Untitled Document

I've been away from the list for a few months, so I missed the discussion of sRGB vs AdobeRGB, but it does bring me back to my question of what space would we use if we started from scratch rather than debating which space is "good enough". I generally use ColorMatch for files I send to clients which will be printed in CMYK- less danger of them hosing the files than AdobeRGB in my experience, but no "stigma" like sRGB. I sometimes use AdobeRGB for clients when I the end results will be printed as Translumes or more than 4 color and I think some colors might be outside of the ColorMatch gamut. I don't actually test in any way to see if the colors are out of the ColorMatch or sRGB or AdobeRGB gamut- this seems too technical to me!

However, I have been interested in finding a better space. I had asked here about L*Star RGB a while back and just saw this interesting post on a different list (I hope Jack Lowe will not object to my posting this here):

"I have used L-Star for quite a while now and the best way to describe its advantages is perhaps to use Karl Koch's (the inventor's) words. He also created ECi-RGB, a space I had been using successfully beforehand:

"I must say that eciRGB as a working space has proven its qualities over these 5 years, but is has a major drawback like ALL these working spaces: Its gray axis is defined with a gamma characteristic which results in a grayscale compression on one end and an expansion on the other. If you start to edit an image in one of these working spaces, an increase or decrease of RGB values will have different effects in visually perceived brightness, depending on your starting point (relative to the L* axis).

In LStar-RGB this problem has been overcome, the gray axis is defined in visually equidistant steps like in Lab (or better L*a*b*), hence the name."

ECI-RGB and L-Star are identical in gamut, just the grey axis definition is altered so that it behaves differently when making colour and tonal edits.

What does this mean in the 'real world'? Well, have you ever noticed, when editing in a space such as Adobe RGB, that the shadows seem to block-in very rapidly when you, say, and an S-shaped contrast curve?

Shadow detail is lost too soon, in my opinion, and you have to work that little bit harder at retaining it... that's the compression Karl speaks of in the darker end of the grey axis. L-Star solves that, so that we get a much more predictable result when making edits to files. The finished result has a much more elegant feel than edits made in, say, ARGB. The tones are distributed much more uniformly across the board...

However, on a practical level, its not without its issues simply because recipients of these files don't recognise the name of the emdedded profile 'L-Star RGB' - it isn't 'Adobe RGB' and are inclined to ditch it on opening. Bah!

Still, that doesn't stop me using it as it means I have done my bit that much better and that is my job - my job isn't to educate people in other bureaux or design houses (unless they pay me to do so!) although I do what I can to help things along.

Check it out:

http://www.lstar-rgb.com/

As well as the above, the other advantage, to my mind, is that L-Star RGB (and therefore ECI-RGB) is what I would term a 'robust' working space. Its gamut isn't willdy huge (like some areas of ARGB) and the colours that one therefore works with are always going to be quite 'sensible' for future re-purposing. To that end, it is a great space to work with if your images are going to press as CMYK conversions (to ISO specs, for example) are generally less drastic.

Jack Lowe

http://www.jacklowedigital.co.uk "

Any comments?

Ric Cohn

___________________________________________________________________________

Re L*RGB, I'm interested in knowing whether this download works also for Windows XP operating system, and whether Photoshop CS2 can work with it. Any experience out there?

Mark Segal

___________________________________________________________________________

It's an .icc profile. Definitely works w/ CS2. I assume it works on Windows.

Ric Cohn

___________________________________________________________________________

Of course it will work. It's just a profile.

I've been using "L-star RGB" ever since I first heard of it and it works beautifully. It's neither too small nor too large: it's just right! said Little Red (RGB) Riding Hood.

:-)

Regards,

Terry Wyse

_____________________________

WyseConsul

Color Management Consulting

704.843.0858

http://www.wyseconsul.com

http://www.colormanagementgroup.com

___________________________________________________________________________

Thank you Ric. I see their <doukumentation> is only <auf Deutsch> for now, so my limited German makes this a bit of an obstacle to fully understanding it. Is it your understanding that the colour space is on the whole more limited than Adobe RGB(98)?

Mark Segal

___________________________________________________________________________

That's my impression. Although, looked at in the ColorSync Utility it doesn't necessarily look that way-- more even maybe, but not overall smaller? I'm sure more knowledgeable people on the list could give better info.

Ric Cohn

___________________________________________________________________________

Quite a while ago I started using a similar idea with a grayscale working space. Like AdobeRGB the Gray Gamma 2.2 compresses shadows a lot so I made a Grayscale ICC file that used the L* function for the transfer function. I've found this to be a very nice overall gray working space.

Later on I created an RGB ICC using the L* function for the RGB transfer functions. This is very similar to this L*star RGB profile. The one part that seems fairly arbitrary is the tri-stimulus values. I used the ARGB values, so my space is the same total size as ARGB just a different internal shape. I notice that L*star RGB uses a different set of tri-stimulus values and thus has a different space size.

It seems like a whole family of L* type profiles might potentially be useful to match various commonly used RGB spaces. Any merit to this?

Roy Harrington

www.harrington.com

www.quadtonerip.com

___________________________________________________________________________

Hi Mark.

I have posted a few ColorThink screenshots on my personal web page showing a comparison of L-Star RGB versus AdobeRGB. I am showing one 2D view in Luv space and three views in 3D Lab space:

<http://home.mindspring.com/~marcoug>

As you can see, it's not easy to say which is the "wider" space overall, since both have mismatches in several areas, to no one's clear advantage, in my view. AdobeRGB reaches deeper into the blue-purples, whereas L-Star RGB outflanks it in the red-oranges by a considerable margin. Also, the white point of L-Star RGB is more neutral than that of AdobeRGB (due to the fact that L-Star RGB is a D50 space, whereas AdobeRGB is D65, thus more blue).

L-Star RGB uses L* (a multi-point defined curve with 256-points in all 3 channels) instead of traditional gamma, whereas AdobeRGB is a more traditional gamma 2.2 space. L* has the advantage of a better overall neutrality.

Regards.

--------------

Marco Ugolini

Mill Valley, CA

___________________________________________________________________________

Hi Marco.

Have you looked at PhotoGamutRGB? For printed output, I think it has advantages over most of the others. Unlike sRGB, ARGB, L*RGB and most of the other editing spaces, PhotoGamutRGB is table based (and gray balanced). It was designed to encompass the gamut of all popular printing spaces (press, lightjet, inkjet, laser, etc). But unlike matrix spaces, it doesn't have huge areas of non-printing colors. To me, the advantage is that it is shaped like a typical output space, is designed specifically for printed output, and contains nearly ALL printable colors, regardless of the output device (maybe not a film recorder).

So, when converting from the PhotoGamutRGB editing space to the output space, moves will be smaller. This should put less work on the CMM and the profile and result in more accurate conversions, using either RC or Perceptual rendering.

And being table based, you can convert to it using Perceptual if desired. This allows a user to convert RAW files using ProPhoto, then convert to PhotoGamut RGB within Photoshop, using any desired intent.

Here's the link if interested.

http://www.photogamut.org/E_ICC_profile.html

Lou Dina

___________________________________________________________________________

On Feb 21, 2006, at 6:05 AM, Louis Dina wrote:

http://www.photogamut.org/E_ICC_profile.html

Very interesting site. PhotoGamutRGB certainly looks appealing in theory. Anyone want to offer their views on the advantages or disadvantages of using a profile like this one? I'm not clear from the site how "finished" the profile is, what the dangers might be in using a "beta" profile, and how recently it's been updated.

It's interesting (to me) how the imaging community seems to be splitting into "faiths". Here's how I see them.

1. The idealists: Capture and keep everything that the device is capable of in case it's needed later. I'd say these people would like to see the image as a window that would be indistinguishable from reality or even tint it until it's "better" than reality.

2. The pragmatists: Perfection isn't possible or necessary. Look at the tools you have and choose the best one for your immediate needs.

3. The optimists: There must be a practical choice that will give me everything I need- no more and no less. I want to avail the pitfalls I've experienced with wide gamut spaces and yet keep the data I need for all existing needs.

I suppose I fall into the 3rd camp. However, as Groucho Marx said, "I'd never want to join a club that would have me as a member", so I'm not ready to proclaim this the best practice moving forward.

Maybe what I need is a charismatic "preacher" to lead me <g>. Dan's newfound belief in sRGB certainly has me questioning my "faith".

Wouldn't it be ironic if sRGB did turn out to be the perfect working space?!

Ric Cohn

___________________________________________________________________________

From: "Ric Cohn"

Well I'm certainly more comfortable about using it more often now after reading these recent posts. We always did use sRGB for our scans from color negatives anyway and we never suffered for it. Now I like to evaluate the image a little more carefully before deciding to use our normal wide gamut space which has been DonRGB, skipping right over AdobeRGB which has the disadvantage of having a D65 white point.

john castronovo

___________________________________________________________________________

John,

Sorry, but I didn't get it.

Why is the D65 white point a disadvantage for AdobeRGB?

Isn't D65 becoming the standard in the press industry?

Clic!o Barroso

*****************

___________________________________________________________________________

Hi,

I have been loosely following this thread. And I want to follow up of the question about D65 by asking a few other question.

Who exactly is setting D65 as a standard?

Does this info come from some type of imaging expert(s)?

Is is a business?

Is it a standards group?

Or has it become a defacto standard somehow?

Or because is it built into certain working spaces that it is precieved to be a standard?

Or is someone just suggesting to use D65 instead of D50?

Or is it a perceived standard that might be thought of because a lot of folks are considering using D65?

I use D65 to set up monitor calibration and profiling sometimes and that can become an internal shop standard for a client but I find that D60 is more to my tastes.

And from my point of view its ok if any of these questions ring true. I am just curious. Any insight to this would be nice.

Thanks

Jim Rich

___________________________________________________________________________

Jim, I'm not an expert in this stuff, but for what it's worth, the explanation I've heard for using D65 on the monitor when trying to match prints that will be viewed under D50 (e.g. Solux) lighting conditions is that the human brain finds this pair to be visually equivalent for reasons that we don't understand - one life's little mysteries. This explanation comes from people in the colour management industry and experienced professional photographers, so regardless of whether it is or isn't any kind of official standard, it has some credibility behind it.

Mark Segal

___________________________________________________________________________

On Feb 22, 2006, at 9:48 AM, Jim Rich wrote:

For what it's worth, when Bruce Lindbloom created BetaRGB this was part of his decision process:

Quote: "Adobe Photoshop and the ICC profile specifications both use D50 as a reference white, this was the logical choice. If instead, a non-D50 white was chosen, then both the creation of, and the use of the working space would require adaptation, which opens the door just a crack for mistakes to be made. Specifying the working space directly in D50 avoids this possibility for error." End quote.

I don't know that the ICC profile specs still use D50, but if still true that would argue against D65 becoming a standard. http://

brucelindbloom.com/

I don't agree with everything Bruce has ever said (wink, wink), but I would say he has some qualifications to have opinions on this.

Ric Cohn

___________________________________________________________________________

Hi jc.

I do not understand why you see that as a disadvantage: through chromatic

adaptation transforms, the white point always remains perceptually neutral.

So, why exactly do you see it as a negative feature?

Best regards.

--------------

Marco Ugolini

Mill Valley, CA

___________________________________________________________________________

In a message dated 2/21/06 7:07 AM, Ric Cohn wrote:

Hi Ric.

"Idealists"? If you are an artist, and care about the quality of your work, you don't stand still, but move with the advances in technology and methodologies. To make one small example, Ansel Adams did not stop evolving his ideas about how to print his work, as a recent show has proven by offering concrete examples of how he kept reinterpreting certain images throughout his career.

It has nothing to do with "idealism," and a lot to do with a pursuit of excellence.

And "better than reality" is a highly subjective view, besides the fact that not even "normal" renditions of reality are perfectly faithful to the original, and alter it in ways that, though "inaccurate" strictly speaking, "feel" right to the observer anyway.

This could be the world of commercial image production, where people are used to the very real limitations of printing presses and to working "by the numbers." It is a view very different from that of the artist, and a very legitimate point of view too, but in no way "better" or "worse" than that of the "idealist". It all depends on whose goals are being served. No one solution fits all.

The "optimist" is the one that, not necessarily through any fault of his own, has no clear idea which final use the work will have, and covers all bases, or tries to. That's a legitimate concern too.

If I may say so, all one needs is to be clear on what final aim is being sought. Once that comes into focus, the rest follows, in my opinion.

If I may be so bold, I personally prefer a reasoned choice to a blind faith in the opinion of others.

And sad too. ;-)

Best regards.

--------------

Marco Ugolini

Mill Valley, CA

___________________________________________________________________________

on 2/23/06 4:59 AM, Marco Ugolin wrote:

I agree !!

Lee Clawson

2/\V/\7 Studio

___________________________________________________________________________

Beautifully stated Marco.I would add that the technology itself is a moving target. Stand still or rather stop looking and you have lost the goal which is to make great images.

Most people look for the solution and stick with it regardless. What has always drawn me to Dan's work is that he questions the process in a useful way.

Ann-Marie Stillion

___________________________________________________________________________

On Feb 23, 2006, at 3:59 AM, Marco Ugolini wrote:

Hi Marco, I'm not sure I'm getting your point. AFAIK the reinterpretation that Ansel Adams did in the course of his career had nothing to do with advances in technology or methodology and everything to do with a change in/advance in his personal "vision"- probably with a healthy dose of looking at what others were doing that he found interesting. Then again I personally feel he was a better teacher than artist, so I think his early efforts, while technically impeccable, had a lot of room for improvement. I guess I don't see what this has to do with technical decisions like color spaces unless you mean it shows that what's important are the ideas and not the technical choices.

Clear as mud in mine <g>. Please tell me where your "focus" has led you.

Ric Cohn

___________________________________________________________________________

Ric Cohn writes:

I was going to let this slide, but since two other list members have now quoted the above, I must correct it before it becomes another urban legend.

I am certainly not a "believer" in sRGB. To reiterate what I have said many times:

Prior to 1998, we had a system whereby everyone's RGB was roughly the same. The definition was not a particularly good one, but it was a standard. The few people who did not wish to use that definition had to make special arrangements with the next person who got their file to make sure that the file was understood correctly.

In the name of avoiding these special arrangements, a system was arranged whereby a special arrangement became necessary *whenever* an RGB file was passed from stranger to stranger. In the name of improving the admittedly mediocre 1998 standard, a system has developed where it has been replaced by a choice of two RGB spaces, each one of which is worse than what we had in 1998.

The only good thing about sRGB and Adobe RGB is that they are standards, which we need to have. Because they are standards, we are now only slightly worse off than we were in 1998, as opposed to the chaotic conditions of, say, 2002, when we were much worse off.

What I've said in support of sRGB is that

1) IF you assume that it and Adobe RGB are the only choices in the world, and

2) IF you are ONLY concerned with CMYK output, and

3) IF you are reasonably good at color correction and are prepared to spend the time to get better images,

THEN sRGB is a better choice than Adobe RGB.

As for other spaces, I certainly agree that the ones mentioned are better choices than Adobe RGB in a perfect world. However, now that things are relatively stable (almost everybody uses either sRGB or Adobe RGB) the last thing we need is to start a debate that will invite the geniuses who came up with the present system to have another shot at imposing an even worse one.

Dan Margulis

___________________________________________________________________________

All good point to clarify your position Dan and well worth the post. And, I think that most of us agree with you as well. However, I believe that we're better off today because now there's an understanding among most professionals of what a profile is and does. At least many more people have enough knowledge now to select a working space based on the needs of the image and not simply ignorant of the fact that a profile is in play at all. Even a year ago, most professionals would give me a blank stare when I asked them for their working space when they sent me an image to print. Now it's only pressmen. <g>

john castronovo

___________________________________________________________________________

On Feb 24, 2006, at 7:46 AM, dmargulisnj wrote:

Dan, thanks for the clarifications. For me, while CMYK is very important, your #2 is not true 100%. I frequently have images used for wider gamut outputs for which (if I read you correctly) you agree that sRGB might not be the best choice (although, if i read you correctly you don't have examples of images which have been prevented from printing colors in the images due to being in sRGB?).

I know that in many cases the smaller gamut spaces such as sRGB or ColorMatchRGB would be fine, but it seems to me that there appear to be enough cases where this is not true to warrant a space that is well designed and that covers a reasonably wider gamut.

I also believe that profiles are more understood now than 5 years ago, so I do not feel that AdobeRGB and sRGB need to be the last working spaces suggested to the educated Photoshop user.

IMHO we already have a worse one from this group- ProPhotoRGB. A few years ago I first saw this space recommended for very specific uses, not it's a default Photoshop space. Many now swear by it as they swore by AdobeRGB. AFAIK, out of the dozens of working spaces out there, the only working space that these people ever disliked is sRGB

<g>.

I guess rather than stick with what we've been given, I'd like to take some control away from those same people you disagree with, and take back some control to try and make an "educated" choice.

Ric Cohn

___________________________________________________________________________

Ric writes,

Correct. I don't have any images where sRGB is unduly restrictive under any conditions. I have some undersea pictures and shots of bananas and yellow peppers that fall within CMYK but outside of sRGB; however such special-case images should clearly be handled in CMYK regardless of one's RGB workspace.

(Note: Later that year, list member VladimirYelisseev provided an image of a brilliantly yellow flower that I conceded was a problem for sRGB. The threads are found in the same section as this one. Additionally, in 2007, in the Metropolitan Museum of Art, I found two antiquities, one from France, one from China, that were dark, saturated cyan, and parts would certainly have been printable in CMYK but out of the sRGB gamut—D.M.)

The better the output conditions the more likely there is to be a problem, that's for sure, so I would be uncomfortable with sRGB for very good conditions in spite of the fact that I have no current images in which it would cause difficulty.

I would say that the most skilled users probably realize more than they did five years ago the constructive uses of profiles, and that they are more prone to treat the profile as their servant and not their master. I do not agree that anyone else understands the issue better than they did five years ago, and I certainly think they understand it less than they did in the year or two after Photoshop 5 was released.

As for suggesting something else to users. I think you have to be fairly good at color correction before the choice of RGB even makes a difference--probably good enough so that you don't need my endorsement to realize that neither sRGB nor Adobe RGB is very good. OTOH, the last eight years have clearly shown how disastrous it can be for nonexpert users when their work is misinterpreted by strangers who don't understand their workflows. The danger of this happening far exceeds the minor benefits of using a better RGB. So I think there is a strong case for nonexperts using what the rest of the world does. Right now, the world uses sRGB and Adobe RGB.

That's correct. This is an example of how everyone's life could be made worse. However, for every time you hear a reference to an intelligently designed RGB like the one in this thread's title, you hear five hundred from shills touting whatever RGB some vendor has

paid them to hype.

History is against this notion. We don't have the bandwidth to drown out the paid spokesmen. I believe that everyone agrees that if we wanted something with a gamut that approximates that of sRGB, there would have been better choices than sRGB. And for sure, if we had wanted something to give us approximately what Adobe RGB does, then Adobe RGB is a poor choice. But in spite of the fact that these are clearly undesirable defaults, they have won out.

People who don't understand color but fancy that they do can make matters a whole lot worse for us if they are allowed to change standards. They have done so once already. I would prefer not to let them do it again. The best way to avoid this, IMHO, is to insist on sticking with something that works, rather than trying to outshout them with new proposals.

Dan Margulis