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.

Newb: out of gamut colors and color shift

Reply
 
Thread Tools
  #21  
Old 01-17-2011, 03:10 AM
creativeretouch creativeretouch is offline
Senior Member
 
Join Date: Jan 2010
Location: London
Posts: 502
Re: Newb: out of gamut colors and color shift

"Its far more effective to soft proof and SEE what the colors will look like based on the rendering intent."

What kind of soft proof for CMYK do you mean? A RGB monitor? As I understand the "gamut warning" tool shows me all colors which are out of gamut (based on profiles which I have set up) and I should be aware of.

I have a feeling you are trying to tell us "do not worry, create any kind of color you wish and the rendering intent will take care of it"

Why is a bad Idea to see colors which are out of gamut?
Reply With Quote top
  #22  
Old 01-17-2011, 04:05 AM
Markzebra's Avatar
Markzebra Markzebra is offline
Senior Member
 
Join Date: Aug 2001
Location: London, UK
Posts: 718
Re: Newb: out of gamut colors and color shift

Quote:
Today, most users have to work with multiple output needs. Its really silly to funnel RGB into a small color space for one use (CMYK).
Thats not the case if your intended output is print. Most professional retouchers have print as their destination, and objective. And no one is suggesting 'funneling' anything completely into the narrow CMYK gamut, thats not necessary. Only that its best to keep saturation within the bounds of common sense. Especially if you are using a calibrated monitor, and want to see something related to what you are likely to get in print as part of your process.

Quote:
NONE of this has any bearing on the gamut warning overlay! Or to put it another way, the gamut overlay has no role in the gamut mapping, it simply shows you all out of gamut colors.
Did someone suggest that Gamut Warning had a role in gamut mapping? If they did I didn't see it. Its very quick to toggle with command shift Y, why not do it? Its showing you crudely what colors are out of the CMYK space that you have in Color Settings. That certainly isn't a valueless feature, that should be removed. Retouching while in Proof colors is also useful, and they are very well set up in Photoshop for people working primarily for print. These are certainly not "legacy" features.

Quote:
Provide a demonstrative way to prove that point will you?
I've already done that. General policy is that its better to remain somewhere respecting CMYK gamut. For the very good reasons I have already sated above Ö
Quote:
you will likely be sending samples to your client with colors that are not produceable on conversion. And judgments on the image while working within these highly saturated colors will be compromised.
Retouching in most RGB spaces with the exception of sRGB, with no awareness of the limitations of CMYK or the destination Gamut that it represents, has all kinds of problems. And should not be recommended or encouraged, because most people are unaware of these issues. Thats why I was forced to chip in.

Quote:
Some reading material:
Disappointed, but mainly in your attempt to patronize your way out of the argument.

Quote:
Rendering intents: Relative tries for the most saturated version of a color it can produce in CMYK, and as a result sometimes does a bad job of transitions in saturation. Perceptual: will bring everything in to order in a smoother more graduated way. But its more liable to under-saturate colors.
That is an accurate statement of the way Photoshops Relative and Perceptual rendering intents work, when you actually convert using them. For Retouchers, not for color Scientists. Instead of implying the quote is wrong in some way, be more specific, would you care to point out what part of that quote is incorrect in your view? I will be interested to hear this answer.

"Lazy" or not, not everyone needs, or wants to to be a color nerd. Retouchers are dealing with Photographic sourced imagery. Any reference to the other Rendering Intents for example is purely technical speak, and more than likely designed to show off than to actually help anyone.
Reply With Quote top
  #23  
Old 01-17-2011, 04:14 AM
Markzebra's Avatar
Markzebra Markzebra is offline
Senior Member
 
Join Date: Aug 2001
Location: London, UK
Posts: 718
Re: Misinformation Warning

Quote:
Originally Posted by 5nap5hot View Post
I was always told that I "should" be using ProPhoto RGB to process my images.
ProPhoto contains within it all, or most of the colors you camera can capture. Thats why its recommended, because the theory is that you are editing in a space which more resembles the input. If you are working for print however, and you know thats going to be your intended output, then its not a good move. In an 8 bit image you only have 256 values in each channel, and these colors cover that entire gamut. You can see that on editing, it may also cause banding problems. Thats why 16 bit is recommended for ProPhoto. For complex retouching unfortunately 16 bit can create a large time overhead.
Reply With Quote top
  #24  
Old 01-17-2011, 04:17 AM
Markzebra's Avatar
Markzebra Markzebra is offline
Senior Member
 
Join Date: Aug 2001
Location: London, UK
Posts: 718
Re: Newb: out of gamut colors and color shift

Quote:
I have a feeling you are trying to tell us "do not worry, create any kind of color you wish and the rendering intent will take care of it"
Thats what I'm worried about too CreativeRetouch, and from experience I know that that would be very incorrect advice to give. For previewing reasons, and the reasons I mention above.
Reply With Quote top
  #25  
Old 01-17-2011, 06:35 AM
Repairman's Avatar
Repairman Repairman is offline
Senior Member
 
Join Date: Dec 2010
Location: Bristol UK
Posts: 683
Re: Newb: out of gamut colors and color shift

All of my retouching work ends up being printed. Some of it ends up on websites which I don't worry about because the different browser/monitor variables do not warrant the fuss. Before digital photography took over two typical work flows would be a) tranny, CMYK scan, retouch, seps. Generally the scan would remove the col vibrancy issue. If it didn't I would fix it properly not by a colour wash. b) tranny, scan, retouch, digital tranny. If I needed tranny output I would make sure the scan and the output were made by the same company. At £250 a pop I couldn't afford colour matching issues to land on my doorstep. I always trusted my scans even if the monitor display raised doubts. (Before digital we had the original photo/tranny for comparison). Digital is great but often the colour files I get are not good. Many snappers rely too much on fiddling with the files after the event in order to make them look pretty on screen - probably the least relevant place in the process.
R.
Reply With Quote top
  #26  
Old 01-17-2011, 10:50 AM
andrewrodney's Avatar
andrewrodney andrewrodney is offline
Senior Member
 
Join Date: May 2010
Location: Santa Fe
Posts: 1,090
Re: Misinformation Warning

Quote:
Originally Posted by 5nap5hot View Post
Ok I need an education.

I will state for the record that I am a photographer first. I was always told that I "should" be using ProPhoto RGB to process my images. Because of this, I use ProPhoto RGB as my default colorspace in photoshop.

Could someone explain why ProPhoto is bad? Why would people recommend it?
Its not bad, you should be using it as described in the Adobe article whoís URL is above. The caveat is, do the work in high bit (what Photoshop calls 16-bit). The article explains why. If you use a smaller working space, especially in Adobe raw conversion products, you are throwing away color info that the capture device probably captured. Both Lightroom and ACR use ProPhoto RGB (with a linear gamma) for all processing as well. You can and likely will spin off an iteration in a smaller color space when the need arises. Like posting images to the internet in sRGB. You can always go from a larger to a smaller color space just as you can pour a cup of water from a gallon container into a quart container. Going the opposite way buys you nothing.
Reply With Quote top
  #27  
Old 01-17-2011, 11:45 AM
andrewrodney's Avatar
andrewrodney andrewrodney is offline
Senior Member
 
Join Date: May 2010
Location: Santa Fe
Posts: 1,090
Re: Newb: out of gamut colors and color shift

Quote:
Originally Posted by Markzebra View Post
Thats not the case if your intended output is print.
Once again, you are unclear and one would assume purposely vague. Print can be to an ink jet, a chromagenic sliver process, a dye sub, a press, a press that uses more than four CMYK colorants, etc. Just CMYK ink on paper is a big pile of differing gamuts and needs. An Indigo is nothing like a sheetfed press which is nothing like a web press or a press using Hexachrome. Funneling your master data into a small gamut space is inflexible and unnecessary.

Quote:
Most professional retouchers have print as their destination, and objective.
Print to what, and when?. So if someone whoís spent hours retouching an image in (god forbid) sRGB and has a client that now needs a higher quality or differing quality output (from billboard to print on a wall), all youíve done is funnel the initial data into something smaller from the get go for what purpose? You continue to avoid answering simple questions about workflow such as why not archive high gamut data for any possible output instead of tossing away color gamut? Youíve yet to explain the benefits to a small gamut working space master. Which Iíll add has zero to do with this original topic that the gamut overlay is basically useless.

Quote:
And no one is suggesting 'funneling' anything completely into the narrow CMYK gamut, thats not necessary. Only that its best to keep saturation within the bounds of common sense.
WTF does that mean? Common sense to me means saving all the color the capture device and the process (scan or raw conversion) possible. For a raw workflow, that means ProPhoto RGB 16-bit. For a scan that means the scanners native color space described by its input profile. What does it mean to you (and more importantly why)? Now as to your common sense, its yet to be defined despite the questions asked of you.

Quote:
Especially if you are using a calibrated monitor, and want to see something related to what you are likely to get in print as part of your process.
Again, this is vague. What kind of display? One that is limited to sRGB or the ones I and others use that exceed Adobe RGB gamut? And are you suggesting that today in 2011, we throw away all colors outside display gamut, colors that can be output because the current display gamut is small? Talk about throwing away the baby with the bath water. And yet, you propose that the gamut overlay, an ugly solid color that hides the colors you can see is somehow useful? How?

Quote:
Did someone suggest that Gamut Warning had a role in gamut mapping?
Well it does, its just basically useless for the reasons Iíve explained and you have avoided. Again, it treats everything identically, it places an ugly color over the actual colors, it canít do squat to change the effect of a color space conversion using ICC profiles. The question Iíve asked you over and over again, one you continue to avoid answering is just what does the overlay provide thatís useful? What does it allow you to do that changes what the conversion with ICC profiles and various rendering intents will apply to the data?

Quote:
Its very quick to toggle with command shift Y, why not do it?
Command Y toggles the soft proof which is VERY useful. Now turning on the Gamut Warning hides this useful soft proof with a solid color. The question you refuse to answer is, whatís the reason anyone would do this considering what a soft proof provides. Again, its basically a useless preview (the overlay). I said that in my first post and until you can explain what Iíd do now with this overlay, Iíll stick to that opinion.

Quote:
Retouching while in Proof colors is also useful, and they are very well set up in Photoshop for people working primarily for print. These are certainly not "legacy" features.
Unfortunately you are misunderstanding the difference and usefulness between the soft proof (Customize Proof Setup) which we got in Photoshop 5 in 1998 and the Gamut Warning overlay that is a legacy feature from prior to Photoshop 5. You absolutely do not need to invoke the Gamut Warning while using the Command Y Customized Soft Proof. And in fact youíve as yet explained why hiding the soft proof with the ugly overlay is useful or what you would do with this.

Quote:
General policy is that its better to remain somewhere respecting CMYK gamut.
Whoís general policy besides apparently yours? And as Iíve said and will be happy to illustrate, thatís not possible with any RGB working space without using a larger gamut working space. Do you think that even sRGB has a gamut that matches (respects) CMYK gamut (as if every CMYK device had the same or similar gamut, a huge misunderstanding on your part). First define the CMYK process. Lets even say SWOP V2 based on the older TR001 specifications. There are Cyanís and greens in V2 that fall well outside sRGB. They do not fall outside the gamut using Adobe RGB (1998)! This alone bodes for a larger working space not a smaller one! Now lets get past SWOP V2 and consider an Indigo using more than 4 colors, an Epson Ink Jet etc. Adobe RGB is too small. ProPhoto isnít.

Quote:
Retouching in most RGB spaces with the exception of sRGB, with no awareness of the limitations of CMYK or the destination Gamut that it represents, has all kinds of problems.
OK, instead of being an alarmist, why donít you provide some actual examples we can test ourselves. Instead of using words like ďdamageĒ, why donít you show us, provide files we can use to test this theory. As Iíll show you, sRGB which you above seem to propose canít contain the CMYK gamut of SWOP V2. And it canít contain the gamut of lots of other output color spaces, CMYK or otherwise. So you are proposing using a larger gamut working space? But wait, that would mean apparently ďproblemsĒ because its far wider than the limitations of CMYK. So what are you proposing for a working space?

Quote:
And should not be recommended or encouraged, because most people are unaware of these issues.
Enlighten us then.

Quote:
That is an accurate statement of the way Photoshops Relative and Perceptual rendering intents work, when you actually convert using them. For Retouchers, not for color Scientists
Retoucher, photographer or color scientist, using the incorrect terminology and explaining the roles of rendering intents incorrectly is no excuse. You are entitled to your own opinions, you are not entitled to your own facts!

Quote:
Instead of implying the quote is wrong in some way, be more specific, would you care to point out what part of that quote is incorrect in your view?
Sure.

Quote:
Relative tries for the most saturated version of a color it can produce in CMYK, and as a result sometimes does a bad job of transitions in saturation.
First off, the idea it does a bad job is simply untrue for all conversions using all profiles. Second, what RelCol does is take all out of gamut colors from the source color space and map them to the edge of the output color space gamut. I wrote this above.

Quote:
Perceptual: will bring everything in to order in a smoother more graduated way. But its more liable to under-saturate colors.
Sloppy. As I also wrote above, the Perceptual intent try to maintain cool relationships (they key word is try as I told you that every perceptual mapping is proprietary). It affects both in and out of gamut colors in the mapping, ANOTHER reason why the Gamut Warning is especially not useful here. In Gamut colors that will undergo mapping using Perceptual but not using RelCol are not shown within this overlay! Again, there are no specifications in how a perceptual rendering intent should be conducted, at least with profiles built to version 2 (v2 ICC profile format), which at this time are most prevalent. Each manufacturer of a product that builds ICC profiles that can conduct a perceptual rendering intent can use whatever techniques they feel will produce the most pleasing colors. Two profiles built from the same printer made by two different profile packages will almost always produce two different prints when a user selects the perceptual rendering intent. Talking about a Perceptual intent as being better (smoother) or worse without understanding this fact is critical since you are making a black and white statement that is simply too simplistic to accept.

The saturation rendering intent was at one time recommended for solid graphics like logos or pie charts (sometimes referred to as business graphics) and the gamut mapping is weighted to produce the most vivid saturated colors (hence the name). For this reason, using this intent on images could produce less than desirable results. However, depending on the profile and how it was built, the saturation intent might be fine for some images so donít dismiss it outright.

The relative colorimetric rendering intent uses the gamut clipping technique and takes into account the white of both the source and destination color spaces when converting colors. The ICC profile knows all about the device it describes, including the paper white of the device. A color in the source is perceived relative to the white of its paper. This method takes into account this relative perception and calculates how the same color will appear relative to the destination paper color. The colors that fall within gamut are not affected at all. I find that the relative colorimetric intent works quite well when the source and destination gamuts are similar. Picking an intent is both image- and profile-dependent.

Profiles don't know squat about images! They see your image as a huge pile of solid colors (pixels with Lab values) having no idea about the solid color next to it. Profiles have no idea about colors in context. The rendering intent you use is image specific! You must view the differing rendering intents using the soft proof and pick the one you visually prefer. One image may produce a superior color appearance using Perceptual while the next may call for RelCol (or perhaps Saturation). This is WHY we use the Customize Proof Setup to view the intents (and why the Gamut overlay which now blocks this makes it basically useless).

Quote:
"Lazy" or not, not everyone needs, or wants to to be a color nerd.
Yup and they should not be doing color space conversions! Or deciding what encoding color space the original image comes to them in. They should retouch and leave color management and prepress to those who understand the process!

Now, perhaps youíll stop digressing from the original topic and explain to us the various questions asked of you about the Gamut overlay and the damage you propose. I think the lurkers here now have the necessary info about rendering intents and encoding color spaces.
Reply With Quote top
  #28  
Old 01-17-2011, 11:56 AM
andrewrodney's Avatar
andrewrodney andrewrodney is offline
Senior Member
 
Join Date: May 2010
Location: Santa Fe
Posts: 1,090
Re: Newb: out of gamut colors and color shift

Quote:
Originally Posted by creativeretouch View Post
What kind of soft proof for CMYK do you mean?
Any CMYK output device you have a good ICC profile for, that defines that output device.

Quote:
As I understand the "gamut warning" tool shows me all colors which are out of gamut (based on profiles which I have set up) and I should be aware of.
To now do what? The question I canít get an answer to. You have a document in say Adobe RGB (1998). You load a soft proof for SWOP V2 (presumably that defines the actual print process). You MUST convert from Adobe RGB to SWOP V2. The gamut overlay shows you, god forbid, there are colors in Adobe RGB that fall outside the gamut of SWOP (no big surprise). Now what do you do? You must covert. You must use that profile. You must map out of gamut colors of the working space into the gamut of the output device. You see an ugly overlay. Reality sucks.

What you do is setup the Customize Proof Setup with the correct profile and toggle the rendering intents while you examine their effect on the image. You select the intent that produces the most pleasing color. The profiles have no idea if this is a black cat on coal or a white dog on snow or a very colorful scene. YOU (or someone else) has to view the pixels in context and decide on the rendering intent. But the out of gamut colors fall into gamut, differently using differing intents but they fall into gamut. In what way did the ugly overlay help?

The basic point I made is, the overlay, a legacy feature provided prior to ICC profile support in 1998, prior to rendering intents for conversions, prior to the soft proof is basically useless. As yet, no one has told me why thatís incorrect, what I could possible use the overlay for.

Quote:
I have a feeling you are trying to tell us "do not worry, create any kind of color you wish and the rendering intent will take care of it"
In a nutshell yes. What else can you do? You have a destination color space with a fixed gamut. Youíve got colors that are outside that gamut (but maybe not a differing output device). If you have good profiles, look at the effect of the rendering intent and pick the one you prefer and convert.
Reply With Quote top
  #29  
Old 01-17-2011, 12:46 PM
SilvaFox SilvaFox is offline
Senior Member
 
Join Date: Apr 2008
Location: Florida
Posts: 128
Re: Newb: out of gamut colors and color shift

Never fails. Whenever I see "CMYK, RGB, ICC or GAMUT", a total $#!%storm ensues.

After all these years of doing this, do I really need to read the book, or go to school to know what looks good and what doesn't? (Sounds like YES)

Can I still manipulate colors on my calibrated monitor so they look good, have the correct values, and have the best chance of looking that way on the final product? Whatever is is. Oh, and can I do it anyway I want as long as (that's right) it looks good? (Sounds like a big NO.)

A lot of us are not able to EXPLAIN (scientifically) why something looks awful.

Sorry. Time to get back under my bridge.
Reply With Quote top
  #30  
Old 01-17-2011, 01:03 PM
creativeretouch creativeretouch is offline
Senior Member
 
Join Date: Jan 2010
Location: London
Posts: 502
Re: Newb: out of gamut colors and color shift

Andrew, have you ever heard about "luminosity masks" or "saturation masks" and other tools which will help you to adjust colors & contrast ??? This "ugly overlay" - how you call a gamut warning - will help you to define areas you have to continue to work on - to get them inside your final gamut as much as posible and keep all colors in a desired balance.
In my opinion, playing with ICC profiles without knowledge how to fix the source file will give you ugly final result ... as we can see everywhere around us every day ... unfortunately ...
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



All times are GMT -6. The time now is 08:31 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