing these is the path forward.

From here, growing means working with bigger and more complex LED systems, specializing (LED crew chief, creative, media server), and building toward senior LED and creative-tech roles. LED is a major, growing, in-demand field at the heart of modern visual production. The basics this course provides — how LED works and is set up — are the foundation; growth comes from hands-on LED work and specialization. Keep working with LED, keep building your skills, and you grow into an LED professional in the modern visual world.

Try this. Consider the path of an image onto an LED wall: a source feeds a processor, the processor maps and sends the image to the panels over data, the panels light up. Picture how the wall's layout has to be told to the processor so the image lands correctly.

When the wall is the set

A wall used as a display shows content to an audience who know they are looking at a screen. A wall used as scenery is trying to be a place, and everything about the job changes when that is the intention.

The first change is who decides the content. A display shows what the client brought. A set shows what a production designer and a lighting designer and the video department agreed on, because the wall is now part of the room's light. A bright blue scene throws blue on the performers standing in front of it, and that is either deliberate or it is a problem discovered in a technical rehearsal.

The second is brightness, and it runs the opposite way from instinct. A display wants to be bright. A set wants to sit at the level the rest of the stage lighting is working at, which is usually far lower — thirty to sixty percent of maximum on camera, and sometimes less to the eye. A wall run at full behind a presenter turns them into a silhouette, and no amount of front light rescues it without washing out the wall.

The third is that there is no fallback. A display that fails leaves a dark rectangle and an awkward moment. A set that fails leaves a hole where the world was, in front of an audience who were being asked to believe in it. That is why redundant processing and redundant data paths stop being an upsell on this kind of job and start being the specification.

Working with the people who make the content

The single most common cause of a wall looking wrong is content built for the wrong canvas, and it is preventable with one conversation held early enough.

The wall has a native resolution — the exact pixel count of the grid you are building — and it is almost never a standard one. A wall twelve cabinets wide by six high, in cabinets that are a hundred and sixty pixels square, is 1,920 by 960. That is not a broadcast format, and content delivered as 1,920 by 1,080 will be either squashed, cropped or letterboxed, whichever the processor is set to do. Every one of those looks like a mistake to the client, and all three are avoidable by sending the number before the content is made.

Send it in writing, and send it as pixels rather than as metres. The designer does not care that the wall is six metres wide; they care that the canvas is 1,920 by 960, and that anything important should stay away from the outer few pixels because the edge of a wall is where masking and bezels live.

Ask what frame rate the content is being made at, and make it agree with the system. A file at thirty frames going into a system running at fifty judders in a way that looks like a fault in the wall and is a fault in the plan.

And ask for a still. One frame of the real content, at the real canvas size, viewed on the real wall during a build, catches more problems than any amount of specification — because it is the first moment anybody sees what the audience will see.

Where LED work goes next

The skills in this book are the ones every LED job needs, and they open onto three directions that are genuinely different from each other.

The first is scale. Arena and festival work is the same arithmetic with bigger numbers and less margin: more universes of data, more power, structures that need engineers rather than opinions, and a load-in measured in days. Nothing conceptual changes. What changes is that every mistake is expensive and most of them are made on paper weeks earlier.

The second is broadcast and virtual production, where the wall stops being something an audience looks at and becomes something a camera photographs. Refresh rate stops being a specification and becomes the whole job, genlock arrives, and the wall has to agree with the camera's shutter as well as with the lighting. This is the fastest-growing use of LED and the one that pays best, and it demands the camera knowledge as much as the wall knowledge.

The third is installation rather than events. A wall that goes up once and stays up is a different discipline: it is surveyed, it is commissioned to a standard, it is maintained on a schedule, and somebody signs for it. Less adrenaline, more paperwork, and a career that does not depend on being available at weekends.

None of the three is a promotion from the others. They are different rooms, and the reason to name them is so you can choose rather than drift.

What changes when the wall becomes the set

Once a wall is scenery rather than a display, decisions that used to belong to video start belonging to several departments at once.

Virtual production, briefly

• In virtual production the wall displays an environment that the camera photographs as if it were real.
• The image is rendered in real time and shifts with the camera's tracked position, so perspective stays correct.
• This demands very high refresh rates, fine pitch, accurate colour, and tight synchronisation with the camera.
• It also demands genlock, so the wall, the render, and the cameras all agree on timing.
• It is a specialism built on the fundamentals in this book, not a replacement for them.

On the job

A concert design where the LED is the set: a curved upstage wall, floor tiles, and columns. There is no scenic backdrop behind it and no plan B.

The technical work is the same as any wall, but the responsibility is different. If a projector fails you lose an image. If this wall fails you lose the set, in every wide shot, on the stream, for the rest of the show.

So the decisions change. Redundant processing, because a single processor failure is now a show-stopper. Redundant data paths, so one cable does not dark a column. Spare cabinets on site and a plan for swapping one during a song. Power from more than one source where possible. And a lighting design that has been built knowing what the wall contributes, because at these sizes the wall is one of the largest light sources in the room and the lighting designer is depending on it.

The craft is the same. The stakes and the number of people affected by your failure are not.

How to do it

Working on a show where the wall is the set.

1. Establish early that there is no scenic fallback, and design for redundancy accordingly.
2. Specify redundant processing and redundant data paths for anything the show cannot lose.
3. Carry spare cabinets and a rehearsed procedure for swapping one quickly.
4. Agree with lighting how much light the wall contributes, and at what level it will run.
5. Agree with the director what happens if a section fails mid-show, and have a content state ready for it.
6. Test the failure modes in rehearsal: pull a data cable and see what actually goes dark.
7. Document the whole system so somebody who did not build it can fix it at speed.

What goes wrong

Practice it

Answer before checking.

Glossary

Quick Reference

Every number here is worked through somewhere in the book. They are gathered on one page because a load-in is not the moment to go looking for them.

The numbers you will actually reach for

• Minimum comfortable viewing distance in metres is roughly the pixel pitch in millimetres. A 3.9 mm wall wants about 4 metres.
• Wall resolution equals wall width in millimetres divided by pixel pitch.
• A 500 by 500 mm panel at 3.9 mm pitch is 128 by 128 pixels.
• Indoor walls run about 800 to 1,500 nits. Outdoor walls need 4,500 to 10,000.
• For camera, you want at least 3,840 Hz refresh, and higher is better.
• Average power draw is roughly a third of the panel's stated maximum, but breakers must be sized for maximum.
• Plan on the order of 650,000 pixels per gigabit data port, and confirm against your processor.
• Cat5e and Cat6 runs are limited to 100 metres. Use fiber beyond that.

About Reboot AV

Reboot AV exists because the audiovisual industry has a map problem. The work is there and the ceiling is high, but almost nobody entering it can see the route from where they are to where they want to be.

What the gear actually looks like

Every device named in this book, in one place. Recognition is the whole point of Chapter 4 — use this page to attach a picture to each name before you meet it on a dock at six in the morning.

Audio

Dynamic microphone

Condenser microphone

Wireless handheld

Radio rack

Mixing console

Stage box / snake

Power amplifier

Main loudspeaker

Subwoofer

Stage monitor wedge

In-ear monitors

DI box

Video

Camera

Video switcher

Scaler / processor

Projector

LED wall panel

Confidence monitor

Lighting & control

Conventional fixture

Moving-head fixture

Lighting console

Dimmer

Show controller

Media server

Networking & infrastructure

Network switch

Patch bay

Equipment rack

Power distribution

Cable tester

A properly coiled cable

Connectors, close up

XLR

TRS / TS ¼-inch

SDI on BNC

HDMI

Cat5e / Cat6 (RJ45)

Fiber optic

DMX (5-pin XLR)

AC power — not signal