Editing With Layers

Intro to Layers in Luma

With Luma Layers, you can take an existing image and separate its contents into individually editable elements, layers. A poster can become a background, headline, character, logo, and graphic treatment. A product ad can become a product, shadow, CTA, text block, and backdrop.

Each extracted element receives a transparent background and becomes part of an editable layer stack. You can move it, reorder it, hide it, modify it, replace it, remix it, remake it, or take it out of the composition entirely.

Instead of treating an image as one finished object, Layers turns a flat graphic into its own creative layer modification editor.

What Layers makes possible

Layers lets you preserve the parts of an image that already work while changing only what needs to change.

You can:

• Move a product without regenerating the background.
• Replace a headline without recreating the entire ad.
• Change the subject while preserving the typography and layout.
• Separate an object and reuse it elsewhere on your board.
• Recompose an existing design for a different aspect ratio.
• Create several variations from the same layered source.
• Edit text directly, including its font, size, and color.
• Bring additional assets from your board into the composition.

This is the key difference between editing a layer and generating another image: the change is scoped to the element you selected rather than the entire composition.

How to extract layers from an image

Step 1: Select the image

Start with any image on your Luma board.

This could be:

• An image created in Luma
• An uploaded photograph
• A finished advertisement
• A poster or social graphic
• A product composition
• An illustration
• An existing campaign asset

Right-click the image, or Control-click on supported devices, to open its menu.

Step 2: Choose Extract Layers

Select Extract Layers from the menu.

The extraction window will open with a field where you can describe what you want separated.

Adding an instruction is optional. You can let Luma determine the most useful layers, or tell it exactly which elements you need.

Step 3: Describe the layers you want

Your instruction can be broad:

Separate this image into the most useful editable layers.

or

Break this image into the main visual elements I would need to rearrange or replace.

or

Identify the most important parts of this composition and place each one on its own layer.

It can describe your intended edit:

I want to replace the product and headline while preserving the rest of the design.

or

Separate the elements I would need to change the subject and update the copy without affecting the background.

or

Create layers that will let me swap the main product, change the CTA, and keep the original layout intact.

Or it can specify the exact layer structure:

Extract the headline, car, tree, house, clouds, logo, and background as separate layers.

or

Separate the title, vehicle, foreground plants, building, sky, brand mark, and background into individual layers.

or

Create separate layers for the main text, car, tree, house, cloud layer, logo, and full background.

Another useful structure is:

Extract these layers: the product, product shadow, headline, CTA, logo, foreground graphics, and background.

or

Create individual layers for the bottle, bottle shadow, title, supporting copy, call to action, logo, decorative elements, and backdrop.

or

Separate the composition into the main product, its shadow, primary text, button text, brand logo, foreground design elements, and background.

You do not need to use special syntax. Describe the result naturally, as though you were asking a designer to organize the file for you.

Broad instructions versus exact instructions

Both approaches are useful.

Use a broad instruction when:

• You are exploring an unfamiliar image.
• You are not yet sure what you want to edit.
• You want Luma to identify the main visual components.
• You want a fast starting point.

Example instructions:

Break this poster into the main elements I would need to create new variations.

or

Separate this design into the most useful layers for editing and rearranging the composition.

or

Identify the major visual and text elements in this poster and place them on individual layers.

Use an exact instruction when:

• You already know which elements will change.
• You need a specific production-ready structure.
• Certain objects should remain together.
• You want to avoid unnecessary fragmentation.

Example instructions:

Extract the headline, subheadline, bottle, bottle shadow, flowers, logo, and background. Keep the bottle label attached to the bottle.

or

Separate the title, supporting text, product, product shadow, floral elements, brand mark, and backdrop. Keep all packaging text grouped with the product.

or

Create individual layers for the headline, description, bottle, shadow, flowers, logo, and background. Treat the bottle and its printed label as one complete layer.

or

The most useful extraction is not necessarily the one with the largest number of layers. Separate the elements you need to control while keeping naturally connected elements together.

or

For example, you may want a product and its label to remain one layer, but its shadow to be separate so that the product can be repositioned more easily.

Choosing 1K or 2K extraction

Before processing, you can choose between 1K and 2K resolution.

Use 1K when:

• You are experimenting.
• You expect to create several early variations.
• The source does not contain extremely fine details.
• You are building a rough layout before finalizing it.

Use 2K when:

• You need more detail in the extracted elements.
• The asset contains fine typography or product details.
• You are working closer to final production quality.
• The composition may need to support a larger final export.

A 2K extraction costs twice as much as a 1K extraction, so 1K is often the practical choice for early iteration. Reserve 2K for work that requires more detail or is approaching final production.

What happens after extraction

Once Luma finishes processing the image, it opens the composition inside the Layers editor.

The original flat image is now represented by a stack of named layers. Each extracted image layer includes transparency around the isolated element, allowing it to sit naturally over the layers below it.

The order of the stack determines what appears in front and what appears behind:

• Layers near the top appear in front.
• Layers lower in the stack appear behind.
• The background normally sits at or near the bottom.

You can select a layer directly from the stack or click its visible content inside the editor.

Moving elements around the composition

Each layer can move freely along the horizontal and vertical axes.

Select an element and drag it to a new position. Because the element has been isolated from the rest of the image, the other layers remain in place.

This makes it possible to:

• Shift a headline without affecting the artwork.
• Move a product closer to the center.
• Create more space for copy.
• Adjust the relationship between a subject and the background.
• Rebuild the visual hierarchy of an existing design.

Think of the Layers editor as its own miniature design canvas inside your broader Luma board.

Bringing other board assets into Layers

Extraction is only the starting point.

You can drag another image from your board into the Layers editor and incorporate it into the composition. The added image becomes another layer that can be moved and positioned alongside the extracted elements.

For example, you could:

• Drag in a new product image.
• Add an alternate logo.
• Introduce another character.
• Place a texture behind the headline.
• Add a new foreground element.
• Bring an approved brand asset into an existing design.

You can also begin with an empty Layers editor and drag images or text into it manually. This allows you to construct a layered composition without first extracting a flat image.

Taking layers back onto the board

The relationship between the board and the Layers editor works in both directions.

You can drag an individual layer out of the editor and place it onto your board as its own asset. Once on the board, it behaves like a regular image asset.

You can then:

• Download it
• Crop it
• Modify it
• Use it as a reference
• Place it in another composition
• Combine it with other workflows

This is especially useful when extraction isolates an element you want to reuse, such as a product, character, logo treatment, or graphic motif.

A helpful reminder

Extraction does not require you to describe every visible detail. Focus on the elements you want to move, replace, edit, or reuse.

Complete, thorough instructions always produce complete and thorough results. Explain what should become a layer, what should remain grouped, and what you intend to change later.

When you are unsure, begin with a broad extraction at 1K. Once you understand how you want the composition organized, you can create a more specific extraction or move to 2K for higher-detail work.

Editing With Layers

Understanding the Layers editor

The editor contains two connected views:

• The composition itself
• The ordered layer stack

The composition shows how all visible layers currently fit together. The layer stack shows each individual element and its position relative to the others.

You can use the stack to:

• Select a layer
• Change its order
• Group related layers
• Control visibility
• Delete an element

You can use the composition area to:

• Select visible elements
• Move them horizontally or vertically
• Resize or reposition them
• Arrange imported assets
• Evaluate the finished layout

Changes made in either view are reflected in the other.

Selecting and moving a layer

Click an element in the composition or select its name from the stack.

Once selected, you can reposition it inside the editor. Only that layer moves, while the remaining elements stay where they are.

This makes small layout adjustments much easier. For example, you can move a headline upward to create room for a CTA without disturbing the subject, logo, or background.

Changing the stacking order

The order of the layers determines which elements appear in front.

Drag a layer higher in the stack to bring it forward. Drag it lower to place it behind other elements.

This is often referred to as the layer’s z-order.

You might use it to:

• Move typography in front of a background graphic.
• Bring a foreground object in front of the subject.
• Tuck a decorative element partially behind the headline.
• Place an imported asset between two extracted layers.

If an element appears to have disappeared, check whether it has been placed behind a larger layer.

Grouping related layers

Layers can be grouped by shift+clicking them in the stack, then using the ‘G’ keyboard shortcut or the Group button above the framed selection.

Grouping is useful when several elements belong to one logical component, such as:

• A product and its shadow
• A headline and subheadline
• A subject and a foreground accessory
• A logo and its surrounding graphic treatment
• Several elements that make up a CTA

Groups help keep more complex compositions organized and reduce the chance of accidentally moving one connected element without the others.

Showing and hiding layers

Each layer has a visibility control, it’s the eye icon to the right of the layer name.

Hiding a layer removes it from the visible composition without deleting it. This is useful when comparing variations or testing whether an element is necessary.

For example, you could:

• Compare a poster with and without supporting copy.
• Test several headline treatments.
• Hide the original product after creating a replacement.
• Evaluate alternate backgrounds.
• Create several options inside one editor.

Visibility controls make the Layers editor a useful space for controlled experimentation.

Deleting layers

You can delete any layer that is no longer needed.

Use deletion when you are confident the element should be removed from the composition. Use visibility instead when you may want to bring it back later.

For a safer iteration process, it is usually best to hide uncertain elements first and delete them only after the direction has been approved.

Modifying one image layer

There are many different ways you can prompt to modify or create variants of a layer.

Each image layer includes an instruction field for making a targeted change.

Select the layer you want to edit and describe what that element should become.

Here are some prompt examples:

Replace the red running shoe with a white leather sneaker photographed from the same angle. Preserve the existing lighting direction and shadow placement.

or

Change the selected shoe into a clean white leather sneaker. Match the original camera angle, scale, lighting, and position.

or

Create a white leather version of this sneaker while keeping its current orientation, perspective, proportions, and studio lighting.

Select a background layer and enter:

Change this background into a warm cream studio backdrop with a subtle paper texture.

or

Replace the selected background with a soft cream-colored studio wall featuring a very light paper grain.

or

Create a warm neutral backdrop in a pale cream tone. Keep the texture minimal and preserve the existing lighting direction.

Select a person and enter:

Change the jacket to dark green while preserving the person, pose, expression, and lighting.

or

Replace the selected jacket with a deep green version. Keep the subject’s identity, body position, facial expression, and lighting unchanged.

or

Make the jacket a dark forest green while preserving the person’s appearance, pose, framing, and all other clothing details.

The instruction is applied to the selected layer rather than the entire image.

This is what makes Layers particularly valuable for precision editing: the background, layout, text, and every unselected element can remain intact.

How layer variations are preserved

When you modify an image layer, the newly generated result is added inside the editor as another layer.

The original becomes hidden rather than being immediately discarded. The original and regenerated version are grouped together, allowing you to retain both states.

This gives you a built-in comparison and fallback system:

• Show the original to return to the previous version.
• Show the variation to use the new result.
• Preserve both while continuing to evaluate.
• Create additional variations without losing the source.

This is a more controlled process than repeatedly overwriting an image.

Modifying multiple layers

You can also modify multiple image layers during the same editing process.

This is useful when a requested change affects several related elements.

For example:

Change the product packaging to dark blue and update the product shadow to match its new shape and position.

or

Redesign the selected package in a deep navy color, then adjust the selected shadow so it accurately matches the updated packaging.

or

Make the product packaging dark blue while preserving its branding and perspective. Update the shadow layer so its shape, direction, and softness remain physically consistent.

Another example:

Turn the background into an evening scene and adjust the window reflections to match the new lighting.

or

Change the selected environment to early evening and update the selected reflections so they match the darker sky and warmer interior lights.

or

Transform the background into a blue-hour setting. Adjust the window reflections and highlights to remain consistent with the new time of day.

Selecting only the relevant layers helps keep the change focused while protecting everything else in the composition.

Working with image layers as regular assets

Every image layer behaves like a regular image asset.

You can:

• Download it
• Inspect its file information
• Crop it
• Modify it
• Move it onto your board
• Use it elsewhere in your project

This means Layers is not a closed editing environment. An extracted element can continue through other Luma workflows.

You might extract a character from a poster, drag it onto the board, use it as a reference for another image, and later bring the new result back into the original Layers editor.

Adding images from the board

Drag any compatible image from your board into the Layers editor to add it to the composition.

Once inside, the imported image can be:

• Repositioned
• Reordered
• Grouped
• Hidden
• Deleted
• Cropped or modified

Editing text directly

Depending on the selected text, you can update:

• The written content
• Font
• Font size
• Color
• Position
• Stacking order

Using custom fonts

Layers supports custom fonts so teams can work with their own typography rather than relying only on default options.

• Admin roles can upload and delete custom fonts.
• Team members can view available fonts and use them on any board.

Changing the aspect ratio

The Layers editor supports common aspect-ratio combinations.

This allows the composition itself to change shape, giving you space to rebuild the layout for another format.

The aspect-ratio control can be locked or unlocked.

Locked

When the editor is locked, the content scales with the editor.

Use this when you want the entire composition to grow or shrink proportionally as the editor changes.

Unlocked

When the editor is unlocked, the editor can change size without automatically scaling its contents.

Use this when adapting a design to a new format. The expanded or reduced canvas gives you room to reposition each layer independently.

For example, when converting a square design into a vertical composition, unlocking the editor lets you extend the working area and then move the headline, subject, product, and logo into a new vertical layout.

Copying and pasting an editor

The entire Layers editor can be duplicated by option+click+dragging a new copy on your board.

This is useful before making a major change. One editor can remain as the approved master while the copy becomes a new variation. Keeping a clean master makes it easier to branch into many variations without losing the original composition.

Layer Workflows

The greatest value of Layers is not only that it makes one image easier to edit. It lets one image become the source for an entire family of assets.

Layers enables you to lock-in what works and vary only what changes.

Reusing extracted elements elsewhere

Individual layers can be dragged out of the editor and placed on the main board.

This allows useful reference components to become reusable assets.

For example, you might extract and reuse:

• A product cutout
• A character
• A typographic treatment
• A logo lockup
• A texture
• A background
• A graphic frame
• A decorative motif

Those assets can be modified, cropped, downloaded, or incorporated into other compositions before being brought back into a Layers editor.

The result is a more connected production workflow rather than a series of disconnected exports.

Choosing resolution throughout the workflow

Resolution decisions should reflect the stage of production.

Early development

Use 1K extraction for:

• Layout exploration
• Early campaign branching
• Copy testing
• Rough localization
• Initial aspect-ratio studies

Final refinement

Use 2K extraction when:

• Fine details matter.
• Text or product features require clearer separation.
• The asset is approaching final approval.
• A higher-quality source is needed for the intended output.

Because 2K extraction costs twice as much as 1K, you can conserve resources by proving the structure and direction at 1K before committing to higher-resolution work.

Preserve a master before branching

Before creating a large variation set, keep one clean version of the Layers editor as the master.

Do not use the master for experimental edits.

Instead:

1. Copy the editor.
2. Assign the copy to a specific purpose.
3. Make the required changes.
4. Retain the original layer organization wherever possible.
5. Flatten only after that variation is approved.

This keeps the production structure understandable as the number of outputs grows.

Flattening the final composition

When the layer edits are complete, select the general Layers editor window and choose the Flatten button.

Flattening combines the visible layer stack into one final image.

The flattened output is:

• WYSIWYG: What you see in the editor is what appears in the result.
• Exported as a PNG.
• Created at the same resolution as the highest-resolution image layer in the composition.

Hidden layers do not appear in the visible final composition, so review visibility before flattening.

Also check:

• Layer order
• Text accuracy
• Font selection
• Color and contrast
• Product placement
• Edge quality
• Aspect ratio
• Final composition boundaries

The flattened image is added to your board, where it can continue through the rest of your workflow.

Keep the layered source after exporting

Flattening creates the final image, but the layered composition remains the more valuable production asset.

Keep it available for future revisions.

You may get a request for a different headline, a new offer, another product, or an additional format. With the layered source, you can return to the composition and update only the affected elements.

The flattened image is the deliverable. The Layers editor is the reusable system behind it.