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

Photographer

Vollman

Posts: 2

Nashville, Arkansas, US

After talking with Sean (cool guy) I decided to post my method for isolating certain tonal ranges for clean sharpening. Hope you guys think it's cool.

Tutorial:

http://farm4.static.flickr.com/3378/347 … 1385_b.jpg

Sorry for typos etc. I didn't have a lot of time to put it together... Feel free to write with any questions.

-Matt.

Apr 25 09 04:04 am Link

Retoucher

Kevin_Connery

Posts: 3307

Fullerton, California, US

Vollman wrote:
After talking with Sean (cool guy) I decided to post my method for isolating certain tonal ranges for clean sharpening. Hope you guys think it's cool.

If you're only limiting the highlights and the shadows, have you considered just using the Blend If sliders to exclude the brightest and darkest areas?

Apr 25 09 04:53 am Link

Photographer

Vollman

Posts: 2

Nashville, Arkansas, US

You could do that. I dig seeing the affected tones with the threshold tool though.

Also, I can't really find a good way to feather the mask that the Blend if sliders create.

-Matt.

Apr 25 09 05:37 am Link

Photographer

MEK Photography

Posts: 6571

Westminster, Maryland, US

Vollman wrote:
You could do that. I dig seeing the affected tones with the threshold tool though.

Also, I can't really find a good way to feather the mask that the Blend if sliders create.

-Matt.

I haven't tried this part yet, so I'm just tossing in a guess, but with CS4 and the masks pallet, could you not feather the mask created from Blend If there??

Apr 25 09 06:20 am Link

Retoucher

Kevin_Connery

Posts: 3307

Fullerton, California, US

Vollman wrote:
You could do that. I dig seeing the affected tones with the threshold tool though.

Also, I can't really find a good way to feather the mask that the Blend if sliders create.

Hoid the Option or Alt key down when dragging the slider; that splits the triangle into a pair which lets you control how the limit kicks in.

If you need to manipulate the controls manually, obviously you can use the masks, but if it's just for range limiting, Blend If is faster.

Apr 25 09 07:40 am Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean,

I'm curious why you move this to a separate document in your action to accomplish this, then move the results back. Can this also be done with the same results by copying the background layer and performing the rest on that instead? It seems to work fine for me that way. I did it both ways and the results appear the same to my eye. Unless I'm missing something.

Mike

Apr 26 09 05:58 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

Photons 2 Pixels Images wrote:
Sean,

I'm curious why you move this to a separate document in your action to accomplish this, then move the results back. Can this also be done with the same results by copying the background layer and performing the rest on that instead? It seems to work fine for me that way. I did it both ways and the results appear the same to my eye. Unless I'm missing something.

Mike

Mike,

Good question.  I wanted to create one action which everyone could use, which operated within my limited action-creating abilities, and which reliably separated the image with high accuracy.  I couldn't figure out how to detect the bit mode from within the action (nor how to branch the action's behavior on this basis), and I didn't think it would be ethical to convert everyone's documents to 16bit.  Consequently, I went with creating a new copy in another 16bit window where I could perform the separation accurately and the pieces could be brought back over - kudos to Biggs for having demonstrated how to do this automatically.  In the end, one action reliably creates the most accurate separation we know how to do.

Hope that makes sense; I need some coffee.

Apr 26 09 06:12 am Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean Baker wrote:

Mike,

Good question.  I wanted to create one action which everyone could use, which operated within my limited action-creating abilities, and which reliably separated the image with high accuracy.  I couldn't figure out how to detect the bit mode from within the action (nor how to branch the action's behavior on this basis), and I didn't think it would be ethical to convert everyone's documents to 16bit.  Consequently, I went with creating a new copy in another 16bit window where I could perform the separation accurately and the pieces could be brought back over - kudos to Biggs for having demonstrated how to do this automatically.  In the end, one action reliably creates the most accurate separation we know how to do.

Hope that makes sense; I need some coffee.

Sean,

I had a feeling that might be the reason since there was a convert mode step in the action. There is a conditional mode change feature that can be added into the action set, though I have not found a reason to use it yet so I'm not real familiar with it but might be another way of doing this.

I'm going to play around with this and perhaps come up with a more involved and interactive action/script combination. Is there a specific formula used for the GB radius based on image size? Or is it tune to taste? That's something I can incorporate into a script.

On another note, something interesting to try:

When you make the Low Frequency layer, convert it to a Smart Object before applying GB to it. Then after the High Frequency layer is created, go back and adjust the radius on the GB layer to see the effects. Try with the initial GB setting higher than normal and pull it back after, then with the initial GB setting normal and push it higher after.

Neat stuff.

Thanks again for this very useful technique.

Mike

Apr 26 09 06:50 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

Photons 2 Pixels Images wrote:
I'm going to play around with this and perhaps come up with a more involved and interactive action/script combination. Is there a specific formula used for the GB radius based on image size? Or is it tune to taste? That's something I can incorporate into a script.

This technique is always designed to be tuned to taste, IMO.  Some folks on the internet will espouse that there are ideal radii for all sharpening, and others who believe that inversions of the same can be used for skin smoothing - always using the same radii (a very popular and very profitable action set is based on just this premise).  It's all bull; only images from the same body in the same conditions of the same subject need the same tuning.  Everything else needs to be adjusted.

The effect you describe is also quite interesting.  I'm going to have to play around with it some, but I think I'm going to run up against some blend mode issues for what I'd really like to do with that.  Hopefully Adobe are going to run with the concepts behind smart objects + group blend modes and by CS5 allow us to do some very cool things very easily.

Can't wait to see what you come up with for a script / action combo.

Sean

Apr 26 09 07:21 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Thanks for posting this, it's an amazingly effective way to edit skin. The theory seems pretty straightforward, but i'm not sure i could have figured out the steps in photoshop.

I'm having a lot of fun editing skin with this. I can take out blotches and uneven tone on the low frequency layer, and then take out pimples and large pores on the high frequency layer.

Apr 27 09 11:23 pm Link

Photographer

Sean Peragine

Posts: 75

Boston, Massachusetts, US

another 'thank you' for posting this from someone who loves (or loved?) the good ole high pass filter.  i finally got around to trying out this technique tonight and after some initial confusion i have it working.  although i would imagine that mastering the application of this could take a long time and a lot of experimentation, i can already see really impressive results in sharpening with a large amount of control.  i have to admit, im not exactly clear on the nuances of how it works, though.  if someone wants to point me in the direction of some good resources it would be appreciated.

Apr 28 09 07:54 pm Link

Photographer

Robert Randall

Posts: 13873

Chicago, Illinois, US

grahamsz wrote:
Thanks for posting this, it's an amazingly effective way to edit skin. The theory seems pretty straightforward, but i'm not sure i could have figured out the steps in photoshop.

I'm having a lot of fun editing skin with this. I can take out blotches and uneven tone on the low frequency layer, and then take out pimples and large pores on the high frequency layer.

Can you elaborate on your process a bit further.

Apr 28 09 08:40 pm Link

Photographer

STUDIO2401

Posts: 1335

Manila, National Capital Region, Philippines

A slice of oblivion wrote:
On a website full of Gaussian blurred faces, threads like these never get the praise they deserve.

Thanks for writing this Sean.

Agree !

Apr 28 09 09:33 pm Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Robert Randall wrote:
Can you elaborate on your process a bit further.

Sure. I took a 12MP image, and have been playing with doing a 4 to 8 pixel radius gaussian blur for the low frequency layer and then creating the high frequency layer as sean described.

Then i work on the low frequency layer and try to smooth out blotchy skin and even out the tone of the skin. Not sure how best to do this, but usually between the clone tool and paint brush with a very soft brush and 10% opacity I can smooth it out.

Then i switch on the high frequency layer and use the clone tool on 100% opacity to clean up pimples, pores and such. It also works well on cracked lips. Keeping it on 100% avoids ending up blurring the texture of the skin, although I occasionally clone on a lower opacity if i need to soften creases without removing them altogether.

Once that's done, i use the usual dodge and burn adjustment layers to recreate any highlights and shadows that were lost in the adjustment of the low frequency layer.

Here's something i did last night, cropped it tight so hopefully the model will forgive me smile

https://graha.ms/mm/spatial_frequency_separation_skin.jpg

It's still a lot of work, but since I can clone a clean patch of high frequency skin onto an area with a completely different low frequency tonality gives me a lot more options.

Apr 28 09 10:20 pm Link

Photographer

J C ModeFotografie

Posts: 14718

Los Angeles, California, US

Paul Dempsey wrote:
sounds to me like you're taking the "art" out of photography and turning it into a math equation....ther'e s abook you should check out ..."Drawing on the Right Side of the Brain"  It's designed to help people get away from the over analytical thinking and rely more on the creative side (the right side ) of your brain...plus, it's just way more fun to look at photography as an art form rather than a math problem.

I second that!

Not that the process the OP describes is necessarily a bad thing - it does seem to sell well in the commercial market.

Ars Gratia Artis

Apr 29 09 03:56 am Link

Photographer

Jeremy I

Posts: 2201

Charlotte, North Carolina, US

Sean Baker wrote:
Bottom Line Up Front: High Pass is an inaccurate spatial-frequency based separation technique.  The steps below and the actions provided effect a much more accurate result than can be accomplished otherwise.

Background: Ultimately, this is a spinoff of Mr. Connery's thread on deconvolution as an image sharpening technique here, wherein I set forth my method for finer control of such while both remaining within Photoshop and not spending additional $$ on plugins.  Key to its success is the separation of high and low spatial frequency image data and the fine-tuning of each.

Separation of spatial frequency data has a number of applications in image editing.  Whether for the oft-suggested "High Pass Sharpening", to recover detail lost in an OOF / moving image, or to enhance local contrast throughout the image, the accurate separation of frequency data is relevant for anyone who is a stickler for image quality.

Findings / Technique: In my own experimentation, I've found that HP gives differences as high as 2670/32k per pixel when separating high and low frequency information.  The proposed alternative technique for working with 8bit image data is as follows:

  1.) Start with two copies of the image to be separated.
  2.) Working on the bottom copy, run the gaussian blur filter at the intended pixel frequency (same as you would input into the HP filter).
  3.) Selecting the top copy, choose Apply Image from the Image menu at top (Shift+Ctrl+A on PC; Cmd+Shift+A for Mac).
  4.) In the Layer dialog, select the bottom layer which you blurred in step 2.
  5.) In the Blending dialog, choose Subtract.
  6.) Enter '2' into the Scale box, and '128' for the Offset.
  7.) Preserve Transparency, Mask, and Invert should not be checked.
  8.) Choose OK.
  9.) Your top layer will now look much as a HP result, albeit a bit flatter.  Set the Blend Mode to Linear Light.  Opacity should remain at 100%.

Working in 8bit mode, this will give you an accurate frequency separation to about 115/32k.

Working in 16bit mode, however, we encounter a problem with the offset factor applied in Step 6 above (50% grey in a 16bit system cannot be expressed by an 8bit number).  This can be overcome using the alternate technique which follows:

  1.) Start with two copies of the image to be separated.
  2.) Working on the bottom copy, run the gaussian blur filter at the intended pixel frequency (same as you would input into the HP filter).
  3.) Selecting the top copy, choose Apply Image from the Image menu at top (Shift+Ctrl+A on PC; Cmd+Shift+A for Mac).
  4.) In the Layer dialog, select the bottom layer which you blurred in step 2.
  5.) In the Blending dialog, choose Add.
  6.) Check the Invert box.
  6.) Enter '2' into the Scale box, and '0' for the Offset.
  7.) Preserve Transparency and Mask should not be checked.
  8.) Choose OK.
  9.) Your top layer will now look much as a HP result, albeit a bit flatter.  Set the Blend Mode to Linear Light.  Opacity should remain at 100%.

Conclusion: Accuracy for this technique is fantastic, with a maximum difference from the original of 1/32767.  Unfortunately, this technique is not valid while working in 8bit mode.  I can't explain the phenomenon, but the shadows and midtones get mucked up by about 1-2/256.  Perhaps those with a bit more insight into PS's inner workings can offer thoughts on why this is.

Afterword: Walking through all these steps can get a bit tiring, and I've created an action which I'll link below to run you through them.  It works in a separate window utilizing the 16bit technique above to retain an accurate separation, but, as I'm not an Actions Wizard, it will not move the data back into your original document for you.  If someone more adept than I would like to add that in I will gladly update this post; otherwise it remains to the user to bring both or either layer into your document for use.

What will you do with it?
  - "High Pass Sharpen"?
  - Create a high-spatial-frequency mask?
  - Create a low-spatial-frequency mask?
  - Something else?  Post your results and findings here.  It's time to up the ante on sharing what we find.

Action

Please Note: As outlined above, this technique is designed to end with the image on screen looking exactly the same as it did when you started.  The difference is that your image data is now separated onto two separate layers (separated by size of detail - spatial frequency), allowing you to edit them independently.  If you want to apply "high pass sharpening" with this technique, simply disable the low frequency ('blur') layer and it will appear immediately.

This is one of the more interesting posts I have seen in this joint.

Apr 29 09 04:14 am Link

Photographer

JeF Briguet

Posts: 119

Chamoson, Valais, Switzerland

grahamsz wrote:
It's still a lot of work, but since I can clone a clean patch of high frequency skin onto an area with a completely different low frequency tonality gives me a lot more options.

Wow you just opened a Pandora's Box. Why i didn't think about this before!?!?

Thank you, thank you, thank you [you too Sean].

Edit: Seriously this is huuuge (time saver). Thank you again (for the emphasis).

Apr 29 09 04:31 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

Sean Peragine wrote:
if someone wants to point me in the direction of some good resources it would be appreciated.

Do you want to know why it has the sharpening effect it does, or where to go for more information about spatial frequencies?  Or both? smile

grahamsz wrote:
I'm having a lot of fun editing skin with this. I can take out blotches and uneven tone on the low frequency layer, and then take out pimples and large pores on the high frequency layer.

Thank you for this.  I'd been using the healing brush on the high frequency level, but for no (good) reason was ignoring the lower frequency data.  Much appreciated.

Apr 29 09 06:20 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

JeF Briguet wrote:
Wow you just opened a Pandora's Box. Why i didn't think about this before!?!?

That's exactly how I felt when i saw seans write-up. In many ways it's bloody obvious (at least for someone who's done a few introductory digital signal processing classes), but I never thought to actually figure out how to do it.

Sean Baker wrote:
Thank you for this.  I'd been using the healing brush on the high frequency level, but for no (good) reason was ignoring the lower frequency data.  Much appreciated.

It's pretty neat because you can spot things like where the skin loses it's pink tone and turns gray right before heading into a highlight. That's the sort of thing that will bite you later if you try to use the burn tool on it, but you can subtly recolor it for a smoother effect.

The problem with editing the low frequency layer is that you can end up inserting data that's a higher frequency than is needed, and since the Gaussian blur isn't a clean low pass filter you cant just reblur without further blurring the existing colors.

I think there's maybe a solution in painting your changes to the low frequency layer into a clean layer, then merging those changes back with the original image and reapplying the same radius Gaussian blur as you did in the first step.

Apr 29 09 07:22 am Link

Photographer

Robert Randall

Posts: 13873

Chicago, Illinois, US

J C ModeFotografie wrote:
I second that!

Not that the process the OP describes is necessarily a bad thing - it does seem to sell well in the commercial market.

Ars Gratia Artis

Edit... you aren't worth the brig.

Apr 29 09 07:23 am Link

Photographer

Peter Claver

Posts: 27130

Toronto, Ontario, Canada

Robert Randall wrote:
Edit... you aren't worth the brig.

I wonder if the "making art into math" high horse folks have *any* clue whatsoever about how image processing software is written.

This technique is positively simple in comparison to some of the more esoteric stuff.  That's not a slight to the technique.. I find it quite clever.. but there's some pretty out-there stuff in the image processing field.  Hell.. bayer de-mosaicking is a huge field of research all on its own.

Apr 29 09 07:28 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Peter Claver wrote:
I wonder if the "making art into math" high horse folks have *any* clue whatsoever about how image processing software is written.

Or perhaps how exposure, focus and depth of field work. It seems very arbitrary to draw a line in the sand and say that some math is good, but new math is scary.

Apr 29 09 07:30 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

grahamsz wrote:
I think there's maybe a solution in painting your changes to the low frequency layer into a clean layer, then merging those changes back with the original image and reapplying the same radius Gaussian blur as you did in the first step.

How about doing your edits into a Smart Object layer with a GB filter applied to it?

Apr 29 09 07:32 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Sean Baker wrote:
How about doing your edits into a Smart Object layer with a GB filter applied to it?

I didn't know you could do that (been retouching faces for all of a week). Sounds like it should work though

Apr 29 09 07:33 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

grahamsz wrote:

I didn't know you could do that (been retouching faces for all of a week). Sounds like it should work though

Bollocks.  You're right.  It can be done, but isn't going to be worth the hassle.

Apr 29 09 07:36 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Sean Baker wrote:
Bollocks.  You're right.  It can be done, but isn't going to be worth the hassle.

Why isn't it worth it? Too unartistic wink

Apr 29 09 07:39 am Link

Photographer

Sean Peragine

Posts: 75

Boston, Massachusetts, US

Sean Baker wrote:

Sean Peragine wrote:
if someone wants to point me in the direction of some good resources it would be appreciated.

Do you want to know why it has the sharpening effect it does, or where to go for more information about spatial frequencies?  Or both? smile


Thank you for this.  I'd been using the healing brush on the high frequency level, but for no (good) reason was ignoring the lower frequency data.  Much appreciated.

both please!

Apr 29 09 07:57 am Link

Photographer

JeF Briguet

Posts: 119

Chamoson, Valais, Switzerland

grahamsz wrote:
The problem with editing the low frequency layer is that you can end up inserting data that's a higher frequency than is needed, and since the Gaussian blur isn't a clean low pass filter you cant just reblur without further blurring the existing colors.

I think there's maybe a solution in painting your changes to the low frequency layer into a clean layer, then merging those changes back with the original image and reapplying the same radius Gaussian blur as you did in the first step.

If you stay in 16 bit there is less problem with the low frequency layer. But yea you can't push it to far. I'm still searching for a work around.

Anyway, working on the high frequency layer is a real time saver (i still can't believe it)!

Apr 29 09 08:10 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Sean Peragine wrote:
both please!

It's probably easiest to visualize in a 1-dimensional context.

The fourier theorem states that you can break apart any shape of wave into a sum of perfectly smooth Sine waves. Here's one of the common textbook images showing a square waveform can be approximated by adding up three sine waves.

https://www.files.chem.vt.edu/chem-ed/data/graphics/fourier-waves.gif

The means that any audio waveform can be represented as a summation of Sine waves, and if you can break it down into it's components then you can process them differently.

When you hit the bass boost button on your stereo, it uses a Low Pass filter to create a signal with all the sounds < 30Hz (that's a guess, could be wrong) and a High Pass filter with the rest. It then amplifies the low frequency signal and recombines it with the unaltered high frequency signal.

What we are doing to the image is very similar. The gaussian blurred layer is the low frequency, "bass" signal - it alone is enough to recognize the image (just like you can usually recognize music when it's only coming through a subwoofer) but it lacks crispness and detail. Then the subtracted layer contains the missing detail, but none of the low frequency stuff.

To go on a slight tangent - here's an optical illusion that exploits the same thing. It's the high frequency layer of einstein with a low frequency layer of marilyn. (edit - try walking away from your computer while looking at this, your eyes have a built in low pass filter which will kick in as the image becomes smaller in your field of view)

https://img228.imageshack.us/img228/289/einsteinormonroe.jpg

That might help, or it might just confuse you.

Apr 29 09 08:10 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

grahamsz wrote:
Why isn't it worth it? Too unartistic wink

lol Something like that.  Too much of an administrative headache.  I use smart objects rarely and forget that to change the actual data requires reopening it in a separate document, which in order to make things easy would also mean copying additional layers over to see what you're doing, etc.  It could work, sure, but I don't think the juice would be worth the squeeze as they say.

What might do better is a keystroke-associated action to duplicate the working layer, disable the copy, apply a blur, pause for viewing, delete the copy, and restore the original.  Not elegant, but quick & functional.

Fantastic explanation of spatial frequency.

Sean Peragine wrote:
both please!

And to dovetail what grahamsz laid out above, the sharpening comes about because you're increasing the contrast of the fine detail.  Put into the context of the diagram, you're refining the image data to more closely approximate the square wave and the 'perfect' contrast it represents.

JeF Briguet wrote:
If you stay in 16 bit there is less problem with the low frequency layer. But yea you can't push it to far. I'm still searching for a work around.

Anyway, working on the high frequency layer is a real time saver (i still can't believe it)!

Have you tried a generalized or selection-based surface blur for this purpose?

Apr 29 09 08:51 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

Sean Baker wrote:
Fantastic explanation of spatial frequency

Thanks - didn't mean to step on your toes there, but i'd been thinking about it in some detail and wanted to write that up while it was in my mind.

Sean Baker wrote:
Have you tried a generalized or selection-based surface blur for this purpose?

I've got fairly good results just using the waterdrop blur tool to smooth out the low frequency layer.

Apr 29 09 09:20 am Link

Photographer

Giacomo Cirrincioni

Posts: 22229

New York, New York, US

This thread makes me happy.

Apr 29 09 09:53 am Link

Photographer

Eduardo Frances

Posts: 3227

Barcelona, Catalonia, Spain

Paul Dempsey wrote:
sounds to me like you're taking the "art" out of photography and turning it into a math equation....ther'e s abook you should check out ..."Drawing on the Right Side of the Brain"  It's designed to help people get away from the over analytical thinking and rely more on the creative side (the right side ) of your brain...plus, it's just way more fun to look at photography as an art form rather than a math problem.

neutral... Do you realize the OP has posted a post processing technique to manipulate sharpness neutral... And you know post means after, like after the photoshoot (read: after concept, composition, model direction, when the photoshoot has finished, etc.) If the OP has created a process analyzing the image with a mathematical formula is because you will get the best results out of the PP this way... which has NOTHING to do with what you are babbling about hmm.

Apr 29 09 10:15 am Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean,

I'm working on this combo script/action right now. Since I'm not really knowledgeable with retouching in general I'm going to have questions on how best to write this up for the most usability by the most people.

I'll let you know where I am now.

1. Determines bit depth and converts to 16 bit if necessary.
2. Sets up the High and Low layers as in your action, converting the Low to a smart object.
3. Puts High and Low into a Group named Frequency. I do this so I can toggle visible or invisible to see the effects. I find this works better as my eyes adjust to what's there and I may feel it isn't enough when it's actually too much.

I'm now starting on your Export sharpening action set. I think I'm going to put those as smart objects also so they can be adjusted afterward. However, within the script I'm going to set the initial values to your formulas as given in the other thread for GB and SS. I feel this would give a good starting point for minimal adjustment in the smart objects after.

I'm debating whether or not to combine the Frequency actions with the Export actions with branching determined by a dialog. When you start the action it will ask whether it's capture or output sharpening and go from there. I can also have a dialog ask whether the user wants to accept default settings for the GB and SS (but can still adjust later in the smart objects) or set them on the fly.

Your input.....or anyone else's input....would be greatly appreciated. I want to make this user friendly for as many as possible.

Apr 29 09 10:21 am Link

Photographer

grahamsz

Posts: 1039

Boulder, Colorado, US

I think my preferred layer layout would be something like

- Original / Hidden - Handy for comparison and getting the light right
- 50% Gray / Overlay - for dodging and burning
- High Frequency Layer / Linear Light
- Empty Low Frequency adjustment layer / Blend Normal - It's probably better than directly adjusting the low frequency layer
- Low Frequency Layer / Blend Normal

A second action which does the following

- Deletes the low frequency layer
- Merges the original and low frequency adjustment layers to make a new low frequency layer (while preserving both input layers)
- Applies the same gaussian blur to the newly created layer

Just thinking out loud though

Apr 29 09 10:49 am Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

grahamsz wrote:
I think my preferred layer layout would be something like

- Original / Hidden - Handy for comparison and getting the light right
- 50% Gray / Overlay - for dodging and burning
- High Frequency Layer / Linear Light
- Empty Low Frequency adjustment layer / Blend Normal - It's probably better than directly adjusting the low frequency layer
- Low Frequency Layer / Blend Normal

A second action which does the following

- Deletes the low frequency layer
- Merges the original and low frequency adjustment layers to make a new low frequency layer (while preserving both input layers)
- Applies the same gaussian blur to the newly created layer

Just thinking out loud though

Great. The more out loud thinking I see here the more I can put into this action/script.

Thanks.

Apr 29 09 10:56 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

grahamsz wrote:
Thanks - didn't mean to step on your toes there, but i'd been thinking about it in some detail and wanted to write that up while it was in my mind.

Not at all; it's a better explanation than I'm likely to have given.

Photons 2 Pixels Images wrote:
I'm now starting on your Export sharpening action set. I think I'm going to put those as smart objects also so they can be adjusted afterward. However, within the script I'm going to set the initial values to your formulas as given in the other thread for GB and SS. I feel this would give a good starting point for minimal adjustment in the smart objects after.

This makes sense to me so long as the dialogs still come up for tuning to the image at hand.

Photons 2 Pixels Images wrote:
I'm debating whether or not to combine the Frequency actions with the Export actions with branching determined by a dialog.

My vote would be for separate actions, actually.  Just separating the data is a different ballgame and has different reasons for it than trying to sharpen it.  This would allow - at least for those more efficiency oriented / capable than myself - for mapping of each to a keystroke and fit into more folks workflows.  But obviously there are more educated opinions reading this thread than mine smile.

FWIW, your ability to merge all this is impressive.

grahamsz wrote:
- Original / Hidden - Handy for comparison and getting the light right
- 50% Gray / Overlay - for dodging and burning
- High Frequency Layer / Linear Light
- Empty Low Frequency adjustment layer / Blend Normal - It's probably better than directly adjusting the low frequency layer
- Low Frequency Layer / Blend Normal

This is perfect, IMO.

Apr 29 09 02:01 pm Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean Baker wrote:

grahamsz wrote:
Thanks - didn't mean to step on your toes there, but i'd been thinking about it in some detail and wanted to write that up while it was in my mind.

Not at all; it's a better explanation than I'm likely to have given.

Photons 2 Pixels Images wrote:
I'm now starting on your Export sharpening action set. I think I'm going to put those as smart objects also so they can be adjusted afterward. However, within the script I'm going to set the initial values to your formulas as given in the other thread for GB and SS. I feel this would give a good starting point for minimal adjustment in the smart objects after.

This makes sense to me so long as the dialogs still come up for tuning to the image at hand.

Photons 2 Pixels Images wrote:
I'm debating whether or not to combine the Frequency actions with the Export actions with branching determined by a dialog.

My vote would be for separate actions, actually.  Just separating the data is a different ballgame and has different reasons for it than trying to sharpen it.  This would allow - at least for those more efficiency oriented / capable than myself - for mapping of each to a keystroke and fit into more folks workflows.  But obviously there are more educated opinions reading this thread than mine smile.

FWIW, your ability to merge all this is impressive.


This is perfect, IMO.

OK. I'll work on that. I had a bit of trouble with trying to use smart objects and inverting but I may have a solution for that. We'll see how that goes.

So I'll separate these into 2 distinct action sets. Then let you all try them and see how you like them and let me know what else to put in with it.

Apr 29 09 04:32 pm Link

Photographer

Photons 2 Pixels Images

Posts: 17011

Berwick, Pennsylvania, US

Sean Baker wrote:

grahamsz wrote:
Thanks - didn't mean to step on your toes there, but i'd been thinking about it in some detail and wanted to write that up while it was in my mind.

Not at all; it's a better explanation than I'm likely to have given.


This makes sense to me so long as the dialogs still come up for tuning to the image at hand.


My vote would be for separate actions, actually.  Just separating the data is a different ballgame and has different reasons for it than trying to sharpen it.  This would allow - at least for those more efficiency oriented / capable than myself - for mapping of each to a keystroke and fit into more folks workflows.  But obviously there are more educated opinions reading this thread than mine smile.

FWIW, your ability to merge all this is impressive.


This is perfect, IMO.

grahamsz wrote:
- Original / Hidden - Handy for comparison and getting the light right
- 50% Gray / Overlay - for dodging and burning
- High Frequency Layer / Linear Light
- Empty Low Frequency adjustment layer / Blend Normal - It's probably better than directly adjusting the low frequency layer
- Low Frequency Layer / Blend Normal

For this Empty LF adjustment layer, are you talking a curves layer?

I'm also guessing this is the order you want them in from top to bottom.

Apr 30 09 05:48 am Link

Photographer

Sean Baker Photo

Posts: 8044

Fort Knox, Kentucky, US

Photons 2 Pixels Images wrote:
I'm also guessing this is the order you want them in from top to bottom.

My preference would be for the layers in that order.  And if I might speculate, I think that grahamsz was suggesting a truly blank, normal layer.  One in which cloning / healing / real painting could occur.  But I've been wrong before - just ask my wife lol.

Apr 30 09 06:29 am Link