LED Space Editor

The LED Space Editor is a visual tool for building an LED wall and mapping content onto it. You arrange the physical panels freely on a canvas — drag them, resize them, build irregular shapes — and the editor generates two things in one step:

• Output regions — one per panel, all assigned to the screen you choose, so the screen's pixels are spread across the physical wall.
• Fields — content regions on that screen's canvas that match the layout, so the content you play lands exactly on the panels (a field is a mapping from content to a region of the screen).

Opening the editor

In the Outputs section, open the Displays list header and click the grid icon (next to +). The LED Space Editor dialog opens.

Building the wall

You can start from a regular grid and then rearrange it freely:

1. Under Start from a grid, set the panel width/height, columns, rows and optional gap, then click Build grid. This seeds the canvas with a regular tiled wall.
2. On the canvas, drag any panel to reposition it, drag its corner handle to resize, or select it and use the arrow keys to nudge (hold Shift for larger steps, Delete to remove, Esc to deselect). You can also edit its exact X / Y / Width / Height in pixels. Use + Panel to add tiles, Duplicate to copy the selection, and Remove/Clear to delete.
3. Toggle Snap to align panels to an 8 px grid. Use the Preview selector to choose what's drawn across the panels:
   • Live output — the selected screen's real content (the timeline / composition playing now), reconstructed across the panels so you see exactly how content lands. A sample gradient stands in until a frame arrives (e.g. offline). The live frame freezes while you drag or resize a panel — you're changing the layout, not the content — and resumes on release.
   • Alignment grid, Colour bars, Checkerboard — built-in test patterns for verifying panel layout, orientation and image continuity across the wall without playing anything.
   • Gradient — a static sample image; Off hides the content layer.

Irregular layouts — L-shapes, columns, scattered tiles, mixed panel sizes — are all supported. The editor continuously shows the panel count, the content canvas resolution (the bounding box of all panels), and the field count. Overlapping panels are outlined in red and flagged, since two tiles sharing the same desktop space is usually a mistake.

Positioning on the desktop

For the wall to drive the right pixels, its panels must sit where the physical outputs actually are on the desktop. Under Desktop position:

• Set Origin X / Y to place the wall's top-left at a specific desktop pixel (may be negative for outputs at negative coordinates). This moves the whole wall without changing its internal layout.
• Place on output snaps the wall's top-left onto one of the machine's detected display outputs in one click.
• Turn on Desktop in the canvas toolbar to show the machine's physical displays as a backdrop, so you can position the wall against them visually.

The generated fields are canvas-relative, so the content mapping is identical wherever the wall sits on the desktop — only the physical placement changes.

Editing an existing wall

When walls already exist on the selected machine, an Edit existing wall picker appears at the top. Choose a wall to load its panels, name and screen back into the editor and tweak it — re-applying replaces that wall in place (including after a rename) instead of creating a duplicate. Choose + New wall to start fresh.

Settings

Setting	Description
Name	Label for the wall. Re-running the editor with the same name replaces that wall's panels and its generated fields, and leaves all other outputs untouched.
Machine	(multi-client only) The engine the wall is created on — the local machine or a reachable client. Pick the machine the panels are wired to. The screen list updates to that machine's screens.
Content source (screen)	The screen (on the selected machine) whose content is spread across the wall. Every panel is assigned to it.
Wall field	Generate one field covering the whole wall, so a composition bound to it fills every panel.
Per-panel fields	Generate one field per panel, so each tile can play its own content.
Set screen render resolution	Sets the chosen screen's render resolution to the computed content canvas so the source maps 1:1 onto the wall.

Fields — content mapping

A field is a rectangular region of the screen's content canvas that a composition (Playlist or Timeline) plays into. The editor normalises each panel's position over the content canvas and creates matching fields:

• The wall field spans the whole canvas. Bind a composition to it and the picture covers the entire wall.
• A per-panel field covers exactly one tile. Bind a composition to it and that clip shows only on that physical panel.

Generated fields are tagged with the wall's name, so re-applying the same wall replaces them cleanly instead of piling up duplicates. After applying, the fields appear on the screen in the Screens & Fields panel, where you can fine-tune, warp or assign content to them like any other field.

Applying

Click the apply button — labelled Create N panels + M fields for a new wall, or Update wall when editing an existing one. The editor writes the panels into the project's output list as region outputs, creates the selected fields on the target screen, and (optionally) sets the screen resolution. A confirmation shows how many panels and fields were created. Because the panels are region outputs, the wall survives monitor re-detection and restarts (and later output edits preserve them). To change the wall later, open the editor again and pick it from Edit existing wall.

Multi-client

When multi-client is enabled, the mapper shows a Machine selector at the top listing the local machine and every reachable client. Pick the machine the LED panels are physically wired to:

• The screen list then shows the screens that exist on that machine (the mapper fetches them live from that engine), so the wall's content source is a real screen on the target.
• On apply, the wall's region outputs and its fields are written to the chosen machine's engine. For a client the regions are pushed to that client's engine and stored in the project's remote-output-lists (the same sync used for all remote output assignments); for the local machine they go straight to the local output-list. The generated fields are created on the target engine's screen.

The button is disabled while the UI itself is in client mode (clients don't self-manage outputs — drive them from the master).

How it relates to Splitting

Splitting divides one monitor into cells that can each drive a different screen. The LED Space Editor does the opposite: it spreads one screen across many freely-arranged panels, with independent panel sizing and a free desktop origin, and it generates matching content fields so you can address the whole wall or any single panel. See Output Setup for the underlying output-list model.