A few years into my retouching career, a client sent me feedback that I still think about. She said the model’s skin looked “like a wax figure.” I had spent two hours on that image. I had smoothed, blurred, and healed until everything looked even and clean. The problem was, I had treated the skin like a surface to be erased rather than a structure to be understood. That email was uncomfortable to receive, but it pushed me toward the technique that changed how I work: frequency separation.

What You’re Actually Editing When You Retouch Skin

Skin carries two distinct types of information in any photograph. The first is tone and color: the way light falls across the cheek, the warm flush near the nose, the slight shadow under the jaw. The second is texture: the pores, fine lines, the subtle surface irregularity that tells your eye you’re looking at a real human face.

When we apply a standard healing brush or run a blur filter across the skin, we collapse those two channels of information together. We fix a red spot, but we also drag texture from one area into another. We soften a shadow, but we muddy the pores around it. The result is technically “cleaner” but visually unconvincing, because real skin never looks like that. Frequency separation works by splitting those two layers apart so we can edit each one without touching the other.

Setting Up the Layers in Photoshop (With Exact Settings)

The setup takes about ninety seconds once you’ve done it a few times, but the settings matter. I use an action I named “The Shining” because it reveals something hidden, which felt appropriate.

Start with a flattened version of your image. Duplicate the layer twice. Name the bottom copy “Low Frequency” and the top copy “High Frequency.”

Select the Low Frequency layer and run a Gaussian Blur. For a high-resolution file shot at 50 megapixels or more, I use a radius between 6 and 8 pixels. For a standard 24-megapixel file, 4 to 5 pixels is usually right. The goal is to blur until the skin texture disappears but the color and tonal information remains. You should still see where the highlights and shadows sit. You should not be able to see individual pores.

Now select the High Frequency layer. Go to Image, then Apply Image. In the dialog, set the Layer to your Low Frequency layer, set the Blending to Subtract, the Scale to 2, and the Offset to 128. Click OK. Then change the blending mode of the High Frequency layer to Linear Light. What you’re looking at now is only texture, mathematically separated from the tonal information below it.

Editing Each Layer Without Wrecking the Other

This is where frequency separation earns its reputation. On the Low Frequency layer, use a large, soft brush at low opacity, anywhere from 10 to 30 percent, to blend uneven skin tones. If there’s a red patch near the chin, sample a nearby neutral tone and paint gently over it. You’re evening out the color and light without touching a single pore. The texture layer above stays completely intact.

On the High Frequency layer, use the Clone Stamp tool with the alignment set to “aligned” and the sample set to “current layer.” A blemish on this layer is just a small cluster of textured data sitting in the wrong place. You stamp over it using texture from a nearby area of similar skin. The tonal information, the color, the shadow, none of that moves. The fix is surgical.

For skin that needs tone work across a larger area, I sometimes add a Hue/Saturation or Curves adjustment clipped to the Low Frequency layer. That way I can pull down redness or lift shadows with complete precision, and the texture above it never shifts.

Where Most People Go Wrong With This Technique

The most common mistake I see in my workshops is over-editing the High Frequency layer. People see the texture isolated and start stamping aggressively, repeating patterns across the face until the skin looks like it has been tiled. Real skin texture is irregular. It is not uniform. When you clone, sample from multiple areas and vary your source point often. Keep your opacity on the Clone Stamp between 70 and 90 percent so each pass blends slightly rather than stamping a hard copy.

The second mistake is setting the blur radius too low on the Low Frequency layer. If any texture bleeds into the low frequency pass, the separation breaks down and your edits on each layer will interact in ways you didn’t intend. When in doubt, go slightly higher on the blur radius. A 5-pixel radius on a 24-megapixel file is almost always safer than a 3-pixel radius.

I keep my very first frequency separation attempt on a backup drive. The skin looked like stucco. The pores had been cloned in straight rows across the cheek. I had clearly understood the concept but not the restraint it requires. Looking at it occasionally keeps me honest.

When to Reach for It and When to Leave It Alone

Frequency separation is not the right tool for every image. For editorial work with heavy creative direction, or for portraits where the whole point is raw, unretouched skin, it’s the wrong conversation. For beauty campaigns, commercial headshots, and anywhere the brief asks for clean, polished skin that still reads as skin, it’s the most precise tool I know.

The technique asks you to understand what you’re editing before you edit it. That discipline, separating tone from texture and treating each one as its own problem to solve, is what keeps the result looking like a person rather than a rendering.