Frequency Separation: The Game-Changing Technique That Transformed My Portrait Editing

When I first learned about frequency separation, I’ll be honest—it felt intimidating. The name alone made me hesitant. But once I understood what was actually happening, it became the foundation of how I approach every single portrait that comes across my desk. Today, I want to walk you through this technique because I genuinely believe it will change the way you retouch.

What Frequency Separation Actually Does

Frequency separation splits your image into two layers: one containing color and tone information (low frequency), and another containing texture and detail (high frequency). This separation lets us correct skin tone and blemishes without destroying the natural skin texture that makes portraits look alive.

Before I understood this, I’d spend hours trying to clone out a blemish, only to end up with plastic-looking skin. Now? I can remove that same blemish while keeping the skin looking believable. It’s the difference between editing that looks edited and editing that looks natural.

Setting Up Your Layers

Here’s exactly how I set this up in Photoshop:

  1. Duplicate your base layer twice. Name the first duplicate “Low Frequency” and the second “High Frequency.”

  2. Work on the Low Frequency layer first. Go to Filter > Blur > Gaussian Blur and apply a radius between 8-15 pixels (depending on your image resolution and how much smoothing you want). This removes texture while keeping broad color and tone.

  3. Prepare the High Frequency layer. On this layer, go to Filter > Other > High Pass. Use a radius between 2-5 pixels—this extracts all the texture detail.

  4. Change the High Frequency layer’s blend mode to Linear Light or Overlay. I usually start with Linear Light, which gives me more control.

Now you have two separate layers to work with. The magic happens here.

The Actual Retouching Process

On the Low Frequency layer, I use the Healing Brush or Clone tool to address color issues, redness, and uneven tone. Because this layer has no texture, I can be generous with my strokes without creating that blurry, airbrushed look. I’m not worried about destroying detail because detail isn’t there.

On the High Frequency layer, I only work with dodge and burn tools, and sometimes a soft eraser on a very low opacity (10-15%). I never clone or heal on this layer. The goal is to preserve texture while gently reducing the appearance of imperfections through subtle shading.

This two-pronged approach is what separates professional-looking retouching from overdone retouching.

My Practical Tips From Years of Client Work

Start conservative. I apply less blur on the Low Frequency layer than I think I need. I can always apply it again, but I can’t undo over-processing. A 6-8 pixel blur handles most situations beautifully.

Use adjustment layers on top. Once my frequency separation is complete, I add Curves or Levels adjustment layers to fine-tune color and brightness across both layers together. This keeps everything cohesive.

Check your work at 100% zoom. Frequency separation magic can disappear when you’re zoomed out. Always zoom in to verify that texture looks natural before calling it done.

Save before you flatten. Keep your layered file. Client wants a revision? You’ll thank yourself for having those editable layers intact.

Why This Matters

Frequency separation respects the photograph underneath. We’re not painting over reality; we’re refining it. Every person has skin texture, and that texture tells the story of their face. When we preserve it while removing distracting blemishes and evening tone, we create portraits that look like the best version of the person—not like a smoothed-out version of someone else.

I encourage you to try this on your next portrait. Yes, it takes a few extra minutes to set up, but once you see the results, you’ll understand why this technique has become non-negotiable in my workflow.

Your clients will notice the difference. More importantly, you will.