I keep a JPEG from 2017 in a folder on my desktop called “Do Not Delete.” It’s the first headshot I ever delivered to a paying client. The skin is smooth the way plastic is smooth. The eyes have that over-sharpened, almost illustrated quality. The subject’s face looks like it was ironed. She was kind enough to pay me and never hire me again, and I deserved both of those outcomes.

What I didn’t understand then was the difference between removing information and managing it. Most over-retouched headshots aren’t the product of someone trying too hard. They’re the product of someone working in the wrong layer order, with the wrong tools, solving the wrong problem first.

The Real Reason Headshots Go Wrong in the First Place

A headshot is doing a specific job. It has to read as professional at a small size on LinkedIn, hold up at a larger size on a company website, and still look like the actual human being who walks into a room for a job interview. That’s a harder brief than most people realize, and it’s why the retouching stakes are higher than they are for, say, a fashion editorial where stylization is part of the point.

The technical problem is this: skin has two distinct layers of information. There’s the color and tone layer, which includes things like redness, uneven pigmentation, and shadows from texture. And then there’s the texture layer itself, which is the actual surface detail of the skin. When we use a simple blur or healing brush approach, we’re collapsing those two layers into one operation. We smooth the tone, but we also wipe out the texture. The skin stops looking like skin and starts looking like a rendering of skin.

Frequency separation, done correctly, keeps those two layers independent so we can work on each without touching the other.

Setting Up Frequency Separation That Actually Works

I have a Photoshop action called “Eternal Sunshine” that builds my frequency separation stack. Here’s what it’s doing, step by step.

Duplicate your base layer twice. Name the bottom copy “Low Frequency” and the top copy “High Frequency.” On the Low Frequency layer, apply a Gaussian Blur of 6 to 8 pixels for most headshots shot at 300 DPI. If you’re working with a file that came in at 72 DPI for web use only, drop that to 2 to 3 pixels. The goal is to blur just enough that individual pores disappear but the shape of the face stays readable.

On the High Frequency layer, go to Image, Apply Image. Select the Low Frequency layer as your source, set the blending mode to Subtract, scale to 2, offset to 128. Change the High Frequency layer’s blend mode to Linear Light. Now you have a true frequency separation stack.

Work on skin tone using the Clone Stamp at 10 to 15 percent opacity on the Low Frequency layer only. Work on texture using the Healing Brush on the High Frequency layer only. Never cross the streams. That one habit alone will change what your skin looks like on delivery.

Skin Tone Before Anything Else

Before I touch a single blemish, I spend five minutes just looking at the skin tone across the face. This is the step most beginners skip because it feels less satisfying than fixing an obvious problem. But uneven tone is almost always the bigger visual issue in a raw headshot.

I use a Hue/Saturation adjustment layer clipped to the subject. I’ll drop the Reds saturation by 10 to 15 points, which calms down most post-shoot redness without making the skin look gray. Then I’ll use a Color Balance layer, also clipped, and warm up the Midtones by pushing Yellow by plus 3 to 5. This counters the slight green cast that indoor office lighting almost always introduces. These are small numbers. They should feel invisible. If you can see the adjustment, you’ve gone too far.

The Eyes Are Where You Earn Your Rate

I’ve seen retouchers spend forty minutes on skin and four minutes on eyes. That’s backwards for headshots. The eyes are what the viewer locks onto first, and they’re where the “alive vs. rendered” question gets decided.

For eye whites, I use a Curves adjustment layer masked tightly to the sclera only. I’ll raise the midpoint by 5 to 8 points maximum. Anything more and the whites go nuclear and the subject looks like they’re about to levitate. For the iris, I add a very slight Vibrance boost, plus 10 to 15, masked to the iris only. This adds depth without the artificial painted look you get from aggressive Hue/Saturation moves.

The catchlights matter more than people think. If there are two light sources and therefore two catchlights, I’ll often remove the smaller one with a Healing Brush on a stamped layer. One strong catchlight reads as more natural to the eye, even though technically having two is more realistic.

Sharpening at the End, Not the Beginning

Sharpening a headshot before you’ve finished retouching is like painting a wall before you’ve patched the holes. Do it last, on a stamped composite layer.

I use Smart Sharpen over Unsharp Mask for headshots. Amount at 80 to 100 percent, Radius at 0.8 pixels, Reduce Noise at 10 percent. Then I add a layer mask and paint the sharpening back off the skin with a soft black brush at 40 to 50 percent opacity. The eyes, lips, and hair get the full sharpening. The skin gets softened. This single masking step is what makes a headshot look “professionally retouched” rather than just “edited.”

The most important thing I learned from that terrible first delivery isn’t a technique. It’s that restraint is a skill you build, and the way you build it is by learning exactly what each tool is touching before you decide how hard to push it.