Photoshop Channel concepts: The Power of Ten

[Graphics23]

An introduction to the concept of viewing every image as having ten channels. [details]

[DannyRaphael]

Thanks for your efforts here. Great concept and illustration. Clear and well written, too.

FYI: I changed the title of this thread for better visibility.

[Graphics23]

Here's an action for setting up a ten channel workflow.

You can start in any mode except Bitmap, Duotone, or Multichannel.

It'll take a file and create a duplicate.
The dup is converted to your working RGB and labeled RGB.

It'll go back to the original and dup it again.
This dup will be converted to LAB and labeled as such.

Back to the original and dup it one last time.
This dup is converted to your working CMYK and also labeled.

The original image is left unchanged.

This is a great way to become accustomed to what the various channels look like. Gaining an eye for this is an extremely useful skill since it can benefit just about anything one does in Photoshop.

I use this Action when I want to quickly explore what the ten channels have to offer. Quite often I find the start of a mask or an idea for channel blending that might not jump out at me right away.

As my eye becomes more experienced at inspecting channels I've noticed that ideas suggest themselves to me in ways which I most certainly wouldn't have considered previously.

For example, the other day I was working on an image where I wanted to increase the contrast between subject and background. After running the Ten Channel Action I noticed that in the A channel of LAB the subject was light and the background dark. After a quick blend of the A into the L using Overlay mode I got just what I was looking for.

Enjoy,

Michael

[Graphics23]

Updated tutorial.

Added comments regarding Contrast and Color.

Regards,

Michael

[DannyRaphael]

Originally Posted by Graphics23

Here's an action for setting up a ten channel workflow.

Great use of automation applied to an enlightening tutorial. A+.

Many thanks again, Michael.

[wallykid]

Hi Michael,

I know this thread very old but I decided I would try it anyway. I will also post this message on a new tread.

This is my first ever message on this discussion board. My question regards PhotoShop modes.

I am an aspiring photographer who is in the process of designing a website to display and sell some of my work. I therefore use PhotoShop to enhance my images for both web/monitor display and print display.

So my basic question is which mode should I work in? I originally thought this would be an easy question to answer, but the more research I do the more questions I have and the more I think maybe I have to work in more than one mode. On this website recently I read some interesting discussions about the 10 channel theory whereby you work on all channels in RGB, LAB and CMYK modes. I’ll be honest. This nearly made me cry!

Here is my dilemma. I am primarily a photographer so I don’t want to spend hours working in PhotoShop. I don’t find this part of the photography process nearly as satisfying as actually taking the photographs. I do realize its importance though and as a result I am very proficient in PhotoShop. I don’t mind studying new and difficult techniques, as long as they are the right ones for my purposes.

My preferred mode for enhancing for web/monitor display is LAB mode. I know however that for best print display quality I should work in CMYK mode.

Does this mean I have to have 2 separate processes for every image (surely there is a more efficient way) or can I start in LAB and then convert successfully to CMYK?

I know LAB mode has a very wide Gamut range and that converting to CMYK will lose some colour details. But these wide Gamut range colours can’t be displayed on a monitor anyway so is there any real point of working in LAB?

My latest workflow theory is that I could:
• Start my enhancing in LAB (colour, contrast, some masking etc..)
• Save file.
• For web/monitor display stay in LAB and sharpen. Finally convert to RGB to save for web.
• For print display convert to CMYK mode fine tune highlights/shadows and sharpen.

But would this workflow plan result in vastly different results. My aim is to keep my photos on the web the same as the photos I print. Both being of the best quality possible.

Or should I do all my work in CMYK and then convert back to RGB to save for web. Is CMYK as easy and effective as LAB for general image enhancing.

Too many questions! So if anyone out there has any suggestions or advice I would appreciate it very much. I’d even buy you a beer if we ever met!

Take care,

Conor

[Graphics23]

Hey Conor, welcome to RetouchPRO!

You ask; which mode should I work in?

Work in the mode that gives you the best technique to achieve your goal. That's the whole point of this article. All three modes have strengths and weaknesses. Taking advantage of the strengths and avoiding the weaknesses is the important thing.

It depends on the image. Your photographs will be starting out in RGB. If there's nothing to be gained by converting to LAB or CMYK, why do so?

I'm a prepress technician. In the exciting world of digital prepress, time is money. I can't afford to spend hours working in Photoshop, so I'm always looking for the most bang for my buck!

If an RGB curve gets the job done, great! But if I need more color variation, then I'm converting to LAB.

If a channel blend gives me what I need, super! But if a quick sharpening of the blacK channel makes the image pop, then it's CMYK and I'm done.

You state your preferred mode is LAB. Why? LAB is like a sledge hammer. You can't tap in carpet tacks with a sledge hammer! And just because printing is done in CMYK doesn't mean YOU "should work in CMYK mode". You should work in CMYK mode if that channel structure will give you the most leverage to move your pixels.

You should have one RGB image which is properly tagged. Convert this image to sRGB for web display and provide it as is to your printer after explaining that you have tagged it for color management and the tag should be honored. Then let the printer handle the conversion to CMYK.

This is also important when printing to your inkjet printer. Most inkjets expect RGB. If you send it CMYK, most likely it will convert to RGB and then convert again to CMYK.

LAB has a wide gamut, yes. But converting to CMYK won't loose any data if there's nothing in your image that's outside of the CMYK gamut. This general fear of converting due to lose of data strikes me as jumping at shadows. I convert when necessary and very seldom have I ever had gamut issues. Especially going from RGB to LAB and back. When I do have gamut issues there are workarounds.

Don't get hung up on color modes. They all have their place. Keep an open mind and try to remain flexible. As you gain experience things will sort themselves out.

Regards,

Michael

[Richard_Lynch]

Michael...
Interesting. One of the things I have done in books of mine is show how to work with CMYK in RGB mode by separating out the CMYK components based on light theory. I also, in my new book, show how to work with channels as layers -- so that there is really no need for two palettes, AND you have more flexibility with the changes that you make.

In general this type of thing is something most people need to know more about.

[Graphics23]

Hi Richard,

Thanks for chiming in!

Using channels as layers sounds very intriguing.

As you can surmise I'm something of a channel junky, so I'd very much like to discuss the concept.

I'm at a point where finding a new book on Photoshop has become very difficult. At the risk of tooting my own horn, most books simply don't speak to my level of experience. I have no interest at all in another book by "The Photoshop Guys" or any of the myriad recipe type offerings.

Other than works by Dan Margulis, I haven't found anything in quite some time which sparks my interest and also challenges me.

Would your "Layers" book be something an old timer like me find worthwhile? Perhaps if you elaborate on the channels as layers idea you might entice me into buying your book.

Regards,

Michael

[wallykid]

Hi Michael,

Thank you very much for your detailed explanation. I really do appreciate it. It has helped me a lot. Sorry to take up more of your time but I would like to ask you about the following section of your answer:

“You should have one RGB image which is properly tagged. Convert this image to sRGB for web display and provide it as is to your printer after explaining that you have tagged it for color management and the tag should be honored. Then let the printer handle the conversion to CMYK.”


My first question is about the term “tag for color management”. What do you mean by this? Is it a PhotoShop process or tool? I have looked in the PS help and found nothing. I then did some Google searches but can’t find anything helpful that explains what you might mean.

My second question is about PhotoShop working spaces. I was under the impression that for web/monitor output the sRGB working space should be used, but for print output Adobe RGB working space should be used. In one of my PS books it even says avoid using sRGB for print output. But you seem to suggest that working in sRGB is okay for printing.

This is your field of expertise so I would be really interested on what you have to say.

Take care,

Conor

[Graphics23]

Originally Posted by wallykid

Hi Michael,

Thank you very much for your detailed explanation. I really do appreciate it.

You're welcome! I'm happy to help.

My first question is about the term “tag for color management”. What do you mean by this?

In Photoshop, there's only one LAB color mode. But there are many versions of RGB and CMYK.

sRGB, Adobe RGB, ProPhoto RGB, Apple RGB, etc... are ICC Profiles. When an image has an embedded profile it is considered to be tagged for color management.

Color management does two things and only two things; it defines colors and it converts colors.

Without a profile the brightness levels in your image have no meaning. For example; in your RGB image you use the eyedropper tool and place the cursor over a given pixel. Looking at the info palette you see 196r 0g 0b. This is red. But what red; Candy Apple Red, Fire Engine Red, Brick Red, your Red, my Red? That's where the profile comes in. It's the profile that defines what the numbers mean.

Your digital camera captures an image in RGB. Your printer outputs that image in CMYK. At some point there is going to be a conversion. If colors have no definition they can't be converted.

My second question is about PhotoShop working spaces. I was under the impression that for web/monitor output the sRGB working space should be used, but for print output Adobe RGB working space should be used. In one of my Photoshop books it even says avoid using sRGB for print output. But you seem to suggest that working in sRGB is okay for printing.

sRGB is not a working space, it's an ICC Profile. Same goes for Adobe RGB. The Photoshop "working space" is simply the ICC Profile which will be assumed, or used by default, if an image is not already tagged or you don't specifically assign one. The term "working space" is something of a misnomer on Adobe's part.

Working in sRGB is OK and has nothing to do with printing. Printing, with some exceptions, is done using CMYK. The trick is to convert your RGB numbers into CMYK numbers which your printer will interpret in such a way that you get acceptable results at output. That's where properly tagging your image comes in.

Again I point out that I'm a prepress technician. As such, I don't work in any one colorspace, nor do I prefer one over another. I use ICC profiles as part of my "trick bag" of image correction/enhancement techniques.

There are two commands in Photoshop under the Edit menu; Assign Profile & Convert to Profile. I almost always start and end my image editing with one or both of these commands. I use them so much that I have them hotkeyed to F2 and F3 respectively. They both allow you to choose an ICC profile from a list of the profiles installed on your machine.

Assign Profile changes the appearance of an image without changing the numbers, by numbers I mean channel structure (RGB, LAB, CMYK) and brightness levels. A brightness level of 196r in sRGB is not the same color red as 196r in ProPhoto RGB.

Convert to Profile changes the numbers while attempting to not change the appearance, because out of gamut colors will change based on your choice of rendering intent.

To put it another way; Assign Profile tells Photoshop what a given "red" means, Convert to Profile tells Photoshop what numbers to use to get that "red" if it is possible to get that "red" using a specific channel structure.

So for me, asking what color space to use is like asking what filter to use. My answer would be to use the one you need to get the job done. Sound familiar?

Adobe RGB is not superior to sRGB, it's different. There is nothing wrong with using sRGB any more than there is in using Unsharp Mask instead of Smart Sharpen.

My advise is to start with sRGB and stay there until you begin to notice that it isn't giving you what you want. That day may never come, but in the meantime you'll get to know what to expect, you won't have to worry about converting for web, and if you give your files to someone else there is very little chance for unexpected results.

This also brings up a very important issue, handing your files off to someone else for output. Unless you know the person creating your prints understands how to deal with ICC profiles, giving an RGB file that is anything but sRGB is asking for an unpleasant surprise.

For an in depth study of this topic you can't do any better than reading Professional Photoshop by Dan Margulis.

Regards,

Michael

[Richard_Lynch]

Michael,
I don't know about pushing off someone to Dan's book as that was a pretty good primer. I use sRGB almost exclusively (largely because of tests in my workflow) and am glad to see you downplay the importance of color space -- too many, I think get riveted to an idea they've heard. Pre-press people don't have that advantage, as they have to deal with reality ;-). Probably the best thing that ever happened for me having to do with imaging was getting thrown under the pre-press train when I worked for a photography book publisher as a jack of all trades...including layout, color correction, and editing. It was a fabulous way to learn all about digital rendering.
I am curious as to your position stated:

sRGB is not a working space, it's an ICC Profile.

That may confuse some people as Adobe clearly uses profiles as names for working color spaces. I think I know what you mean, but would you care to clarify?

[wallykid]

Hi Michael,

Thanks again for your helpful explanations. I really am grateful.

I have one last question and then I will leave you in peace.

In your last post your compared the 2 PhotoShop commands “Assign Profile” and “Convert to Profile”. I have used “Convert to Profile” many times before, but have never used “Assign Profile”. And after reading your explanation I still can’t see why you would need to use this command. Why would you want to change how your image looks? You mentioned that you use BOTH very often. I found this surprising. Can you please give me some examples of situations when the “Assign Profile” command would be useful.

Take care,

Conor

[Michel B]

Imagine someboby has posted an aRGB picture for the web, and the profile cannot be read in the file exif info, you'll get dull and flat colours. Then you have to assign aRGB profile, and perhaps convert it to sRGB before posting for the web.

Michel B

[Graphics23]

Originally Posted by wallykid

Thanks again for your helpful explanations. I really am grateful.

Glad to do it!

I have one last question and then I will leave you in peace.

Famous last words?

In your last post your compared the 2 PhotoShop commands “Assign Profile” and “Convert to Profile”. I have used “Convert to Profile” many times before, but have never used “Assign Profile”. And after reading your explanation I still can’t see why you would need to use this command. Why would you want to change how your image looks? You mentioned that you use BOTH very often. I found this surprising. Can you please give me some examples of situations when the “Assign Profile” command would be useful.

ICC Profiles are typically used for color management, one of the most vexing and overrated issues in image manipulation. However, for color correction, purposeful use of ICC Profiles can be a tremendous boon to the prepress technician on a tight schedule.

Figure 1a would be a wonderful holiday shot, if only we could see it better. Due to the auto exposure feature of most digital cameras the picture is much too dark. As usual, the harried prepress technician is charged with quickly turning this amateur capture into something that looks like it was shot by a professional.

1a.jpg

Figure 1b was created using Photoshop’s Assign Profile command and choosing a custom ICC profile created for just such an occasion, the occasion being the need to quickly adjust a dark picture… not the Christmas holiday.

1b.jpg

As you can see, with one quick command I've significantly improved this image and made subsequent edits much easier, if they're even needed at all. Now it's a simple matter of using Convert To Profile to prepare the image for output. In this case I converted to sRGB for posting to the web.

Another example is that very often I get images which are not tagged. They open in Photoshop using my default profile, which Photoshop calls the "working space". But this may not be the correct profile for the image. I use Assign Profile to preview various profiles until I feel the color is correct, usually this just means trying sRGB, Apple RGB, Adobe RGB, etc.., after which I do whatever editing is needed, then convert to the output profile. This is very similar to what Michel B outlines above, only in the print world.

Regards,

Michael