I have pointed it out a couple of times during speaking events or workshops I held that I don’t think anyone should use Frequency Separation. Because of that, I get asked a lot to explain why I would say such a heresy. So, when I wanted to do a video on the topic, I realised I wanted to start with this very explanation.
Before I start listing my reasons, let me mention that it can be an immensely useful technique applied in the right situations. Even I — who is somewhat against it — have used it a number of times. Albeit, I used it about once or twice this year — special occasions.
My quarrel with it starts with certain people using it as a gimmick. You see, there are some self-appointed retouching “gurus” out there who like to market their courses and tutorials by saying they will teach you the high-end retouching technique: frequency separation.
The first problem with that is frequency separation is not more of a high-end technique than let’s say using the Curves Adjustment layer is. It is simply just another tool. I’d argue that it’s an inefficient, cumbersome tool at that, but nevertheless a tool.
The second problem is these people extract money out of people who don’t know better by convincing them that this flashy technique is worth it. As a retoucher — who works full-time as an actual retoucher for clients —, I’m here to tell you that it isn’t worth it. We don’t use it that much; don’t give into the hype.
It doesn’t matter whether you do it for clients or not, working non-destructively is extremely important in post-production. However, frequency separation is NOT non-destructive. Hence, anything you do under your group or layers of FS will not show up in your final image. That’s not good because if a client wants to change something you might have to start from the beginning.
Obviously, it is worth mentioning that no healing/cloning method is completely non-destructive as of yet. Hopefully, the people at Adobe will find a way to do that in the future.
If you’ve ever read up on the topic, you know that some people like to create additional layers sandwich in between their Low Pass and High Pass layers. Using these layers, it is possible to even out skin colour and tones. However, using this method, it’s easy to create a divide between tones and colour. In result, this creates a clearly identifiable “frequency separation look.” These are rather undesirable in retouching.
Therefore, since it’s really easy to overdo it, and in reality, it doesn’t bring that much to the table, I’d advise you to stay away from using it. Once you’ve mastered subtle retouching, you can come back to it — just don’t forget that it’s still not the be-all and end-all of high-end techniques.
Lastly, we can’t glance over the fact that using frequency separation is cumbersome and inefficient. As the detail retention of the high pass layer is linked to the blurriness of the low pass layer it is one dimensional: it can only represent one range of detail and tones. (Note: it’s possible to create multiple passes in one separation, but that doesn’t take away from my point.) That means that if we wanted to manipulate different degrees of details or tones we’d need to create new separations again and again.
Not to mention that if we wanted to e.g. remove something that has a prominent presence both in detail and tone, we’d need to jump back and forth between our passes to completely be able to get rid of it.
To reiterate: I’m not saying you should never use it — though, it’s worth noting again how seldom I tend to utilise it. If you are a beginner at retouching, turn to something more useful, like dodging and burning, or the curves layer; revisit frequency separation once you are more advanced. Even then, it’s just a technique.
Other this writing, I have the following video on the same topic in case you are more a visual person. Check it out: