Hello everyone.
Can someone please help with simple complete step by step instructions on how to accomplish TWO frequency separation.

I have looked it up previously online and got advice from others but their instructions seem to be missing a step or two and so it never truly worked. I do get the general idea but when I try it things fall apart a bit and the "combined" image never truly looks like the original.

Thank you.

As no one has piped in, I'll make a suggestion (although most major retouchers are far more qualified).

I'm guessing (hopefully correctly) that when you refer to "two" frequency separation that you're just referring to the skin texture versus the skin tone / colour blending, and how to mix the results well.  If so, then this tutorial is as good a starting place as any:

https://youtu.be/Z8mnzVNaiD0

The Frequency Separation technique is a technique that separates the two main frequencies that make up each image: low and high frequencies.
Separation is very useful because it allows you to work independently on each frequency data without any interference.

At high frequencies there are small details such as skin pores, hair, wrinkles, skin imperfections (acne, scars, wrinkles).
At low frequencies, instead, there are image data that contains information about volume passages, shades of shadows and light areas, colors, and tones.
This is a great way to work image for beauty and glamour. Is professional way to have professional post production.
thera are many video tutorial on web, but you can see this, to start.
https://www.youtube.com/watch?v=H1S9RnfbQrg

RoyMayh wrote:
Hello everyone.
Can someone please help with simple complete step by step instructions on how to accomplish TWO frequency separation.

I have looked it up previously online and got advice from others but their instructions seem to be missing a step or two and so it never truly worked. I do get the general idea but when I try it things fall apart a bit and the "combined" image never truly looks like the original.

Thank you.

I think this needs clarification.

I want to separate into
"high frequencies", like 3-4 px
"mid frequencies" like 10 px

and then broad tones.

That is what I meant by "two frequencies".

3 layers stack.

Thanks, I am familiar with basic separation into "high low"  version.
E

I could be wrong, but I have never seen/heard (from my RGG Edu tutorials/video courses, youtube, etc) of this being a thing before so I don't think it's possible.

Deleted. No longer useful as I was still typing my original "Mid Frequency Separation" question while other responded...

I'm not sure if this is what your talking about but there is a thing called asymmetric frequency separation.

https://www.youtube.com/watch?v=UKe4fbDH2XI

To anyone (preferably with experience) that knows:

What is "Mid Frequency Separation" and what would it be used to accomplish?  Versus the traditional High / Low Separation for skin texture detail / general skin colour tones.

As I had never heard of it (assuming that it is a process that is actually used by some) what does this "Mid Frequency" target instead of the two traditional ends of the scale typically used for Frequency Separation?

I have to admit that I'm a little skeptical as you would think that the mid range would essentially undo the basic separation that splits the fine detail texture (high frequency) from the broad colour swaths (low frequency), so I'm having trouble coming up with what the advantage of such a "Middle" approach would be.

Anyone that knows?   Or is this "Mid Frequency Separation" all just a misunderstanding...

Okay so I put my thinking cap on and think I have my answer.

I can figure out how to create a "Mid Frequency" separation layer to be used with a higher than usual frequency layer and a lower than usual frequency separation layer (neither of which overlaps the new "Mid Frequency" layer.

But I can see only one purpose (at least that I can think of) for having this new middle frequency layer. And that is to keep it as it was originally while modifying the higher and lower frequency separation layers.    That would mean that only the more extreme issues (at the high end and low end) would be corrected.  If you did it the other way around, then you'd only be adjust the minor issues and not the major issues, which makes even less sense.

I strongly suspect that someone is either confusing some terms or confusing their "separation point" (between high and low) by how big or small the actual number is, thinking that there should therefore be a need for another number (separation point) for some other purpose. Or confusing multiple high and low separation layers used for separate parts of the image with different characteristics / needs?   Or the number of useful supporting layers, or something along that line...?

My personal conclusion is that creating a third distinct "Mid Frequency Separation Layer" appears to make little to no (useful) sense.  Which is probably why it's so hard to find anything out there on the topic. But I'd love to learn something new if there's someone out there that is experienced in using it and can fill in what I've missed!!

Go download these actions.

http://photoshopsikho.com/download/high … tion-free/

One of them is called Skin Texture or Quick Skin Texture. Study it by playing it from the actions box. I am not sure if it's what you're asking for, but it might help you a little.

Low and High Frequency Separation is just fine...and it works well.  But, there are some other different things you can do to get different High Pass results.

Details ahead... so watch out if you dont like reading.

Sometimes I will add in these  "other options"  to the standard two frequency separation:

Use Different Radiuses
Sometimes I may add in a second apply image layer that is made from a different version of the blurred layer below... that is  "More OR Less blurred"...  sometimes a blur level of 4 pixels, or 7 pixels or 14 pixels for example... can seem to change the perceived frequency spectrum... and I create additional applied image layers above for those different blurred depths. 
Its all the same base image... just different radiuses of blur and apply that pick up different frequencies.  Doing your own experimentations is best.

Use Different Kinds of Image Blur
I can sometimes also use a "Different Method to Blur".  You often notice that Gaussian Blur produces a very distinctive pixel spread across adjoining pixels... well, surface blur, and lens blur, and the blur created by dust and scratches with zero detail for example... will produce blur patterns that are quite different in visual effect.  They are all blurs... but...  they produce a different and somewhat distinctive quality of blur.  Sometimes one of these other blur methodologies picks up a different range or different area of high pass details for me to work with. 

Low Freqency Light and Dark
One other step in a frequency separation kind of work-up...  is that on a face for example... its not all just one shade of foundation color.  Often there are slightly darker areas of "skin shadow areas" like around the nose or eyes or the lip area...where there are specific darker areas of the face/skin.  This is especially true if the model or MUA has used contouring or sculpting in toning the models makeup for high visual impact.  This is the darkness that creates the depth and detailing and the lighting aspects in the image.  I will often separate out just these darker areas, derived from where they belong and actually are... to put back into the final edit.  This gives back some of the "sculpting" and depth and shadows back into the image just where they belong.  And, of course, those darker shadow areas can also be luminance adjusted,  and masked, and edge blended afterwards... so that it looks smooth and has nicer edge transitions. 

Keep Everything Adjustable
And with most of these end results...  blending modes, layer "opacity" sliders, and selective masking...  can all be used to reduce or adjust the layers visual effects.  I can tend to be a bit heavy handed at times, and have preferences toward a certain high key look ( based on my own interpretation of the art I like to see...for me)... so...  reducing the effects back to a 70% application level... often looks better to some... especially to my wife!  LOL.   

Its up to your own taste,  judgement,  and interpretation of what you consider a good look... so keeping all of these layers adjustable with a mask...  allows you to selectively fine tune the effects afterwards... Interactively adding and taking away the visual effects in real time... without the penalty or hassle of having to work with the History pallet.

Its OK to  "Touch Up"  Your Original Low and High Freq Separations
For most of us...the standard low/high frequency separation is just fine.  However, touching up the low freq layer afterwards to make it even better (re-smoothing it, or better blending it, or re-coloring it)  can really make a great difference.   Dont be afraid to go back and re-adjust your low freq skin tones and colors to get a better end product...and put a mask on that layer too... so you can selectively add or subtract skin smoothing...  so its just perfect and exactly positioned. The mask will allow you to interactively add or remove skin smoothing because different parts of the face may need more or less smoothing to look great.

Also, consider touching up the Hi Freq layer afterwards too... by cloning and patching on it... to get rid of the original poor skin details that were captured on day one.  Re-cleaning the HF layer can give you a better end product too (with less pimples or imperfections in it).