RetouchPRO

Go Back   RetouchPRO > Technique > Photo Retouching
Register Blogs FAQ Site Nav Search Today's Posts Mark Forums Read


Photo Retouching "Improving" photos, post-production, correction, etc.

Automatic Skin Smoothing.

Reply
 
Thread Tools
  #51  
Old 10-28-2010, 10:36 AM
oneredpanther's Avatar
oneredpanther oneredpanther is offline
Senior Member
 
Join Date: May 2010
Location: Copenhagen
Posts: 166
Re: Automatic Skin Smoothing.

If it could be modified to use Apply Image with an offset subtraction layer instead of the horrendous High Pass, then I'd be using it all the time.
Reply With Quote top
  #52  
Old 10-28-2010, 10:50 AM
Flashtones Flashtones is online now
Senior Member
 
Join Date: Apr 2009
Posts: 956
Re: Automatic Skin Smoothing.

Quote:
Originally Posted by oneredpanther View Post
If it could be modified to use Apply Image with an offset subtraction layer instead of the horrendous High Pass, then I'd be using it all the time.

I still don't get this. Asked it of Sean but without answer.

My understanding is, the problem with traditional highpass is a diminishment of upper highlights. Very important when you want to do a split that contains those tones and you wish to render them unchanged. However, when doing an inverted HP one is typically using it in selected areas that rarely contain near-pure highlights, and they do it with the intention of reducing contrast, ie, diminishing highlights.

So I don't see how it's problematic in most of it's intended uses. Are you (or anyone) able to show me in pictures, or is it merely a purists delight that is somehow significant even if unseeable?
Reply With Quote top
  #53  
Old 10-28-2010, 10:59 AM
oneredpanther's Avatar
oneredpanther oneredpanther is offline
Senior Member
 
Join Date: May 2010
Location: Copenhagen
Posts: 166
Re: Automatic Skin Smoothing.

It's very seeable if you know what you're looking for.

At 100% crop I can tell by looking in about 85% of cases whether high pass was used or not.

Offset subtraction leaves no trace, high pass has (for want of a better word) a residue.

When I get home I'll whip up an example.
Reply With Quote top
  #54  
Old 10-28-2010, 11:08 AM
Flashtones Flashtones is online now
Senior Member
 
Join Date: Apr 2009
Posts: 956
Re: Automatic Skin Smoothing.

panther, if you could demonstrate I'd be most grateful.

Just to be clear, I'm not talking about doing simply doing a split that renders the image virtually identical to the original but on two frequency layers (what I call a traditional split). I mean an Inverted HP used selectively to smooth transitions in contrast.
Reply With Quote top
  #55  
Old 10-28-2010, 11:45 AM
Boneappetit's Avatar
Boneappetit Boneappetit is offline
Senior Member
 
Join Date: May 2010
Location: USA
Posts: 762
Re: Automatic Skin Smoothing.

Quote:
Originally Posted by Flashtones View Post
I still don't get this. Asked it of Sean but without answer.

My understanding is, the problem with traditional highpass is a diminishment of upper highlights. Very important when you want to do a split that contains those tones and you wish to render them unchanged. However, when doing an inverted HP one is typically using it in selected areas that rarely contain near-pure highlights, and they do it with the intention of reducing contrast, ie, diminishing highlights.

So I don't see how it's problematic in most of it's intended uses. Are you (or anyone) able to show me in pictures, or is it merely a purists delight that is somehow significant even if unseeable?
Flexmanta (Omar Josef) has a video, where he explains the cons of using High Pass, but the video is in Spanish. Here in RP you can find a link to MM forum where Godmother (Natalia Taffarel) explains why High Pass sucks. Apply image is the way for the use of Split frequencies, and some other things too.

Just in case, here is a link to Omar's video, (spanish version) but maybe you can catch the idea just watching it.

http://www.ojodigital.com/foro/retoq...ecuencias.html

Here is the English version, but this one is a short version and he doesn't go into details of the high pass.
http://hiendworkshops.com/2010/08/18...y-separation-2

Last edited by Boneappetit; 10-28-2010 at 11:55 AM.
Reply With Quote top
  #56  
Old 10-28-2010, 12:19 PM
Flashtones Flashtones is online now
Senior Member
 
Join Date: Apr 2009
Posts: 956
Re: Automatic Skin Smoothing.

Boneappetit, Flex's English version only covers the traditional/conventional split, where I agree the Apply Image approach has benefit. Not sure what he covers in the Spanish version, I couldn't tell from a quick glance.

What I'm talking about is Inverted HighPass used selectively to "degrunge". It's a different beast with a different intended result.

Hopefully Flex, Natalia, Sean or others who've really done side by side comparisons for themselves will weigh in.

My suspicion this is like someone seeing evidence that Curves is better than ColorBalance for a certain application and then assuming that curves will yield a better result in every application and rule out the use of Color Balance, which is nonsense. (And I'm someone who always uses curves instead of Color Balance. But guess what, better retouchers than me use both. It's not about the tool, it's about knowing when, where and how to use it.)
Reply With Quote top
  #57  
Old 10-28-2010, 12:48 PM
mistermonday's Avatar
mistermonday mistermonday is offline
Moderator
Patron
 
Join Date: Dec 2005
Posts: 3,028
Re: Automatic Skin Smoothing.

Hi Flashtones, for frequency separation, there is a significant visible difference between accurate and regular high pass, in the shadows and highlights. However, you are correct that it may result in no visible difference on your inverted high pass / degrunge / contrast attenuation layer (or whatever else you wish to call it). Thats because you typically have a Hide All mask on that layer and you only selectively paint on it over high contrast transitions, and your painting is usually at fairly low opacity. The ultimate result there is a subtle reduction in contrast and detail via controlled blurring of a defined band of frequencies / edges of certain radii. I have never run a benchmark of the two but I suspect little or no visible difference.
Regards, Murray
Reply With Quote top
  #58  
Old 10-28-2010, 01:37 PM
Flashtones Flashtones is online now
Senior Member
 
Join Date: Apr 2009
Posts: 956
Re: Automatic Skin Smoothing.

Quote:
Originally Posted by mistermonday View Post
However, you are correct that it may result in no visible difference on your inverted high pass / degrunge / contrast attenuation layer (or whatever else you wish to call it).
You forgot bandstop.

But thanks for understanding the distinction I was getting at.
Reply With Quote top
  #59  
Old 10-28-2010, 01:48 PM
mistermonday's Avatar
mistermonday mistermonday is offline
Moderator
Patron
 
Join Date: Dec 2005
Posts: 3,028
Re: Automatic Skin Smoothing.

Flashtones, you forgot Bandpass , you are welcome!
Regards, Murray
Reply With Quote top
  #60  
Old 10-28-2010, 02:57 PM
ShadowLight ShadowLight is offline
Senior Member
 
Join Date: Oct 2009
Posts: 397
Cool Re: Automatic Skin Smoothing.

Quote:
Originally Posted by mistermonday View Post
Hi Flashtones, for frequency separation, there is a significant visible difference between accurate and regular high pass, in the shadows and highlights. However, you are correct that it may result in no visible difference on your inverted high pass / degrunge / contrast attenuation layer (or whatever else you wish to call it). Thats because you typically have a Hide All mask on that layer and you only selectively paint on it over high contrast transitions, and your painting is usually at fairly low opacity. The ultimate result there is a subtle reduction in contrast and detail via controlled blurring of a defined band of frequencies / edges of certain radii. I have never run a benchmark of the two but I suspect little or no visible difference.
Regards, Murray
It is not quite simple... actually Sean's technique produces nose in 8bit as it gives values between 127 and 128.

I agree 100% with using Apply when splitting and editing specific frequency (well... I have another way of doing it but that's not the point)... Since we want to eliminate the frequency band even if we truncate the signal down it will be better than to leave some noise (if you use Contrast -50).

Furthermore the "technique" where you use "Contrast -50" legacy, instead of Alpha manipulation, will produce luminosity shift. ... try it yourself if you don't believe me.

I had a huge argument with him about that, but apparently he didn't do the proper testing.
Inverse and Contrast -50, @ lin. light will not give you 127-128 but depending on the image will shift the whole base level.
if you use Contrast -50 as an "adjustment layer" then it does a proper calculation and you get 127/128. but since you are after something that is a solid mid-gray this technique fails as it splits the values between 127/128 leaving traces of noise in 8bit.
using Alpha does not have this problem.

When it comes to split-editing, I'm not aware of an easy way to dynamically adjust "apply image". (I have my own method that does it, but it takes about 20 layers and is very heavy for everyday use).

There are many ways to make the same thing.
If you have a way that will give better results, please show me a test and I'll see if I can put it in the program to make it better.

As it stands:
1) "Apply" is not dynamic (and not even Sean uses it for IHP/degrunge)
2) Using Contrast -50 (not as adjustment layer), produces noise in 8bit, and has luminosity shift (on some images the result can be 160/161)


I know it's hard to believe I know what I'm doing ... but I also did the proper tests.
I don't mind suggestions, but please don't say things that are not relevant, and take a few min and test your suggestion to see if the results are actually better in reality (something Sean forgot apparently to do with Contrast -50)

... enough about the splitting issue....

btw, I think there may be a way to take care of the runaway sliders after clicks... I'll see if it will work this weekend when I have some time.

one more thing... I didn't show it in any demo, but you don't need to make a smoothing for the whole layer.

1) make a selection (even with the lasso if you like) where you see some problems
2) click Create, - (this will make a copy only of the selected region and apply the technique only on it)
3) you can still use masks to adjust it later

... time for my coffee

Edit:
... don't forget to add some contrast after your corrections.
like in here: http://www.mobilefiles.ca/SkinSmooth...h-example2.psd

Last edited by ShadowLight; 10-28-2010 at 05:36 PM.
Reply With Quote top
Reply

  RetouchPRO > Technique > Photo Retouching


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Glamour retouch challenge here!!!! superkoax Photo Retouching 332 02-08-2011 09:52 PM
Automatic Skin Tone Retouching - Seeking feedback Pixarra Photo Retouching 15 11-09-2010 01:38 AM
Skin Smoothing Technique (Again) Novi Photo Retouching 10 10-16-2008 04:25 PM
need help doing pageant eyelashes and skin smoothing w/ photoshop cs sandylavallie Photo Retouching 3 05-06-2005 08:53 AM
skin smoothing action john_opitz Photo Retouching 10 08-21-2002 10:48 PM


All times are GMT -6. The time now is 11:38 PM.


Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
SEO by vBSEO ©2011, Crawlability, Inc.
Copyright © 2016 Doug Nelson. All Rights Reserved