Workshop: Frequency separation

Those two words in the context of photography meant nothing to me, so I had to look it up. It is based in science and effectively means that there are two types of pixels in a digital image that can be separated and altered in order to improve the appearance of skin without destroying the basic appearance of the skin or making it look too artificial.

I started with this image of Alex Kelly.

Following the step by step instructions, my processing progressed as follows:

I can barely tell the difference in the skin text here. The instructions direct you to apply a Gaussian Blur to one of the two layers created for the process “but don’t go too far as to make the face unrecognizable”. What do the numbers represent? What is too far? 2? 60? The default was 1.5 so I tried that and then tried 45 but the difference was minimal. I attributed this to the fact the model in my photograph has great skin that needs almost no retouching.

A quick internet search on Frequency Separation brought back a staggering 151,000,000 results in less than half a second, so clearly this is a common process. It wasn’t a difficult process to follow, but as always, context is required to really understand what the process is about and its place within the industry. Having since watched some YouTube videos, it is starting to make sense and I’m beginning to see why retouching professionals rely on it to improve the images in their care.