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.

L*a*b* question: why can't it be simulated in RGB?

Reply
 
Thread Tools
  #51  
Old 09-11-2010, 12:50 PM
Flashtones Flashtones is offline
Senior Member
 
Join Date: Apr 2009
Posts: 956
Re: L*a*b* question: why can't it be simulated in

Mistermonday, I'm confused. I thought the advantage of the asymetric sep was to keep color off the texture layer precisely to avoid such color smearing.

Chain, if you do make new actions I'd love it if you'd share them, as I already found your old ones useful.
Reply With Quote top
  #52  
Old 09-11-2010, 02:14 PM
Chain's Avatar
Chain Chain is offline
Senior Member
 
Join Date: May 2009
Location: Oslo, Norway
Posts: 551
Re: L*a*b* question: why can't it be simulated in

I'm redoing my Frequency Separation action sets now. Just takes a bit to alter them so they all work perfectly with layers and multiple runs while watching Mythbusters at the same time. ;p

I'll upload them when done

Last edited by Chain; 09-11-2010 at 02:26 PM.
Reply With Quote top
  #53  
Old 09-11-2010, 02:24 PM
Der_W's Avatar
Der_W Der_W is offline
Senior Member
 
Join Date: Jul 2009
Location: Germany
Posts: 558
Re: L*a*b* question: why can't it be simulated in

Quote:
Originally Posted by Chain View Post
Are you me from the future?

[...]

Time for me to make my own updated frequency separation actions. Hehe.
I'm pretty sure, that it'd be rather the other way around. If not, you'd have to be like Benjamin Button probably (guessing you're older than 19 ;-)?).

Looking forward to your updated actions :-).
It's always great to see how this develops as it just seems to have unlimited uses!
Reply With Quote top
  #54  
Old 09-11-2010, 05:21 PM
Chain's Avatar
Chain Chain is offline
Senior Member
 
Join Date: May 2009
Location: Oslo, Norway
Posts: 551
Thumbs up Re: L*a*b* question: why can't it be simulated in

Phew. That took longer than expected, I had to edit them a bit to make the actions stop fighting - and then changed my mind a couple of times about how to do things... :P

I made the set work on both 8 and 16-bit images.
It includes the following actions:
  • Frequency separation - plain separation using Gaussian Blur
    (replace Gaussian Blur with something else if you prefer)
  • Frequency separation (Luminosity) - Keeps the color information out of the separation (as discussed in this thread)
  • Band stop - A band stop layer with a mask.
  • Band stop (Smart Filter) - A band stop layer with a mask, but made using Smart Filters sacrificing some accuracy for convenience.

8 and 16-bit only set are included as well. They will be faster and the results are the same on 16 bit images. The plain 8-bit versions are a bit less accurate (but have proven accurate enough for my everyday needs).

Probably some final polishing to be done, but you guys are all clever enough to do that yourself if you see something you don't like.

Edit:
Here's my action-set for sharpening as well.

Edit:
It now includes separate 16 and 8-bit action sets. They are noticeably quicker than the version that works on both bit depths.
Attached Images
File Type: png preview.png (6.3 KB, 38 views)
Attached Files
File Type: zip Frequency separations by Chain (8 and 16 bit).zip (16.9 KB, 48 views)

Last edited by Chain; 09-21-2010 at 11:47 AM. Reason: more actions!
Reply With Quote top
  #55  
Old 09-11-2010, 09:57 PM
mistermonday's Avatar
mistermonday mistermonday is offline
Moderator
Patron
 
Join Date: Dec 2005
Posts: 3,028
Re: L*a*b* question: why can't it be simulated in

Hi Flashtones, when you heal in the 3 different ways there are different processes going on.
When you do a normal heal on an image (sample attached) and you happen to heal along an edge with different colors on each side, the healing brush in trying to blend the texture into the destination makes a mess by blurring together the colors on both sides of the edge. To make it worse you will also end up with a color and brightness component of the source.
The standard frequency separation does not blur the destination colors along the edge but the source brings color with it and sometimes the destination will blend in some source color - but that depend s on the on a lot of things.
The asynch freq separation has no color and using the healing brush does not blur. If you look closely it does not affect colors on either side of the edge but can add edge contrast depending on the radius and what the underlying colors are.
I suggest you use Chain's actions to create both types of separations on the same test image and compare the two results.
Regards, Murray
Attached Images
File Type: jpg Normal Heal.jpg (155.3 KB, 16 views)
Reply With Quote top
  #56  
Old 09-11-2010, 09:58 PM
mistermonday's Avatar
mistermonday mistermonday is offline
Moderator
Patron
 
Join Date: Dec 2005
Posts: 3,028
Re: L*a*b* question: why can't it be simulated in

Chain, that was kind of you to create and share your actions, thanks.
Regards, Murray
Reply With Quote top
  #57  
Old 09-11-2010, 10:27 PM
Flashtones Flashtones is offline
Senior Member
 
Join Date: Apr 2009
Posts: 956
Re: L*a*b* question: why can't it be simulated in

Quote:
Originally Posted by mistermonday View Post
The asynch freq separation has no color and using the healing brush does not blur. If you look closely it does not affect colors on either side of the edge but can add edge contrast depending on the radius and what the underlying colors are.

Sorry to be a pest. Your words make sense to me, I just don't get how they jive with this image you posted of an asymmetric split and heal:

http://www.retouchpro.com/forums/att...1&d=1284221905

You say above it shouldn't affect colors on either edge, but then there is so much blue inside his ear. Was the wrong image attached or am I missing something obvious?

Edit: just to be sure I understand what you did there: after the split, on the HF layer you healed mane fur around the edge of his ear (with a selection?) and the result yielded was a sharp edge with texture but blue spill?

Last edited by Flashtones; 09-11-2010 at 10:40 PM.
Reply With Quote top
  #58  
Old 09-11-2010, 10:49 PM
mistermonday's Avatar
mistermonday mistermonday is offline
Moderator
Patron
 
Join Date: Dec 2005
Posts: 3,028
Re: L*a*b* question: why can't it be simulated in

Flashtones, sorry for not making it clear. If you zoom in close on the original unmodified image, you will see that the strong blue edge is actually there. It could have resulted from over shapening or a bit of chromatic aberation from the lense. If you look at the asymetric separation, you will see that yes, the blue edge is intensified, but aside from that thin edge, the sky, where the texture has been healed over is still clean blue with texture, the area of ear on the other side of the blue edge retains the same brown color but with texture. The only difference between the symetric and the asymetric is that the blue edge in the center is more contrasted and appears more visible. That's not blurring especially not of the kind that results from a normal heal on an unseparated image. I hope this helps. I think if you create the separations and do you own testing you will get a good feel for the differences and virtues of each type.
Regards, Murray
Reply With Quote top
  #59  
Old 09-11-2010, 11:21 PM
Flashtones Flashtones is offline
Senior Member
 
Join Date: Apr 2009
Posts: 956
Re: L*a*b* question: why can't it be simulated in

Thanks mistermonday, that makes sense. I will certainly do additional testing as well.
Reply With Quote top
  #60  
Old 09-15-2010, 07:51 AM
Zampano Zampano is offline
Junior Member
 
Join Date: Apr 2006
Posts: 11
Re: L*a*b* question: why can't it be simulated in

Although the tone sometimes seems a little bit harsh, this thread is one of the most interesting ones here at RP...
Thanks a lot to all contributers for all the insights and especially for the example files...!
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
Converting CMYK to RGB and back to CMYK. xxxmen Input/Output/Workflow 23 01-06-2011 11:56 AM
my 1st post..red cast on Lacie 324/CS4 RGB to CMYK nadaman Hardware 47 03-10-2010 02:31 PM
Adobe RGB (1998) profile problem * Sinisa Software 7 07-10-2009 11:08 PM
RGB Help PSE JohnTravers Salon 5 10-13-2008 07:28 PM
silly question Tinkerbella197 Photo Restoration 5 09-15-2008 11:46 AM


All times are GMT -6. The time now is 05:13 AM.


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