Frank Romano complains about colour management.

[Deleted member 16349]

On WTT, Mr. Romano again talks about some problems with colour management.

Frank Romano on Color Management - WhatTheyThink

I am glad to hear him discussing this even though he has no solution.

My view is and has been, that the general problem is not about standards but about wrong methods.

If the leaders of the industry have no clue about what is need, you all are in real trouble.

 

[gordo]

Unfortunately vendors do not solve problems unless they feel there is a clear ROI in doing so. The problem that Frank spoke about is not so difficult to solve but the ROI is.

 

[Deleted member 16349]

Unfortunately vendors do not solve problems unless they feel there is a clear ROI in doing so. The problem that Frank spoke about is not so difficult to solve but the ROI is.

Very true. In other industries, many of the solutions come from the research groups at universities. This is partially true for colour science related efforts but for graphic arts technologies it seems the universities don't do much.

One can say that this is due to a lack of investment but even if there was investments, the graphic arts institutions are not capable and the engineering universities lack specific knowledge that they can base their efforts on. Plus they have no interest in such a field.

 

[Deleted member 16349]

Gordon, it is good you have commented on this topic in WTT. I can't since it seems that I am barred from making any comments on WTT. They did seem to like some criticism I gave them recently and they censored it.

Anyhow, it seems Dov Issacs is blaming everyone else for not being able to deal with a messy method to manage colour for graphic arts. Is this the fault of the users or it this the fault of the creators of these systems? I think they have made poor methods to manage the reproduction of colour and can not seem to understand how to solve problems in the most practical way.

They think in terms of colour science for managing colour and that is the problem. One should use as little colour science as possible in the process but one should use the description of colour for the target and the output of measured colour.

I think it is very good that Mr. Issacs agrees with Prof. Ramano but I think both of them have little idea how to get to a practical solution that does not require such expertise.

Maybe this is the underlying reason why things do not get improved in the graphic arts. Printers don't want things to be simple because there expertise will be questioned and colour management suppliers and consultants want things to stay complicated. Since there is not leadership in this industry to question these parties, there is very little opportunity to investigate and demonstrate other methods.

No skunk-works type of groups that can challenge accepted knowledge. No leadership to fund them.

 

[gordo]

Here's the post Erik is referring to (very slightly edited for clarity):

RE: Dov Isaacs wrote: “I question your assertion that we at Adobe “hold most of the cards in this space” from the point of view that

(a) we have no way to force content providers and print service providers to properly use color management or use it at all,"

[Gordo] The way color management is currently implemented in authoring applications means that one has to be a color management expert in order to benefit from it.

Dov Isaacs wrote: "(b) we cannot prevent third parties from adding “secret sauce” solutions to the workflow"

[Gordo] Agreed. But perhaps something could be done within the apps that Adobe does control?

Dov Isaacs wrote: "(c) there is a limit to how much educating Adobe can do without cooperation from the graphic arts education community, much of which has little expertise and/or experience in this area."

[Gordo] Based on my experience, educating the creative folks who use the software in technical matters such as color management is problematic - if not futile.

Dov Isaacs wrote: "What do you think we should or could do?"

[Gordo] Color management as it's currently implemented is similar to the controls that a driver needed to know how to use in order to drive a 1920s car (e.g. How to: Drive a Ford Model T - Feature - Car and Driver ). But unlike the evolution of the automobile, we are adding more controls and more sophisticated ones, rather than eliminating them through automation. At the end of the day, most drivers just want to get from point A to point B. I.e. they want to get to a destination. And I believe that is also true for document creators. They just want to get to their chosen destination without all the complication.

So, would it be possible to automate the application of color management so that it is destination focussed rather than controls focussed? Or perhaps have that as an alternate workflow option so that sophisticated users could still fiddle with the controls if they so desired.

What I'm thinking of is something like a "destination" menu option that the user would use to select the destination of the document they're working on. The destination could be generic (newspaper, the web, sheetfed, etc.) or specific (Joe Blow's Print Shoppe). The authoring application would then apply, behind the scenes, the appropriate color management tags, profiles, whatevers based on the profile of that destination. If a destination tagged image is brought into a different application (say a PShop image into an InDesign page) the creative would be notified if there is a destination mismatch that the creative needs to resolve.

If creatives use an RGB environment then they could perhaps change the destination in their document and behind the scenes the authoring app would, behind the scenes, retag or do whatever is appropriate to purpose the document for the new destination.

It seems that all the components are in place (specified industry standard print conditions, profiles, and file output formats) but is dependent on the operator to bring the pieces together instead of the authoring software (which already has all this info built in). If the destination is non-standard, say Joe Blow's Print Shoppe, perhaps there could be some piece of software available for him to create a his own custom destination profile. So the creative might start with a generic destination profile - e.g. GRACoL, but once he knows he will be printing at Joe Blow's Print Shoppe he would then apply the Joe Blow's Print Shoppe destination profile and all the applicable changes needed would be automatically applied by the authoring app.

Just a thought.

==========End of original post===========

What I find fascinating is that color management and implementation has been an identified problem in the graphic arts since the introduction of desktop publishing in the mid 80s.
In the early days of DtP the computer displays were grayscale - not color. But I had no trouble specifying color with confidence because I (like all graphic designers) was used to specifying color without actually seeing it until I got color proofs from the trade house or printshop. At that point plate-ready film had been imaged so any changes would have cost a fortune. I simply specified color in my page layout application the same way I did on mechanical art - typing the specifications in the app instead of writing them on overlays. It worked very well.

This tying together and automating behind the scenes the application of color management, rendering intents, profiles, output, etc. has been done successfully before and is an offering by at least one vendor's prepress system - so I know it should be workable in the authoring space.

I think that Adobe (Dov Isaacs) is in a unique position to solve this decades old problem - all the components are there, except perhaps the will.

 

[Deleted member 16349]

Here's the post Erik is referring to (very slightly edited for clarity):

RE: Dov Isaacs wrote:

Well they are still arguing this topic on WTT. Good that they are making comments but I still think they have been and still are on the wrong track in thinking what the solution should be to obtain predictable print outcomes.

Eddy Hagen came close when he said that the problem is in the ICC profile. I think this is basically true. How can people expect to get predictable results if they use some generic ICC profile instead of a custom ICC profile for their printing conditions?

It seems to me that there is such a great effort to avoid making ones own custom ICC type profiles but people are willing to go through all the waste of time to do the steps for G7. And G7 does not ensure colour away from the neutral curve, which will also not be used in any images since most use some amount of Gray component removal. People are not printing CMY grays from what I have heard.

I suspect that custom ICC profiles are still not good enough but are closer to what is needed than some generic ones.

I wish I will be able play around with some ideas for this problem in the future. It certainly needs something to be done. And I have to agree with Mr. Romano, that it is now too complicated and I would add confusing. I think it can be made much more simple to obtain what's needed but certain technologies and methods need to be developed. I have been thinking about some to the related issues for the past 10 years.

Maybe I am really foolish but I don't see such a big problem at this time with developing a predictable system. As an example, if you tell me you can print a particular colour and then I ask you to print that colour, why would I not expect to get a good result. Now multiply that by millions of colours. I should be able to get what I want within the millions of colours you say you can print. That is potentially what an image is.

Part of the problem is how one asks for colours. One thing I have come to understand is that one should not use Lab values. Lab values describe colour but are not good for making mathematical calculations, which are required.

The use of multiple colour spaces is confusing and not necessary. Just use one device independent colour space but pick the right one and have everything be described by that colour space.

I find this an interesting problem and maybe because I am not a colour scientist, I can see things from a different direction. I look at it as an engineering problem and think of what is the most practical way to solve this problem without any loyalty to any particular field of study. Colour science is a distraction to solving these problems.

I hope the heated discussion on WTT goes on. Gordon will probably jump in again.

 

[gordo]

Reordered (and edited) for my purposes.

I hope the heated discussion on WTT goes on. Gordon will probably jump in again.

I'll not jump in again. It's a waste of time.

Well they are still arguing this topic on WTT. Good that they are making comments but I still think they have been and still are on the wrong track in thinking what the solution should be to obtain predictable print outcomes.

The fact that there is an argument on something so basic suggests to me that there are core problems that continue to be ignored. If Frank Romano, Frank Cost, Eddy Hagen, and John Clifford, on this thread can't figure it out - how can one expect creatives to do so? Are "they" the color orgs on the wrong track? Yup, I think so.

Eddy Hagen came close when he said that the problem is in the ICC profile. I think this is basically true. How can people expect to get predictable results if they use some generic ICC profile instead of a custom ICC profile for their printing conditions?

The notion is that the press has been aligned to the standard print condition - e.g. ISO 12647-whatever. So a generic profile based on the data set from that print condition should work. But, of course, it's been made much more complicated that it should be.

It seems to me that there is such a great effort to avoid making ones own custom ICC type profiles but people are willing to go through all the waste of time to do the steps for G7. And G7 does not ensure colour away from the neutral curve, which will also not be used in any images since most use some amount of Gray component removal. People are not printing CMY grays from what I have heard.

G7, AFAIK, was never tested against (the poorly written) ISO 12647-whatever specifications - i.e. If one adheres to ISO 12647-whatever do you end up with grey balance without the need for G7? The laughter that accompanied the intro of G7 in Europe suggests that is so.

I suspect that custom ICC profiles are still not good enough but are closer to what is needed than some generic ones.

Custom ICC profiles would just add another layer of confusion and complication. Printers should be able to print to an industry standard (embodied in standard ICC profile and/or to their own standard). Standard profiles where appropriate - custom ones when appropriate.

And I have to agree with Mr. Romano, that it is now too complicated and I would add confusing.

Yup - but not according to Adobe (via the topic thread) - who at the present hold the key cards.

Maybe I am really foolish but I don't see such a big problem at this time with developing a predictable system. As an example, if you tell me you can print a particular colour and then I ask you to print that colour, why would I not expect to get a good result. Now multiply that by millions of colours. I should be able to get what I want within the millions of colours you say you can print. That is potentially what an image is.

Yup.

Part of the problem is how one asks for colours. One thing I have come to understand is that one should not use Lab values. Lab values describe colour but are not good for making mathematical calculations, which are required.

Lab values (being numbers) are theoretically unambiguous. I don't know about making mathematical calculations.

I find this an interesting problem and maybe because I am not a colour scientist, I can see things from a different direction. I look at it as an engineering problem and think of what is the most practical way to solve this problem without any loyalty to any particular field of study. Colour science is a distraction to solving these problems.

You betcha.

 

[BeauchampT]

Other than the fact that so much of this goes over my head.....

The simple fact is that we are hunting for a grail here. With customer expectations and market materials available, we will always be adapting to new conditions. Constant adaptation is no basis for a stable standard, IMO.

At the end of the day, designers design, printers print, and salesman (can't believe i say this) bless them, are able to sort out the mess with the customer and we all move on.

I sometimes feel that the pursuit of the perfect system is simply what allows us to innovate new and better ideas. The fact is, we will never actually achieve the perfect system. Awful lot of fun to listen to the controversy in the meantime!

As always, thanks to all the experts and experienced men and women on this forum that keep me coming back for more!

 

[gordo]

The simple fact is that we are hunting for a grail here. With customer expectations and market materials available, we will always be adapting to new conditions. Constant adaptation is no basis for a stable standard, IMO.

At the end of the day, designers design, printers print, and salesman (can't believe i say this) bless them, are able to sort out the mess with the customer and we all move on.

I sometimes feel that the pursuit of the perfect system is simply what allows us to innovate new and better ideas. The fact is, we will never actually achieve the perfect system. Awful lot of fun to listen to the controversy in the meantime!

I don't think the issue is the pursuit of a perfect system. I would hope that it's about developing a system that allows the clear setting of expectations in presswork and the verification that those expectations have been met (and perhaps for Erik's sake, the technology to make it so). The pursuit of the perfect system, IMHO, doesn't allow anyone to innovate new and better ideas. In this world, the pursuit of profits drives the development of systems/solutions. If there's no ROI in terms of monetary gain there is no product/solution development no matter how important it is for the industry as a whole to have the product/solution.

 

[Keith]

If there's no ROI in terms of monetary gain there is no product/solution development no matter how important it is for the industry as a whole to have the product/solution.

Why doesn't someone develop that color management system and sell it to us printers just like developers creating MIS systems and selling it to the industry? G7 costs a printer money to be certified, correct?

Someone please point out what I am missing.

 

[Deleted member 16349]

Lab values (being numbers) are theoretically unambiguous. I don't know about making mathematical calculations.

.

I asked this question of the mathematical accuracy with using Lab values or some specific tristimulus values, from a colour science group on Linkedin. They confirmed for me what was suitable and what wasn't for obtaining accurate mathematical results. I needed this confirmation to know what kind of colour describing method should be used to deal with the general problem of reproducing an image.

Why is this important? Well if one has an image that is made up of pixels that have been described in some colour description method, one wants to be able to add pixels together and find out what the colour will be for the larger scale area. So if one uses Lab values, and one averages the L, a and b values of all the pixels to get an average Lab value for the larger scale, one will get a false result. It will not accurately result in the colour of the larger scale. Moving accurately from one scale to another is important for the generation of screens of different Lpi or in general to confirm that the colour of an image is correctly managed by sampling different sizes of area in the image.

So the colour scientists are good at providing the critical information needed to develop a practical method but they are not always so good at looking at problems that are slightly out of the specific field.

Anyhow, they confirmed that the basic colour tristimulus values will accurately result in predictable colour when mathematically averaged, which was important to me for my thoughts on how to make a reliable method.

I kind of do this for fun but it would be nice if something of value to the industry would come out of it.

 

[Deleted member 16349]

Why doesn't someone develop that color management system and sell it to us printers just like developers creating MIS systems and selling it to the industry? G7 costs a printer money to be certified, correct?

Someone please point out what I am missing.

Part of the problem is that printers don't demand better. They don't know how to ask the critical questions needed to pressure the suppliers and other organizations that are pushing faulty methods.

So many people are brainwashed into thinking that the developers and suppliers know what they are doing that printers don't question and demand answers. They just wait for some new product and have little idea whether it is the right approach or not.

The unfortunate truth is that the development community does not really know what to do to make a better product. They have been brainwashed too. It is not being done intentionally but just because the science community in the printing industry has little imagination and have not done a good job of developing a valid body of knowledge needed to think through problems.

I have been talking about the problem and solution to the print density variation problem on Printplanet for about 14 years. Has there been any outcry to the press manufacturers to investigate this? No. On the contrary, some printers fight the idea that something can be improved. So how can the industry benefit from new knowledge if it is fought by the very people it is supposed to help.

The same thing will exist for colour management. Too many people are convinced that the colour management community know what they are doing and there is too much invested in doing the wrong things, so there will be a huge problem to change this around.

Even though Mr. Romano has no clear idea what to do, it is of great service to the industry for him to bring this issue out into the open. Kind of late but still a good thing.

 

[Green Printer]

Erik maybe the industry needs a frontal lobe lobotomy.

 

[Deleted member 16349]

Erik maybe the industry needs a frontal lobe lobotomy.

I think I need one. Then I can forget about ..... what was I thinking about? I forgot.

 

[Deleted member 16349]

Again Prof. Romano has made comments about the mess in colour management. These comments were made due to the response to his earlier video. It is here in his WTT video for those who can see it.

Frank Talks Color Management (In a Purple Shirt) - WhatTheyThink

One of his suggestions is that there should be some kind of meeting where all interested parties can get together and decide what is a better approach.

I see a problem with this. The very same people who could not develop simple and reliable methods are not likely to come up with them now. They have too much personal investment in the approaches they already have developing and working with for a long time. It is not some kind of conspiracy but only human nature.

The other problem with this approach is that one will get into the trap of "group think" where everyone starts to think that one approach is right without really thinking about it deeply. We already have problems with this in the industry. As an example, the efforts from IDEAlliance trying to impose their methods as some kind of industry standard that everyone should follow.

This also goes for the groups that have developed Standards. The idea that if a Standard is established, then it MUST be right because a committee developed it. Unfortunately the people in these committees have not been capable of actually solving many problem but they are supposed to be good enough to develop Standards others are supposed to follow.

These groups of industry experts will never come up with the new ideas that are required. They have too much of their personal reputations at stake. They will drowned out and push out anyone that tries to think differently. This is nothing new and it happens in all industries and fields of science.

I think diversity is the probable solution. The industry should be open to having many small groups try to come up with better ideas and demonstrate them. One does not have to develop full blow products to demonstrate new ideas. New ideas can be done with relatively simple tests. New thinking is actually low cost but it requires the ability to think.

Why does this not happen? Well as long as the large printers support the centralized thinking that exists now, there will be little chance for diversity. Large printers, who have resources, need to support new thinking in a diversified way.

The large printers in the industry have failed to meet their responsibility in this area. Large printers should support the effort to develop new knowledge by supporting small groups that have ideas. There will be many failures but there will be some successes. Maybe the ratio is 10 to 1.

Historically, new knowledge has been developed in a messy way. Most people forget or just don't know of all the failures that happened in search of the few successes but we all have benefited from the successes.

For the general printing organization, the opportunities for advancing new knowledge and developing more effective technologies lies not in the direction of things that they think are the right ones but in the direction they think were the wrong right ones. That is why innovation is always a surprise.

 

[gordo]

Of course Erik, where you see a problem others see problem solved. Here, summing it up, is Adobe's response from the original Frank Romano post:

"By Gordon Pritchard on Oct 14, 2014

@Dov, so I think you're saying that as long as content creators leave the application defaults alone then there is no problem?
And in Frank Cost's example - he broke the integrity of the PDF by using High Quality Print setting recommended by CreateSpace rather than the default.

By Dov Isaacs (Adobe) on Oct 14, 2014 @Gordon:

Almost that! Choose the initial default color settings for the Creative Suite / Creative Cloud applications and synchronize and use then globally. For PDF export, the existing PDF/X-4 settings with your favorite output intent profile (matching the CS/CC default CMYK color space) should (famous last words) get you what you need for most all print publishing workflows."

Boom! LOL!