Glance at the summary firstThe summary from the end of the lesson, as a road map before you start reading.
A layer is like a transparent sheet placed on top of the photo; its order decides what covers what, and by keeping every patch, piece of text or part on its own layer, the original photo isn't touched. An adjustment layer carries no pixels, only a correction instruction; it has its own mask, is turned on and off or deleted when wanted, and its effect is kept measured with opacity. Non-destructive editing is kept in a layered file that preserves the layers, a separate flat copy is exported for sharing; every correction made to skin is kept on a separate layer with a mask covering the face, and is checked by turning it on and off.
Learning goal
You will be able to edit a photo with separate layers stacked on top of each other, without touching its original pixels. You will explain that a layer behaves like a transparent sheet placed on top of the ones below, and that the order of the layers decides what covers what; and you will tell apart a layer that carries pixels from an adjustment layer that carries only a correction instruction. You will add the tone curve and color adjustments as adjustment layers, give them a mask as in the previous lesson, keep their effect measured with opacity, and undo each one whenever you like by turning it on and off or deleting it. You will understand the difference between non-destructive editing and editing that changes the original pixels permanently, together with the difference between a layered file that keeps the layers and a flat file exported for sharing, and you will keep every change you make to skin on a separate layer and check it by turning it on and off.
The question
The drummer wants to use the family portrait they shot on a holiday morning on the cover of an invitation. First they burn the edit that brightened their father's face into the photo and save the file, then they open the file in a program that edits pixels. A charging cable is hanging from the table beside their mother; they erase it. They give the photo some warmth, add some text on top, and save over the file at every step.
The next day their mother says their father's face is a little too bright, and the photo too yellow. When the drummer wants to undo the changes, they see it isn't possible: everything they did has been burned into the photo's pixels and saved over the file. The only way is to start again from the very first file. How do you edit without damaging the original? How do you keep every change in a form you can change your mind about later?
Context
The editing programs you used in the basic editing module already protect the original: every adjustment you make is written not to the photo itself but to an instruction list kept beside it, and can be undone whenever you like. But for jobs like cleaning up a face, erasing a cable, combining two photos or adding text, you most often need to move to a program that changes pixels directly. In these programs, if you aren't careful, every change you make changes the pixels permanently.
Layers are the most basic tool developed against this danger. Programs' names and menus change; in all the paid and free programs mentioned at the start of the module, the logic of layers and adjustment layers is similar. This lesson tells this shared logic, and why in the following lessons every job from cleanup to skin retouching should be done on a separate layer.
Simple explanation
What is a layer?
A layer is an editing unit that can be thought of as a transparent sheet placed on top of the photo, carrying its own content. At the very bottom sits the original photo. Every layer placed on top covers the ones below with its own content; through its transparent places, the ones below show. That is why the order of the layers matters: the layer on top is in front of the one below, and changing the order changes what covers what.
A piece of text, a patch painted over an erased cable, or a part taken from another photo can each sit on its own layer. The original photo is never touched. If you don't like a patch, you delete only that layer; if you want to move the text, you move only the text layer. Before starting jobs like cleanup and retouching, taking a copy of the original photo as a separate layer, or doing the corrections on top of an empty layer, is a habit you will use often in the following lessons.
What is an adjustment layer for?
An adjustment layer is a layer that carries no pixels, only a correction instruction to be applied to the layers below it. When you add the tone curve, color adjustments or brightness as an adjustment layer instead of applying them directly to the photo, the correction sits in the stack like a layer. You can turn it on and off, change its setting or delete it completely whenever you like; the photo below always stays intact.
Every adjustment layer has its own mask. This is the previous lesson's mask logic carried over to layers: in the mask's white places the adjustment is applied, in its black places it isn't. To brighten the drummer's father's face, you add a curve adjustment layer and paint its mask white only over the face; the rest of the photo isn't affected.
Opacity is the proportion that decides how visible a layer will be, or how strong an adjustment layer's effect will be. One hundred percent opacity shows the layer fully, fifty percent halfway, zero not at all. When you think a correction is too much, lowering its opacity instead of setting it up again is most often the fastest and most measured way. Layers also have a blend mode; it decides how a layer combines with the one below, and it will be used in following lessons like dodging and burning.
How is non-destructive editing kept?
Non-destructive editing is keeping all corrections in separate layers or instructions, in an undoable form, without ever changing the original photo's pixels. Its opposite is editing that changes pixels directly and can't be undone once saved. The program's history list only works during the open session; when you close the file or save it in a flat format, the history is lost too.
A layered file is a file format that stores layers, masks and adjustment layers separately. A photo you worked on with layers always needs to be saved in such a format, for example as a TIFF file that supports layers, or in the program's own working format. When you save it in a flat format like JPEG, all the layers merge into a single image and can't be separated. That is why keeping two files is a good habit: a working file that keeps the layers, and a flat copy exported for sharing. The original RAW or JPEG file, on the other hand, stays in the archive without ever being changed.
Some programs also let you add a RAW file together with its editing settings as a layer, inside a container that can be reopened and changed later. This container makes it possible to go back to the RAW settings even after moving to the pixel editor. For example, when weeks later you want to readjust the invitation's white balance, you can open and change the RAW settings inside the container without touching the patch, masks and text on top; the container is updated, and the layers on top stay in place. This flexibility makes the file larger, but for basic settings likely to change later it is most often worth this price.
Layers are also important in protecting skin. Keeping every correction made to a face, brightening, color balancing or retouching, on a separate adjustment layer with a mask covering only the face prevents it from spilling onto other places in the photo. When a curve layer brightening dark skin starts to pull the skin toward gray, lowering the layer's opacity or adding a small warmth adjustment to it brings the skin back to its own color. On light skin, when a color layer reducing redness is too much, the face can turn pale; opacity keeps it measured. Comparing the before and after states by turning each layer on and off clearly shows what changes on the skin.
Visual evidence
The drummer goes back to the family portrait's RAW file, and this time, instead of burning the adjustment that brightened their father's face into the photo, adds it to the pixel editor as a curve adjustment layer. They paint the layer's mask white only over their father's face. For the cable they open an empty layer above the original layer and make the patch there; for the warmth they add a separate color adjustment layer, and for the text a separate text layer. The layers panel now shows five rows: at the bottom the original photo, above it the patch, the curve, the color and the text.
When their mother again says their father's face is too bright, the drummer lowers the curve layer's opacity to seventy percent; their father's dark skin is still bright, but it is now clear that he is sitting away from the window. They watch the difference together by turning the color layer, found too yellow, off and on, and lower its opacity to half. The drummer keeps the file together with its layers as a working file, and exports a flat copy for the invitation. A month later, when the text needs to change, they open and correct only the text layer.
Common mistakes
Common mistakeChanging pixels directly and saving over the file.
The changes can't be undone, and the original photo is lost. The fix: make every correction on a separate layer or adjustment layer, and never change the original file.
Common mistakeSaving a photo worked on with layers only as a flat file.
The layers merge and can't be separated. The fix: keep a layered working file, and export a separate flat copy for sharing.
Common mistakeTrying to set up a correction that is too much again from scratch.
Time is lost and consistency breaks. The fix: make the correction measured by lowering the layer's opacity.
Summary
A layer is like a transparent sheet placed on top of the photo; its order decides what covers what, and by keeping every patch, piece of text or part on its own layer, the original photo isn't touched. An adjustment layer carries no pixels, only a correction instruction; it has its own mask, is turned on and off or deleted when wanted, and its effect is kept measured with opacity. Non-destructive editing is kept in a layered file that preserves the layers, a separate flat copy is exported for sharing; every correction made to skin is kept on a separate layer with a mask covering the face, and is checked by turning it on and off.
Mini assignment
In this assignment you will edit a portrait only with layers, never touching the original photo; a free pixel editor works too.
- Open a portrait in a pixel editor and find the layers panel; see that the original photo sits as a locked layer.
- Add a curve adjustment layer, paint its mask white only over the face and brighten the face; compare three results by setting its opacity to one hundred, seventy and forty percent.
- Open an empty layer and clean up a small spot or a cable in the photo on that layer; then turn that layer off and on and see that the original photo hasn't changed.
- Save the file as a layered working file, and also export a flat copy; the next day open the working file and change only one layer.
Check yourself
- Can you explain the difference between a layer and an adjustment layer?
- When you thought a correction was too much, did you try using opacity?
- Did you keep every correction you made to skin on a separate layer and check it by turning it on and off?
Glossary
| Term | Short definition |
|---|---|
| Layer | An editing unit like a transparent sheet placed on top of the photo, carrying its own content. |
| Adjustment layer | A layer that carries no pixels, only a correction instruction to be applied to the layers below. |
| Non-destructive editing | Keeping all corrections in an undoable form without changing the original pixels. |
| Opacity | The proportion that decides how visible a layer will be or how strong an adjustment's effect will be. |
| Layered file | A file format that stores layers, masks and adjustment layers separately. |
Mini quiz
1. What is the basic difference between an adjustment layer and a layer that carries pixels?
- A) An adjustment layer carries no pixels, only a correction instruction to be applied to the ones below
- B) An adjustment layer always sits at the bottom
- C) An adjustment layer has no mask
- D) An adjustment layer makes the file smaller
2. What happens when you save a photo worked on with layers only as a JPEG?
- A) The layers are stored separately
- B) The masks are kept but the adjustment layers are lost
- C) All the layers merge into a single image and can't be separated
- D) Nothing changes
3. If a curve layer brightening a face is too much, which is the fastest and most measured correction?
- A) Delete the layer and set it up again
- B) Lower the layer's opacity
- C) Reduce the exposure of the whole photo
- D) Save the file in a flat format
4. A curve layer brightening a dark-skinned person's face has started to pull the skin toward gray. Which is the most fitting correction?
- A) Merge the layer with the original photo
- B) Raise saturation across the whole photo
- C) Delete the mask covering the face
- D) Lower the layer's opacity, or add a small warmth adjustment to the same mask
Answers
- A — An adjustment layer sits in the stack as an instruction; the pixels below don't change.
- C — A flat format reduces the layers to a single image; separate layers are lost.
- B — Opacity reduces a correction's effect without setting it up again.
- D — Opacity eases the shift toward gray, and a small masked warmth adjustment brings the skin back to its own color.
Summary card
- A layer is a transparent sheet; its order decides what covers what.
- An adjustment layer is an instruction: it is masked, turned on and off, deleted.
- Keep a correction that is too much measured with opacity.
- Keep a layered working file, export a flat copy for sharing.
- Every correction made to skin on a separate layer with a mask; check by turning it on and off.
Bridge to the next lesson
The charging cable wasn't the only thing the drummer wanted to erase for the invitation's cover. There was a small pimple on their mother's forehead, a few dust spots on their father's jacket, and a nail hole in the wall behind. When they opened an empty layer and painted over the cable with a rough brush, a blurry patch was left in its place. How could they remove these small flaws without spoiling the texture around them and without leaving a trace? How do you remove a pimple, dust or a cable? The next lesson tells about the spot removal, cloning and content-aware fill tools and using them on a separate layer.