Feed one 3840 x 2160 canvas from a single player into a 4K video wall controller set to a 2x2 split with the fourth quadrant black. Each of the three screens then shows its own 1920 x 1080 panel at native resolution, with no rescaling and nothing to keep in sync.
Three screens that read as one wide menu need one picture, not three players trying to keep step. Play a single 3840 x 2160 canvas from one player into a 4K video wall controller. Set the controller to a 2x2 split and leave the fourth quadrant black. Three outputs then carry a native 1920 x 1080 picture to the three screens, with no rescaling, no network and no sync to drift. This page works through why the 2x2 setting beats the obvious 3x1 one, what the canvas should look like, what happens when the power comes back, and the four spec lines to check before you buy. The arithmetic fits on a napkin, and doing it first is what keeps the letters on the menu sharp.
How do you put one menu across three TV screens?
You play one 3840 x 2160 canvas from a single player into a 4K video wall controller set to a 2x2 split. Leave the fourth quadrant black and wire three of its outputs to the three screens. Each screen then shows a native 1920 x 1080 picture.
The controller is the piece people miss in their first search. A plain HDMI splitter repeats the same whole picture on every screen, so all three TVs would show the entire menu. A wall controller takes one picture in and hands each output its own piece, which is the difference set out in how an HDMI splitter differs from a video wall processor.
The three screens are ordinary TVs with HDMI inputs. The player is any box that plays a full screen slideshow or a video loop and starts itself after power-up. It never knows its picture is being cut up, so nothing about the player drives this purchase. Because the menu lives as one picture, an update is one export to one device.
If you are still choosing the screens themselves, the basics of size, viewing distance and brightness are in the menu board screens guide; this page starts where that one stops, at driving three screens from one player.
What are the options for driving three screens from one player?
Four setups cover it: a controller on a 2x2 split with one quadrant black, the same controller on a 3x1 split, screens that carry a built-in wall or tile mode, or one player per screen kept in step over wired LAN.
| Option | What each screen gets | Where it wins |
|---|---|---|
| Controller, 2x2 split, fourth quadrant black | A native 1920 x 1080 quadrant, pixel for pixel | A menu that changes a few times a year |
| Controller, 3x1 split | A 1280 x 2160 slice stretched to fill the screen | Almost nothing; the picture is rescaled and distorted |
| Screens with a built-in wall or tile mode | Whatever the screens' own menus allow | New commercial displays that already have the mode |
| One player per screen, synced over wired LAN | A native 1920 x 1080 picture from each player | Screens that show different content or change on their own days |
The third row can be free: before buying anything, open each screen's menu and look for a wall, tile or daisy chain setting, because some commercial displays carry the splitting job built in. Consumer sets usually leave it out.
The fourth row is a separate project with its own wiring; syncing three menu board screens side by side covers the settings that keep separate players in step. For a menu that changes a few times a year, build the first row.
Why does a 2x2 split look sharper than a 3x1 split?
A 2x2 split cuts the 3840 x 2160 canvas into four 1920 x 1080 blocks that map onto 1080p screens with no scaling. A 3x1 split hands each screen a 1280 x 2160 slice stretched 1.5 times across and squashed to half height.
The arithmetic, for a 3x1 split of the 16:9 canvas: 3840 ÷ 3 = 1280 columns per slice, at the full 2160 rows. Filling a 1920 x 1080 screen means stretching 1280 columns to 1920, a 1.5 times stretch, and squeezing 2160 rows to 1080, a 0.5 times squash. Letters end up 1.5 ÷ 0.5 = 3 times wider than they were drawn. Keep the shape instead and the slice fits the height at 1280 x (1080 ÷ 2160) = 640 columns, with black filling the other two thirds of the screen. So a 3x1 split on this canvas gives you distortion or wasted thirds, and nothing in between.
A 2x2 quadrant needs none of that: on a 1080p screen the quadrant and the panel are the same 1920 x 1080, pixel for pixel. One input not carrying enough pixels for every panel is also why a six screen wall drops to HD per screen, and the same logic makes a 2x2 video wall processor worth having whenever the wall is four panels.
What does a worked example look like with real numbers?
Take a cafe with three 1920 x 1080 screens in a row over the counter: the canvas is 3840 x 2160, the drinks panel sits at top left, mains at top right, desserts at bottom left, and the bottom right quadrant stays black.
Wiring: controller output 1 to the left screen, output 2 to the middle, output 3 to the right, output 4 left unplugged.
On the 2x2 build, each screen gets a 1920 x 1080 quadrant and shows it 1:1, so a capital drawn 100 pixels tall on the canvas arrives 100 pixels tall. On a 3x1 split of the same canvas, each screen gets a 1280 x 2160 slice. The same capitals arrive at 100 x 0.5 = 50 pixels tall and 100 x 1.5 = 150 pixels wide, with the slice's 1280 columns of detail spread across 1920 columns of screen, which softens every edge.
| Setting | Piece each screen gets | Scaling to 1920 x 1080 | 100 pixel capitals arrive as |
|---|---|---|---|
| 2x2, quadrant | 1920 x 1080 | none, 1:1 | 100 pixels tall, sharp |
| 3x1, slice | 1280 x 2160 | 1.5x across, 0.5x down | 150 pixels wide, 50 pixels tall |
Same menu, same player: the 2x2 build keeps every letter at the size it was drawn, and the 3x1 build halves its height, widens it half again and blurs it. Build the 2x2.
How should the 4K canvas be laid out for three menu panels?
Divide the 3840 x 2160 canvas into four 1920 x 1080 quadrants, give each screen one self-contained panel, keep every price inside its own quadrant, and set each TV to its full-picture mode so nothing at the edges gets cropped.
Self-contained panels do the cosmetic work: each screen shows a finished panel with its own heading and prices, the bezel gap reads as the gap between panels, and nothing important sits where two screens meet. Consumer TVs trim the edges of an incoming picture by default, and the full-picture setting that turns that off lives in each TV's picture menu.
Keep the canvas at 3840 x 2160 and no larger. That is the most a 4K controller accepts on its input, and it is the same 3840 x 2160 cap and 4K canvas fix a signage player works under. For the drawing work, laying out one menu board canvas across four screens walks the same quadrant grid panel by panel.
Check the build once with a test frame: black background, a thin white outline around each quadrant. Every screen must show exactly one box, edge to edge. If a screen shows the whole menu shrunk into one corner instead, the show was sized for a single screen, and fixing a signage screen that shows slides in one corner is the repair.
What happens when the power goes out?
Two boxes have to come back on their own. Buy a controller whose spec sheet lists power-off memory, so it is designed to return to the 2x2 split at power-up with no menu pressed. Set the player to launch its show at boot, so after a power cut it is designed to come back playing the first slide with nobody touching it. Then prove both with a ten second power cut.
Controllers differ here, so ask the seller one direct question: after the building's power returns, does the unit come back in the same split with nobody touching it? The player side is a setting in its menu: start the show at power-up and loop it. On the screens, set each TV to switch on to its HDMI input.
Then run the test before opening day. Pull the power strip for ten seconds, plug it back in, and time it: picture back inside a minute or two, correct 2x2 split, first slide playing, nobody holding a remote. Run the player, controller and all three screens from one surge-protected strip so they lose and regain power together.
A controller with no memory of its settings needs its mode picked again after each cut, and that failure waits for opening time.
When is one player per screen the better choice?
Give each screen its own player when the screens show different things or change on different days. A shared canvas pushes every edit onto all three screens at once, and a sync group over wired LAN is what keeps separate players in step.
Sync has one hard rule: the players talk over wired LAN. Wi-Fi latency makes them drift apart, so the wireless route fails no matter which players you buy. When drift shows up on a wired build anyway, why sync group screens play video out of sync lists the causes in the order to check them.
There is an editing cost on both sides. A shared canvas turns a one panel price change into a full rebuild and a re-export. Separate players swap one screen's slides and leave the rest alone, which is the pattern four retail signage screens showing different content is built around. So: a menu that changes as one document a few times a year belongs on the controller, and screens with lives of their own belong on players. If a second location ever needs just one sign, card run digital menu boards are the simplest build in the family.
What do you check on the spec sheet before buying?
Check four lines before you pay: HDMI input rated for 3840 x 2160 at 60 Hz, a mode list that includes 2x2, power-off memory, and four outputs, because the spare output is where a fourth screen plugs in later.
The splicing modes on these spec sheets are printed in a row, typically 1x2, 1x3, 1x4, 2x1, 3x1, 4x1 and 2x2, with larger grids when units are chained. If 2x2 is missing from the printed list, the 3x1 route and its rescaling are what is left, so walk away. Two questions close the rest of the gap: does the unit come back in its split after a power cut, and do all four outputs carry 1920 x 1080?
One note about the screens themselves. This build needs displays that accept HDMI in. Our own panel-mount signage displays are wired the other way round: the HDMI port is an output only, so the panel-mount range cannot be the three screens behind a controller. Its job is a standalone sign running its own slideshow from a card, with no subscription.
Recommendation: For a restaurant menu that changes a few times a year, buy a four output 4K controller with 2x2 in its printed mode list and power-off memory on the spec sheet. Feed it one 3840 x 2160 canvas from a single player set to launch at boot. Put three 1920 x 1080 panels in the top left, top right and bottom left quadrants with the fourth black, set each TV to its full-picture mode, and keep one player per screen over wired LAN in reserve for the day a screen needs content of its own.
Questions
Can I add a fourth screen later?
Yes, and it is the simplest expansion in this build. The controller already makes four 1920 x 1080 outputs, so a fourth screen plugs into the spare output. Draw the fourth menu panel into the black bottom right quadrant, rebuild the canvas, copy it to the player, and the wall grows with no new mode and no new device.
Do the three screens have to be the same size?
No. Every output carries the same 1920 x 1080 picture, so a 32 inch screen and a 43 inch screen both show their full quadrant, and the bigger screen simply draws the same panel bigger. Keep text sizes consistent in canvas pixels and hang the panels at the same height, or buy matching screens for a level look across the wall.
What if the screens have to hang portrait?
Portrait breaks the 2x2 mapping. A portrait screen needs a 1080 x 1920 picture, while a 2x2 split produces 1920 x 1080 quadrants, and three portrait screens in a row want 3240 x 1920 of picture, which one 3840 x 2160 input cannot slice cleanly. Keep the screens landscape, or move to one player per screen over wired LAN.
Planning information, not legal or engineering advice. Rules differ by place: check codes and permits with your local authority and your licensed trades.