Frequency separation is a Photoshop technique used to retouch skin. If you’ve ever wanted to learn this technique and apply it to your images, you’re in luck. In this tutorial, we’ll show you how to set up the layers to perform frequency separation in Photoshop. Let’s get started.
What Is Frequency Separation?
Frequency separation is a retouching method by which an image is divided into two layers; low and high frequencies. The low frequency represents the color information while the high frequency represents mostly the texture details, though in some instances you will see some color there too.
The idea is that when you use frequency separation, you’ll work on each layer independently. This way, editing is more localized and serves to be more effective in removing blemishes, skin discolorations, bags under the eyes, and so on.
Frequency separation is an older retouching method that used to be employed by professional retouchers but has since been largely sidelined in favor of other, more modern techniques. But many photographers still use frequency separation because it’s a relatively quick and easy way to retouch skin without having to rely on third-party plugins or professional retouchers.
8-Bit vs. 16-Bit Frequency Separation
Two common setups are used in frequency separation, one for 8-bit and one for 16-bit images. The traditional way to determine which one to use is to simply go by how your image was rendered. Most phones and DSLRs have a default 8-bit mode, but many advanced systems also have 14 or even 16-bit options.
You can see which one your image is by checking the image tab at the top in Photoshop that contains the file name. At the end of the name, you’ll see either RGB/8 for 8-bit or RGB/16 for 16-bit. If your image is a 14-bit image or higher, you’ll want to use 16-bit frequency separation.
It’s worth noting that over the years a few Photoshop users have noticed that they really couldn’t tell a difference between using one method or the other in side-by-side comparisons, regardless of bit count. That being said, we’ll include both methods in our tutorial.
Step 1: How to Set Up Frequency Separation
There are several variations of this setup. We have chosen one of the basic methods of frequency separation that use as few layers as possible to get the best results quickly.
We’re going to show you how to create two separate background copy layers representing the low and high frequencies. We’ll name each accordingly and apply a blur to the low frequency and use an Apply Image rendering on the high frequency.
With a few other minor adjustments, it’s not a complex setup by any means but one that could be automated by creating a Photoshop action for it. Lastly, we’ll group these layers for organizational purposes so that additional editing can be done if needed.
If you want to follow along with us, you can download the example image from Pexels.
- Press Ctrl + J twice to make two copies of the background layer.
- Double-click on the first copy and name it “Low Frequency”.
- Double-click on the second copy and name it “High Frequency”.
- Unckeck the box of the High Frequency layer.
- Select the Low Frequency layer. Go to Filter > Blur > Gaussian Blur.
- Change the Radius to a pixel size that is appropriate for the image. This will vary, but you’ll want to adjust it so that the texture details start to fade. For this image, we chose 8. Then click OK.
- Check the box of the High Frequency layer and then select the layer.
- Go to Image > Apply Image.
- Here’s how to change the settings for 8-bit images. In the Apply Image menu, change the Layer to Low Frequency (or the first layer copy), change the Blending to Subtract, uncheck the Invert box, change the Scale to 2, and Offset to 128.
- For 16-bit images, change the Layer to Low, check the Invert box, change the Blending to Add, and Scale to 2. Leave Offset at 0.
- Change the Blend mode to Linear Light.
- Click on the High Frequency layer and Shift + click on the Low Frequency layer to select both layers.
- Press Ctrl + G to group the layers together.
- Name the Group “FS” for frequency separation.
If you’re a fan of third-party plugins to make your photo editing workflow more streamlined, we show you how to use PortraitPro to transform your portraits.
Step 2: Edit the High Frequency Layer
Now that we have the frequency separation layers prepped in Photoshop, we can retouch this image. We’ll start with the High Frequency layer, where we’ll use the Clone Stamp tool (S) to remove blemishes, stray hairs, and other small troubled regions.
For most of the work, you’ll want to choose a Hard Round brush, with Flow set to 100%. Make sure Sample is set to Current Layer.
To start, zoom in by pressing the Z key. Move the mouse left or right to enlarge the image to full screen. Start with the top of the face. Identify the areas that need to be retouched, like the strand of hair on the forehead, the black mark on the woman’s right eye, and the small patches of rough skin.
We cover how to use the Clone Stamp tool in detail if you need a refresher.
Then continue down the image, cleaning up any blemishes and rough spots. When you need to use the Clone Stamp tool near lines and borders between different pixels, you can switch to a Soft Round brush to avoid hard edges from causing pixel distortion.
You don’t have to get everything perfect on your first pass. Even when you move to the next step, you may find it necessary to come back to the High Frequency layer to do more work.
There may be certain spots that would be better solved with other tools like the Healing and Spot Healing tools. If you’re a beginner, experiment to see which other tools work for you.
Step 3: Edit the Low Frequency Layer
The Low Frequency layer is where most of the color information resides. It’s here where we’ll smooth out transitions between colors to achieve a professional look.
There are a couple of ways to approach smoothing out patchy areas. One way is to work directly on the Low Frequency layer using the Lasso tool and applying Gaussian Blur with every selection.
But if you want to work non-destructively, so you can go back later to correct mistakes, you can create an additional layer above the Low Frequency layer and convert it to a Smart Object.
Here are the steps:
- Select the Low Frequency layer. Press Ctrl + J to duplicate it.
- Right-click on the new layer and select Convert to Smart Object.
- Go to Filter > Blur > Gaussian Blur.
- Once again, choose a value where the texture is faded. We chose 18 for this image. Press OK.
- Hold and press the Alt key and the Mask icon to create a negative mask (black).
Then with a Soft Round brush with the Flow set to a lower amount, like 50%, start brushing the effect wherever a smoothing effect would be useful. Avoid brushing over the edges because this would likely affect the dimensionality and cause the image to look flat.
If you need a refresher, check out our guide for using the Brush tool in Photoshop.
Here’s our image before working with the Low Frequency layer converted to a Smart Object:
Here’s our completed image:
Additionally, you can reduce the Opacity of the layer if you want to reduce the effects. We reduced it to 75%.
Frequency Separation Is a Photographer’s Best Friend
If you want to retouch your pictures but don’t want to spend countless hours doing so, frequency separation in Photoshop is the perfect choice for photographers who want to get the job done quickly and effectively.
Not only will a technique like frequency separation save you time, but you also don’t have to send your pictures to a professional retoucher.
Read the full article here