Forums > Digital Art and Retouching > HighPass Sucks (+ solution)

Retoucher

James Minshall

Posts: 218

Bedford, Indiana, US

Great great work here folks!

I've been using something similar for a couple of years.. just didn't realize it had some fancy name like spacial separation.

I had:

HP 3px
HP 5px
Dodge/Burn
Blank Retouch
Surface and Gaussian Blur

Your method is slightly easier than what I've been doing.  Thanks alot!

May 04 09 07:52 am Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

syd47421 wrote:
Great great work here folks!

I've been using something similar for a couple of years.. just didn't realize it had some fancy name like spacial separation.

I had:

HP 3px
HP 5px
Dodge/Burn
Blank Retouch
Surface and Gaussian Blur

Your method is slightly easier than what I've been doing.  Thanks alot!

Shutterslam wrote:
Do we get to call this the Sean Baker effect?

Noooooo.  As syd notes, variations on the techniques explained within this thread have been around for a long time.  The earliest references I know to bandpass / bandstop type skin retouching are from byRo over on the RTPro forums.  Another gentleman independently discovered the same and discussed it as part of his retouching workflow on his blog here.  I suspect that grahamsz also came up with this on his own, and certainly his combination of this with the frequency separation technique which started this all off is the first I've seen it described openly.

What I outlined at the outset of this thread isn't revolutionary.  It's a replacement for a high pass filter, and not all of it original work.  What's important is how everyone is using it.  How you're experimenting; learning; and sharing your findings.  I'll be in and out of the loop for the next few weeks, but I hope this carries on.  Post what you find; ask a question if you have one.  Just share the information.

May 04 09 09:46 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Ahh cool - I tried to google my technique and couldn't really come up with any references to it anywhere. From what little i remember of signal processing i think it's probably best described as either a "spatial bandstop" or "spatial band-attenuation" but it sounds like it's functionally identical to the linked technique

May 04 09 02:07 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

A few more observations with implications for skin processing while I take a break from loathing my professional education choices smile.

  1.) Something funny happens with PS when you try to apply a second degree of frequency separation to the initial high frequency data, knocking accuracy down by a few orders of magnitude (slightly visible in some areas; still a bit more accurate than HP).  Three total separations results in clearly visible degradation.  I'm not entirely sure what's causing it, but suspect there may be a solution - slicing off the 'smallest' data first and continuing to make new separations off the lower slice vs the upper.  Please share if you test this; it'll probably be a week or two for me to get the time.

   2.) If the above doesn't pan out and another solution to get around it can't be found (where're the Adobe engineers when you need someone to make a code fix for rounding errors? smile ) my own workflow (for quantity edits at least) will likely evolve to using two layer groupings of quick-select'd patches of skin.  Run the 'new-HP' (nHP?), invert, and GB to taste (retain fine detail).  Create a faux-texture layer to go overtop where we're taking out the old detail... I'm thinking overlay blend mode, light direction to match pores, smart layer w/ HP & GB filters applied to match what was removed, and opacity tuned again to taste.  Setup as an action with automatic selection of the brush tool to tune the group mask and healing brush to perfect the retained fine texture.

  3.) I'm a bit concerned about color shift from the inserted texture... I don't want to have to merge the skin retouches to change their blending to luminosity, but I can't think of another solution - ?

May 04 09 02:20 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

grahamsz wrote:
From what little i remember of signal processing i think it's probably best described as either a "spatial bandstop" or "spatial band-attenuation" but it sounds like it's functionally identical to the linked technique

Well, at risk of taking all the fun out of things, I'll adopt those terms myself then, too smile.  FWIW, the way you've combined the concepts, yours is IMO the most accurate technique currently being discussed.

Perhaps also of note - at one point the guys on RTPro had a very weakly-functional bandstop plugin up and running in PS (through a third party plugin engine), but it was never stable on my box.  I'm sure it's still out there though if anyone else can get it to work.

May 04 09 02:24 pm Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Sean Baker wrote:
Well, at risk of taking all the fun out of things, I'll adopt those terms myself then, too smile.  FWIW, the way you've combined the concepts, yours is IMO the most accurate technique currently being discussed.

Mine's only accurate because it builds on your work. I'm thrilled with the combined effort though - the image you commented on my portfolio maybe had 20 mins of retouching work done on it.

Sean Baker wrote:
Perhaps also of note - at one point the guys on RTPro had a very weakly-functional bandstop plugin up and running in PS (through a third party plugin engine), but it was never stable on my box.  I'm sure it's still out there though if anyone else can get it to work.

I was thinking of the same thing actually, or moreover I was thinking of building a full scale graphic equalizer that would let you mask your image then adjust the different frequency bands. Unfortunately i have far too many projects to want to do that now, and I haven't tried writing a photoshop plugin since ps3 (and it was a bitch back then).

May 04 09 02:56 pm Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

grahamsz wrote:

Sean Baker wrote:
Well, at risk of taking all the fun out of things, I'll adopt those terms myself then, too smile.  FWIW, the way you've combined the concepts, yours is IMO the most accurate technique currently being discussed.

Mine's only accurate because it builds on your work. I'm thrilled with the combined effort though - the image you commented on my portfolio maybe had 20 mins of retouching work done on it.

Sean Baker wrote:
Perhaps also of note - at one point the guys on RTPro had a very weakly-functional bandstop plugin up and running in PS (through a third party plugin engine), but it was never stable on my box.  I'm sure it's still out there though if anyone else can get it to work.

I was thinking of the same thing actually, or moreover I was thinking of building a full scale graphic equalizer that would let you mask your image then adjust the different frequency bands. Unfortunately i have far too many projects to want to do that now, and I haven't tried writing a photoshop plugin since ps3 (and it was a bitch back then).

As soon as I learn more of exactly what is going on with this, I'll try writing a plugin like you describe here.

I'll have a lot of questions, though. smile

May 04 09 06:14 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

FWIW, noise + difference clouds in a 50% gray smart object layer with HP + GB + emboss filters does make for a very nice substitute skin texture.  Uniform noise seems to be better than Gaussian, but that may just be me.  Really wish I'd thought of this smart object thing a year ago....

Edit: Faux texture survives resizing leaving a pleasant texture intact.

May 04 09 06:18 pm Link

Photographer

A_Nova_Photography

Posts: 8652

Winston-Salem, North Carolina, US

Ok, I've been playing with this and I don't know if I'm doing it wrong... I remember in the OP that it was discussed having the HF layer at 135% (give or take). If I run the action (been playing with different GB %) I really don't see a sharpness difference at any GB level until I duplicate the HF layer and then back off the opacity till it looks good...

May 05 09 08:36 am Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

ACPhotography wrote:
Ok, I've been playing with this and I don't know if I'm doing it wrong... I remember in the OP that it was discussed having the HF layer at 135% (give or take). If I run the action (been playing with different GB %) I really don't see a sharpness difference at any GB level until I duplicate the HF layer and then back off the opacity till it looks good...

That just means that you are doing it right. Remember that the original intent of this is tonseparate the frequencies while retaining the original image (which is why it looks the same). Duplicating the high frequency portion to increase contrast is one of the many creative applications for it.

May 05 09 09:25 am Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean, grahamsz,

Just so you guys know, I've been fighting with this one image for months now trying to get it to look the way I wanted. I had all but given up. It was giving me a substantial amount of difficulty since my retouching skills aren't so great. It's now my avatar. Granted, it does need some more work, but I'm extremely happy with it as-is and it's much better than any other attempt I've made on it yet.

So a big thank you to both of you for all the information. And a friendly bump to the thread so others won't miss it. smile

May 05 09 06:21 pm Link

Photographer

MEK Photography

Posts: 6571

Westminster, Maryland, US

An amalgam of Sean and Grahamsz' techniques:

Thanks to both!!

https://modelmayhm-1.vo.llnwd.net/d1/photos/090506/07/4a01a4e098d28.jpg

May 06 09 08:13 am Link

Retoucher

James Minshall

Posts: 218

Bedford, Indiana, US

grahamsz wrote:

Sean Baker wrote:
Well, at risk of taking all the fun out of things, I'll adopt those terms myself then, too smile.  FWIW, the way you've combined the concepts, yours is IMO the most accurate technique currently being discussed.

Mine's only accurate because it builds on your work. I'm thrilled with the combined effort though - the image you commented on my portfolio maybe had 20 mins of retouching work done on it.


I was thinking of the same thing actually, or moreover I was thinking of building a full scale graphic equalizer that would let you mask your image then adjust the different frequency bands. Unfortunately i have far too many projects to want to do that now, and I haven't tried writing a photoshop plugin since ps3 (and it was a bitch back then).

I tried writing what you'd call a 'graphic equalizer' for images a couple of years ago using Highpass and other tricks.  It KINDA worked.. but was clunky and awkward to use.  Maybe I'll try it again with this new method.. I'll keep the thread posted!

May 06 09 12:12 pm Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean Baker wrote:
A few more observations with implications for skin processing while I take a break from loathing my professional education choices smile.

  1.) Something funny happens with PS when you try to apply a second degree of frequency separation to the initial high frequency data, knocking accuracy down by a few orders of magnitude (slightly visible in some areas; still a bit more accurate than HP).  Three total separations results in clearly visible degradation.  I'm not entirely sure what's causing it, but suspect there may be a solution - slicing off the 'smallest' data first and continuing to make new separations off the lower slice vs the upper.  Please share if you test this; it'll probably be a week or two for me to get the time.

   2.) If the above doesn't pan out and another solution to get around it can't be found (where're the Adobe engineers when you need someone to make a code fix for rounding errors? smile ) my own workflow (for quantity edits at least) will likely evolve to using two layer groupings of quick-select'd patches of skin.  Run the 'new-HP' (nHP?), invert, and GB to taste (retain fine detail).  Create a faux-texture layer to go overtop where we're taking out the old detail... I'm thinking overlay blend mode, light direction to match pores, smart layer w/ HP & GB filters applied to match what was removed, and opacity tuned again to taste.  Setup as an action with automatic selection of the brush tool to tune the group mask and healing brush to perfect the retained fine texture.

  3.) I'm a bit concerned about color shift from the inserted texture... I don't want to have to merge the skin retouches to change their blending to luminosity, but I can't think of another solution - ?

I'm curious about this second degree of separation. Whenever I try it, I get blotchy results. Almost like turning up the contrast a bit too much. I haven't even tried a third separation. I have been experimenting with different ways of separating out the spatial frequencies with little success. I'm trying to come up with a way like grahamsz described similar to an equalizer. Now it seems syd47421 may have the starts of a method.

I'll try experimenting more. There may be some ratio involved that becomes additive with successive separations.

May 06 09 12:28 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

syd47421 wrote:
I tried writing what you'd call a 'graphic equalizer' for images a couple of years ago using Highpass and other tricks.  It KINDA worked.. but was clunky and awkward to use.  Maybe I'll try it again with this new method.. I'll keep the thread posted!

You reminded me of something I read once, so I had to dig it up:

This link may be of some interest to anyone who's found the rest of this worthwhile, and I think the 'equalizer' it links to may also be of some use (if only as a template).

Photons 2 Pixels Images wrote:
I'm curious about this second degree of separation. Whenever I try it, I get blotchy results. Almost like turning up the contrast a bit too much. I haven't even tried a third separation. I have been experimenting with different ways of separating out the spatial frequencies with little success. I'm trying to come up with a way like grahamsz described similar to an equalizer. Now it seems syd47421 may have the starts of a method.

I'll try experimenting more. There may be some ratio involved that becomes additive with successive separations.

For myself, the greatest difficulty was just remembering which layer I was working with at the time and being sure to blur & subtract with / from the right layer at the right time.  In the end, it does work; it's just a logistic PITA.  There's probably a more elegant way of doing it with scripting and working in another window; i.e. returning the 'slices' to the main file while working with a progressively deconstructed version in a temporary window.  More time than I have ATM sad.

Mike & P2P - Great work, and thanks for sharing it!

May 06 09 12:35 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

The preceding link also led me to find this.  I've not been able to test it, but for those who work in 8bit it may be right up your alley for a simple, one step solution.

And of course for Bob Randall, we have this link so that he can work in LAB too big_smile.

May 06 09 12:42 pm Link

Photographer

Kevin Connery

Posts: 17824

El Segundo, California, US

Sean Baker wrote:
The preceding link also led me to find this.  I've not been able to test it, but for those who work in 8bit it may be right up your alley for a simple, one step solution.

And of course for Bob Randall, we have this link so that he can work in LAB too big_smile.

The author of the tech-slop articles hangs out on photoshoptechniques.com's forums. If you're interested in some fascinating--if esoteric--image processing techniques, that's one place to go. The RetouchPro forums tend not to get that analytical.

May 06 09 01:10 pm Link

Retoucher

James Minshall

Posts: 218

Bedford, Indiana, US

regarding the 'graphic equalizer'

just toying around.. might push someone in the right direction...

layer order:

frequency 2
frequency 5
retouch
blur 2
blur 5

frequency layers are 'apply image, add the correct blur layer, invert with scale 2.  blending mode to linear.  the problem is when you stack two linear blending modes, it gets all sharp and contrasty.  to solve, on frequency 2 and frequency 5 layer, bring up LEVELS and adjust OUTPUT LEVELS to 50 and 205 on BOTH layers.  That adds a tiny bit of contrast, but I'd imagine if you tweak those numbers a bit, they'll go right together seamlessly.

now.. what I can't remember is why I tried doing this in the first place, LOL!

Anyone have ideas for practical uses for a 5 frequency equalizer?

May 06 09 02:00 pm Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean Baker wrote:

syd47421 wrote:
I tried writing what you'd call a 'graphic equalizer' for images a couple of years ago using Highpass and other tricks.  It KINDA worked.. but was clunky and awkward to use.  Maybe I'll try it again with this new method.. I'll keep the thread posted!

You reminded me of something I read once, so I had to dig it up:

This link may be of some interest to anyone who's found the rest of this worthwhile, and I think the 'equalizer' it links to may also be of some use (if only as a template).


For myself, the greatest difficulty was just remembering which layer I was working with at the time and being sure to blur & subtract with / from the right layer at the right time.  In the end, it does work; it's just a logistic PITA.  There's probably a more elegant way of doing it with scripting and working in another window; i.e. returning the 'slices' to the main file while working with a progressively deconstructed version in a temporary window.  More time than I have ATM sad.

Mike & P2P - Great work, and thanks for sharing it!

I'm up for the task of writing up the script/action for it. I just don't know the exact procedure you're using past the first separation.

I've tried something similar to that link you posted to RetouchPro with 5 high frequency separated layers. I used a single GB layer as a smart filter and progressively turned up the blur radius before applying to each successive HF layer. The blur radii I used were: 0.8, 1.6, 3.2, 6.4, and 12.8.

It seemed to work OK and I haven't tried it on a range of images yet, but it doesn't seem quite right. I'll have to play with it more and see.

May 06 09 02:05 pm Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

syd47421 wrote:
regarding the 'graphic equalizer'

just toying around.. might push someone in the right direction...

layer order:

frequency 2
frequency 5
retouch
blur 2
blur 5

frequency layers are 'apply image, add the correct blur layer, invert with scale 2.  blending mode to linear.  the problem is when you stack two linear blending modes, it gets all sharp and contrasty.  to solve, on frequency 2 and frequency 5 layer, bring up LEVELS and adjust OUTPUT LEVELS to 50 and 205 on BOTH layers.  That adds a tiny bit of contrast, but I'd imagine if you tweak those numbers a bit, they'll go right together seamlessly.

now.. what I can't remember is why I tried doing this in the first place, LOL!

Anyone have ideas for practical uses for a 5 frequency equalizer?

Actually, I hadn't thought of that. Using a levels or curves adjustment on the HF layers I made. I'll have to see if that does anything.

I can see a use as the resolution of images goes up more and more. This will give more control of exactly what you're working on and I would image speed up image processing since you only have to concentrate on a specific range at any given time.

So far just the single separation has speeded up my workflow while making it easier at the same time. The healing brush seems more forgiving on the separated layers than it does on the combined layer.

May 06 09 02:12 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

In a few days I'll try to sit down and write a better explanation for the repeat seps; I apologize that I've not been able to better communicate it to this point.  At a guess, the oddities you may be seeing may be from not re-blurring the intermediate separations?  From your description it sounds like you might be duplicating (increasing contrast in) each of the successively more detailed areas.  Complete guess, though.

For syd - also a wild ass guess - the output #s you're looking for are likely 63 & 191 if you're doing what I think you are.  You could also blend the layer with a 50% gray for the same effect... not sure offhand which will give a more accurate result.

And for anyone else looking at skin techniques, one more you could try would be:

  1.) Do the frequency sep as at the beginning of this thread; separating the details you want to keep (high frequency layer).
  2.) Make another copy of the original image + surface blur until all the nasties are gone.
  3.) Gaussian blur that same surface-blurred copy at the radius you separated at.
  4.) Stack the high frequency layer atop the SB + GB layer, mask away & paint back in where you want clean skin.

This doesn't solve the intermediate frequencies issue (oddities with some resizing), but does handle the edge transitions nearly perfectly.

May 06 09 02:16 pm Link

Photographer

Jerry Bennett

Posts: 2223

Philadelphia, Pennsylvania, US

Wow! This thread just keeps getting better! Thank you!

May 06 09 02:51 pm Link

Photographer

Soliz Photography

Posts: 22

Palo Cedro, California, US

It seems to be a bit to sharp for my needs. I still continue to use these steps from
http://www.naturescapes.net/docs/index. … or-the-web

But lean more towards Ken's settings: http://www.kenrockwell.com/tech/photosh … pening.htm

May 06 09 04:48 pm Link

Photographer

Robert Randall

Posts: 13890

Chicago, Illinois, US

Soliz Photography wrote:
It seems to be a bit to sharp for my needs. I still continue to use these steps from
http://www.naturescapes.net/docs/index. … or-the-web

But lean more towards Ken's settings: http://www.kenrockwell.com/tech/photosh … pening.htm

I can't comment on the system you prefer because I haven't used any of it, but in terms of flexibility, I think you might be missing something with this threads procedure. In 20 years of computer imaging, I have never come across such a versatile and beautiful sharpening method.

May 06 09 06:21 pm Link

Photographer

Jerry Bennett

Posts: 2223

Philadelphia, Pennsylvania, US

Photons 2 Pixels Images wrote:
OK. New action set up that does what you're asking. At least I hope it does. smile

http://www.nunuvyer.biz/Photoshop/Frequency2.zip

This one just has the actions...the script is the same.

I can't get it to do anything. When I try to  press the play button I get the circle with a line through it. sad
Never mind, I figured it out. I'm new to running actions, been doing everything the old fashioned slow way!

May 06 09 08:07 pm Link

Photographer

Jerry Bennett

Posts: 2223

Philadelphia, Pennsylvania, US

ps- Shouldn't the high frequency layer be desaturated? I always do that with high pass so there's no color artifacts. Or am I just being Captain Obvious?

May 06 09 08:17 pm Link

Photographer

Jerry Bennett

Posts: 2223

Philadelphia, Pennsylvania, US

Damn! I just tried it and no, you have to leave the color in there. Hmmm....

May 06 09 08:23 pm Link

Photographer

Bobs Fine Art

Posts: 1371

Falls Church, Virginia, US

boy, oh boy, just stop wasting your time with all this nonsense, you'd spend alot less time just using the adjustment brush in lightroom.. you want -50 clarity, and +10 sharness to smooth skin, nuff said, and it gots alot faster!

May 06 09 09:17 pm Link

Photographer

Stacy Leigh

Posts: 3064

New York, New York, US

Great examples in this thread to illustrate your technique. I will have to read that several times to absorb 2/3 of it (apparently I am not that smart!).


Thank you tons for sharing the wealth...

smile
Stacy Leigh

May 06 09 10:17 pm Link

Photographer

Jerry Bennett

Posts: 2223

Philadelphia, Pennsylvania, US

Just playing around with it the first time and used it to even out skin a little with barely any effort:


https://farm4.static.flickr.com/3548/3508902767_8d15413543_o.jpg



This is one awesome tool!

May 06 09 10:22 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

Jerry Bennett wrote:
ps- Shouldn't the high frequency layer be desaturated? I always do that with high pass so there's no color artifacts. Or am I just being Captain Obvious?

If you're going to drop out the low frequency layer to use the HF data as a 'traditional' HP layer, then yes, you should take the color out of it to avoid shift.  But as you noticed, if you're just doing a frequency separation it's important to leave the color in as it is a part of that HF image data.

Great example with the skin work as well.  I suspect you'll be finding even more uses for the principles of this shortly.

isphotography wrote:
boy, oh boy, just stop wasting your time with all this nonsense, you'd spend alot less time just using the adjustment brush in lightroom.. you want -50 clarity, and +10 sharness to smooth skin, nuff said, and it gots alot faster!

Do what works for your workflow.  For 99% of the things I do, I loathe the adjustment brush in LR - too clunky for my purposes.  But like I said, I don't think anyone here will dispute that you should be doing what works for you.

Hey Bob - there's your Kelbyism lol.

Stacy - just ask if you have questions; it's good to have as many open minds contributing to this as we can get.

May 06 09 11:41 pm Link

Photographer

Ruben Vasquez

Posts: 3117

Las Vegas, Nevada, US

Robert Randall wrote:
Sean helped me work through the process earlier today, and I came up with a varient on full sized files. I have never seen sharpness like I achieved today, and I've been doing this for 20 years.

Would you care to share your "varient?"

May 07 09 04:46 am Link

Retoucher

James Minshall

Posts: 218

Bedford, Indiana, US

I've got a working 3 band equalizer.. but it could just as easily be 50, the process is the same.  Only thing is, I really don't see a use for it.  Its complicated using masks and clone layers, etc etc...  why mask/clone an image 5x when you can get a damn good result with just 1?

Do we really need to separate skin this much?  Are you guys thinking of other uses instead of skin?  Show me a practical use please! smile

May 07 09 10:40 am Link

Photographer

Soliz Photography

Posts: 22

Palo Cedro, California, US

http://psd.tutsplus.com/tutorials/photo … etouching/

This one does what you're all trying to do and then some wink

May 07 09 01:28 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

Soliz Photography wrote:
http://psd.tutsplus.com/tutorials/photo-effects-tutorials/super-fast-and-easy-facial-retouching/

This one does what you're all trying to do and then some wink

I try not to ask much of those who find my drivel, but please do try to read the thread next time before posting - the technique explained here by grahamsz does what that site is trying to do, just much better wink.

May 07 09 01:34 pm Link

Photographer

Sean Baker Photo

Posts: 8044

San Antonio, Texas, US

syd47421 wrote:
I've got a working 3 band equalizer.. but it could just as easily be 50, the process is the same.  Only thing is, I really don't see a use for it.  Its complicated using masks and clone layers, etc etc...  why mask/clone an image 5x when you can get a damn good result with just 1?

Do we really need to separate skin this much?  Are you guys thinking of other uses instead of skin?  Show me a practical use please! smile

The value wouldn't be for skin processing; rather for other creative applications (many of which I'm aware that I'm not right-brained enough to discover).  But you're right - for skin, there is really no point in taking it further than has been demonstrated.

May 07 09 01:36 pm Link

Retoucher

James Minshall

Posts: 218

Bedford, Indiana, US

Sean Baker wrote:

The value wouldn't be for skin processing; rather for other creative applications (many of which I'm aware that I'm not right-brained enough to discover).  But you're right - for skin, there is really no point in taking it further than has been demonstrated.

Its difficult sometimes to think outside of what my normal workflow requires.  Someone plant a seed in my brain.. please! smile

May 07 09 01:47 pm Link

Retoucher

James Minshall

Posts: 218

Bedford, Indiana, US

Syd's Graphic Equalizer

needs a couple of tweaks in the levels outputs, but the thing works.. and you can use the theory of it to make as many separations as you need.

May 07 09 01:55 pm Link

Photographer

Soliz Photography

Posts: 22

Palo Cedro, California, US

Alright I read into what this is all about, very interesting to say the least.
Now with the Action Script 'Frequency.atn'

The High Frequency layer and Low Frequency layer, which tools do I use on which layer? I started to do some test healing on the High and Low, noticed they both did close to the same thing.

Still going to need to back track to understand, but any other information would be great.

-NTS

May 07 09 03:01 pm Link

Photographer

AC Fashion Fusion

Posts: 134

Toronto, Ontario, Canada

Thanks..

May 07 09 03:06 pm Link