> ## Documentation Index
> Fetch the complete documentation index at: https://comfytv.org/llms.txt
> Use this file to discover all available pages before exploring further.
# Layer Editor
> A full raster + vector image editor built into the node — layers, masks, selections, brushes, shapes, text and transforms — like a compact GIMP/Photoshop right inside your ComfyTV graph.
## What this node does
Layer Editor is a rich in-browser editor that produces images for your graph. You build a multi-layer document on an artboard — paint, mask, add shapes and text, transform and warp — and the editor captures the flattened composite (and, when you have several layers, a per-layer batch) as ComfyTV image snapshots. Nothing runs on the GPU; all editing and capture happen live in the browser, so there is no ▶ Run wait for the drawing itself.
Because it sits in the `ComfyTV/Input` category, it is a starting point: draw or assemble something and feed it downstream to generation, editing, or compositing stages. Its outputs are ComfyTV snapshots, not native tensors — insert a **Bridge** to feed native ComfyUI nodes ([bridges docs](https://github.com/jtydhr88/ComfyTV/blob/main/docs/bridges.md)).
## When to use it
* Assemble a rough composite (background + subject + logo) before sending it to an edit or inpaint stage.
* Paint a precise mask by hand — brush onto a layer's mask channel, use the magic wand, or convert a selection.
* Add title text, shapes, or annotations to an image without leaving ComfyTV.
* Bring in a PSD, edit its layers, and export the result — or export your work back out as a PSD.
## The editor
The card is a three-zone editor with a top toolbar, a left tool strip, the canvas, and a right layers panel. A maximize button (top-right) blows it up to fullscreen; the same button toggles back.
**Tool strip (left).** One button per tool: **Select** (move layers), **Transform** (scale/rotate), **Marquee** and **Ellipse marquee**, **Lasso**, **Magic wand**, **Brush**, **Eraser**, **Paint bucket**, **Shape**, **Warp**, and **Text**.
**Context toolbar (top).** Shows the active tool's name and its options, which change per tool:
* **Brush / Eraser** — a **Content vs Mask** target toggle (paint the pixels or paint the layer's mask), plus **size** (2–400), **hardness** (0–1), **opacity** (0–1), and a **color** swatch (color shows for the brush on content).
* **Shape** — pick a kind: **rectangle, ellipse, line, polygon, star, arc, spiral**. Choose **New layer** vs **Combine** into the active layer; polygon/star expose **sides** (3–24), star a **point ratio**, spiral **turns**; toggle **fill** and **stroke** with their colors and a **stroke width** (1–100).
* **Selection tools** (marquee/ellipse/lasso/wand/bucket) — the wand/bucket expose **threshold** (0.01–1), **anti-alias**, and **contiguous**; other selection tools expose a **radius** and selection-modify actions **feather / grow / shrink / border**, plus **Fill** and **Stroke** the selection with the current color.
* **Transform** — **Apply** or **Cancel** the pending scale/rotate.
* **Warp** — a mesh warp with a **3×3 / 4×4 / 5×5** grid and **Apply / Cancel**.
On the right of the toolbar sit **Undo / Redo**, **Import PSD**, **Export PSD**, and **Fit view**.
**Canvas (center).** The artboard. Hold space to pan; scroll to zoom; drag with the active tool. Double-clicking a text layer opens an inline text-edit popup.
**Layers panel (right).** The layer stack. Add layers from your asset library or from a file, reorder them, set each layer's **blend mode** and **opacity** (0–100%), and toggle locks (such as lock-alpha on raster layers). Layers can be raster, shape, or text.
## Workflow / step by step
1. Drop the Layer Editor node onto the canvas. It opens with a blank artboard (default 1024×1024; the size is stored internally).
2. Add a base layer — upload an image or add one from the asset library via the layers panel.
3. Build up: paint on a new layer, add a shape or text layer, or bring in more images. Reorder and set blend mode/opacity per layer.
4. Refine with masks and selections — brush onto a layer's **Mask** target, or make a selection with the marquee/lasso/wand and feather/grow it.
5. Transform or warp pieces into place; Apply when happy.
6. The editor continuously captures the composite. The flattened result flows to **image**; when the document has more than one layer, a per-layer batch flows to **images**.
7. Connect **image** (or **images**) downstream, or Export PSD to keep an editable copy.
## Inputs and outputs
| Port | Type | Meaning |
| ---------------- | ---------------- | ---------------------------------------------------------------- |
| **image** (out) | `COMFYTV_IMAGE` | The flattened composite capture |
| **images** (out) | `COMFYTV_IMAGES` | Per-layer batch (only when the document has more than one layer) |
The document itself (`layer_state`), the artboard `width`/`height`, and the capture URLs are stored internally on the node and are not sockets you wire.
## Tips
* The **Content vs Mask** toggle is the heart of hand-masking: switch to **Mask** and brush white to reveal / black to hide, without touching the pixels.
* **Import PSD** brings a layered file in with its layers intact; **Export PSD** sends your document out as a real PSD you can reopen elsewhere.
* Undo/redo cover editing steps — use them freely; the layers panel locks (like lock-alpha) prevent accidental paint on a finished layer.
* The per-layer **images** batch only appears once you have two or more layers, so a single-layer document just outputs the flattened **image**.
## Related nodes
* **Storyboard Editor** — reuses this same editor engine, one instance per shot board.
* **Image Edit / Inpaint** — natural downstream stages for a composite or hand-painted mask.
* **Bridge (from Image)** — convert the capture to a native ComfyUI `IMAGE` for core nodes.