Three side-by-side menu boards stay in step two ways: as one canvas split across the screens, so every change lands on the same frame, or as three playlists of exactly equal loop length started on a shared signal. Pick the method before the hardware, and manage drift with the frame math in this article.
Three 1920 x 1080 menu boards hung side by side make one 5760 x 1080 canvas, and the whole job is keeping their changes on the same frame, within one frame at 30 fps, or 0.033 s. Two setups do it. Run all three as one canvas split across the screens, and every change lands on the same frame. Run three playlists whose loop lengths match to the frame and start them on a shared signal, and the boards change together for as long as the loop math holds. The hardware matters less than the method, so ask any player you are evaluating which of the two it supports, in those words. This page gives the decision rules, the drift arithmetic, and the questions to put on the quote.
Pick the method before the hardware
The menu design picks the method. Buying screens and players first leaves you asking what the hardware can do; the order that works is the reverse.
- One canvas when the wall carries one continuous design: a menu strip flowing across all three boards, or one full-bleed background spanning the wall. One source, one timeline, no drift.
- A sync group when each board carries its own section and rotates its own promotions. Each board gets a player, every playlist matches the others in total duration, and a shared start signal ties the loops together.
- No sync when the menu is stable. Two boards hold static sections and one board rotates the promos. Boards that never change cannot fall out of step.
For timing, one frame at 30 fps is 0.033 s. A canvas lands every change inside one frame. A sync group lands inside whatever its start method allows, and the player's documentation states that number; ask for it.
One canvas split across three screens
Build the artwork once at the combined resolution. Three landscape panels at 1920 x 1080 side by side make one 5760 x 1080 canvas, and each board shows its 1920-pixel slice.
Three ways to drive it: one player feeding a video wall processor that cuts the picture across the screens, displays with a daisy-chain input that pass the picture along the row, or a cloud player with a canvas layout that assigns one screen per region. When a processor earns its place and when it does not is worked through in the 2x2 video wall processor article.
Everything changes on the same frame because there is one timeline. Fades and wipes are edits on that timeline, so their timing is exact. The cost is the seam: draw guides at 1920 px and 3840 px and hold every line of text inside one third, so nothing crosses a bezel. You also give up per-board playlists; the wall is one picture cut in three. The layout mechanics are worked out for a four-screen wall in laying out one menu board canvas across four screens, and the grid math carries over to three boards.
Three playlists of equal length, started together
Each board plays its own playlist, and two rules do all the work.
Rule 1: equal total duration, to the frame. Pick a loop length, say 120 s, which is 3,600 frames at 30 fps. Every board's loop must render to exactly 3,600 frames. The reliable way is to render each board's loop as one video file with the transitions baked in, then read the file length off the render. Player-applied transitions can change the total: three 0.5 s wipes appended to slide time run a four-slide loop 1.5 s longer than the same loop with straight cuts. A one-frame difference means that board finishes every loop 0.033 s late.
Rule 2: one shared start. The boards need a common start: a sync-group or synced-playback feature that sends a start over the network, or a scheduled start where all three begin their loop at the same clock time. Whether the group re-aligns at every loop boundary or only at the first start is stated in the player's documentation. Get that sentence before you buy, because the drift math below is what happens when the group locks once and lets go.
When a group still staggers, the causes are worked through in why sync group screens play video out of sync.
The cheapest option: two static boards and one rotating
A menu that changes a few times a year does not need three timed players. Hang two boards with static sections, drinks and sides for instance, built as single slides, and let the third board run a rotating playlist for promotions. Static slides never drift, and the promo board only has to keep time with itself. Updates stay simple: the static boards change when prices change, and the promo board changes on its own schedule. This is the setup to start with when the menu is settled, and it upgrades cleanly: a player with a sync-group feature takes the existing slides into a group with no redesign.
Scenario: three 55-inch boards over a counter
The wall. Three 55-inch 16:9 landscape panels sit side by side above the order counter. Each panel measures 47.9 in wide by 27.0 in high, a 55 in diagonal at 16:9, so the active area runs 143.8 in, about 12 ft of board, and the counter sits 12 ft from it.
Text size. The fast-reading figure is 20 ft of viewing distance per inch of capital height, and the FHWA planning figure is 30 ft per inch. At 12 ft, that sets 0.6 in capitals for fast reading and 0.4 in as the floor. Line spacing is 1.7 times capital height, about 1.0 in here. A 1920 x 1080 panel puts 40 pixels in an inch, 1080 pixels over 27.0 in, so 0.6 in capitals are 24 pixels tall and the 0.4 in floor is 16 pixels. Menu item names clear it easily; the small print, a calorie line for instance, is what to watch, and it stays at or above 0.4 in.
Canvas math. One canvas across the wall is 5760 x 1080, and each section gets 1920 px of width, enough for a photo column, item names and prices.
Drift math. The target loop is 120.0 s, which is 3,600 frames at 30 fps. The drinks board's promo video renders 3,604 frames, 120.1 s. That board restarts on its own at each loop end, so it finishes every loop 0.13 s late. After 8 loops, 16 minutes, it is about 1 s behind. Over a 12-hour day at 120 s loops, 360 loops, it ends the day 48 s behind, and anyone in line sees three walls changing at three different moments.
Result. Trim the promo render to 3,600 frames, or put the three boards in a sync group that re-aligns each loop, or schedule a group restart at opening. With the frame counts equal, the boards change on the same frame from open to close, and the 0.6 in capitals read from 12 ft.
Rigging note: three panels this size are a two-person lift, and each mount point lands on blocking rated for the mount's load.
Hardware options compared
| Setup | Sync method | How tight | Per-board content | Remote updates | Pick it when |
|---|---|---|---|---|---|
| One player plus a video wall processor | One canvas cut to the screens | Same frame | No, slices of one picture | Yes, with a cloud player | One continuous design |
| Displays with built-in players and a sync mode | Group start, per the model's sync feature | Whatever the display's sync spec states | Yes, group playlists or a split canvas | Yes | Few boxes, short signal runs |
| Separate players with a sync-group feature | Start signal over the network | Locks at group start; re-lock per the player's documentation | Yes | Yes | Each board rotates its own promos |
| Three independent players, matched loops, restart at open | Scheduled start, no signal between boards | Drift grows between restarts at the loop-mismatch rate | Yes | Yes | No sync feature available, simple content |
What to ask for before you buy
Put these six on the quote, in writing:
- The sync method by name, one canvas or a sync group, taken from the player's own documentation, not a sales page.
- Re-lock behavior: does the group re-align at every loop boundary, or only at the first start?
- Canvas support at 5760 x 1080 for three 1080p boards, if you go the canvas route.
- Transition handling: whether a baked-in video keeps loop length exact, and whether player-applied fades or wipes extend it.
- The update path: cloud pushes over the internet, or a card swap on site. If you will not be on site to update, you need the network path; card-based offline boards update by swapping the card.
- A scheduled group restart, which resets any accumulated drift every morning.
We sell wall, counter and hanging displays, including larger wall models through 75 inches, quoted per project; the quote states the software, the sync behavior, the ports and the power, because those decide a three-board build. Our panel-mount series tops out at 23.8 inches, plays images and video from a card offline with no subscription, and suits a single board updated by card swap. Browse the display range, and for sizing a screen against the menu start from the digital menu boards guide.
Recommendation: run the three boards as one 5760 x 1080 canvas when the design is a single continuous menu strip and you want every change on the same frame. Use a sync group of three equal-length loops that re-aligns at each boundary when each board carries its own section and rotates its own promos. Keep two boards static and rotate only the promo board when the menu is settled, which removes sync from the job entirely.
Questions
Can one player run all three menu boards?
Yes. One player outputs the whole canvas, and a video wall processor or the displays' daisy-chain inputs split that one picture across the three screens. All boards show slices of one timeline, so every change lands on the same frame, and there is one content edit to make. Check the split path carries 5760 x 1080 before you plan the cabling.
How close do the boards have to be for changes to look simultaneous?
On one canvas, changes land on the same frame. For three playlists, use the drift rule: one frame of mismatch per loop at 30 fps is 0.033 s, which grows to 1 s after 30 loops, one hour at 120 s loops. Once the gap reaches a few tenths of a second, anyone watching sees one board change first. Equal frame counts or a group that re-aligns each loop holds the gap at the start offset.
Do the three boards need to be the same size?
For one canvas, yes. The picture is cut on pixel columns, so mixed sizes or resolutions break the grid; match resolution and alignment, and hold text inside each third so nothing crosses the bezel lines. For three playlists, sizes can differ. Keep every loop's total duration equal and start them on the same signal, and boards of different sizes still change together.
Planning information, not legal or engineering advice. Rules differ by place: check codes and permits with your local authority and your licensed trades.