Dan Margulis Applied Color Theory

Who's at Fault?

Date: Fri, 27 Jun 2003 15:35:53 EDT
   From: Dan Margulis
Subject: Who's at fault?

*****Forwarded message*****
Date: Fri, 27 Jun 2003 18:54:12 +0100 (BST)
Subject: Who's at fault?
To: Dan Margulis

I recently had to send a psd file to a printer because
my Mac was dead. I removed the storage drive and
salvaged the file but was unable to convert to TIFF
due to the hardware problems I was having.

I explained to the printer that the file was a psd and
told him "the design work is done it just needs
converting" (word for word). He said he would look at
it and I told him to call me if there were any
problems.

There were no phone calls so I assumed everything had
gone OK. Anyway, turns out he converted the file to a
TIFF but didn't convert to CMYK. I needed to use some
lighting effects in PS and had set up in Adobe 98.
Needless the say the prints turned out bad! They have
refused to reprint and want payment for the job.
 
Who is at fault here?

Alain Corf
________________________________________________________________________

Date: Fri, 27 Jun 2003 17:47:06 -0400
   From: Preston Earle
Subject: Re: Who's at fault?
 
What sort of print job (offset, ink-jet, digital, other)? What sort of color proof? If no color proof, why not? (One of the purposes of a color proof is to catch errors like this, isn't it?.) What about some compromise on the reprint, like you pay for materials and he provides the labor, etc. or you pay for the cost of the reprint but the he issues you a credit against future business (at say $.10 to $.20 on the dollar) for the cost of the reprint such that he keeps a customer and you ultimately get your reprint for free?

Preston Earle
________________________________________________________________________

Date: Fri, 27 Jun 2003 17:18:14 -0700
   From: Jono Moore
Subject: Re: Who's at fault?

You should have signed off a proof. If you told the printer to go ahead and print without a proof then you are kinda stuck.

If the picture looked obviously bad then I would question the printer on why someone didn't mention the picture looked bad. But if the picture looked ok, they wouldn't know what it is supposed to look like.

As Preston mentioned in his message you may be able to work some kind of deal on a reprint.

It's going to depend on how big a print job it is and how much the printer wants to keep your business.
 
--
Jono Moore
udoprinting.com
 ________________________________________________________________________

Date: Fri, 27 Jun 2003 23:47:19 -0400
   From: Terry Kruska
Subject: Re: Who's at fault?

Sorry but you are.  While a good printer probably would have anticipated problems and contacted you. They simply followed your instructions.
 ________________________________________________________________________

Date: Sat, 28 Jun 2003 08:12:26 -0500
   From: Henry Segalini
Subject: Who's at Fault?

We are offset lithographers and our customers' expectation levels vary widely.

We have one (repeat "one") customer who wants to save money and doesn't get a proof.  However, what they get (final product) is what they get and they pay for it.  No discussion.

Running a job without a proof is so abnormal for us that I (upper management) sometimes have to intervene, despite written instructions on the job ticket,  to assure the press crews that it is OK to proceed without a proof.

So, your verbal instructions are worth the paper on which they are printed.

Henry Segalini
St. Louis  MO USA
 ________________________________________________________________________

Date: Sat, 28 Jun 2003 13:43:38 EDT
   From: Dan Margulis
Subject: Re: Who's at fault?

Alain writes,

There were no phone calls so I assumed everything had gone OK. Anyway,
turns out he converted the file to a TIFF but didn't convert to CMYK. I needed to
use some lighting effects in PS and had set up in Adobe 98. Needless the say
the prints turned out bad! They have refused to reprint and want payment for
the job. Who is at fault here?

I would have to agree with others who have posted, that the blame is primarily yours although the printer shares some of it.

This scenario has played itself out several times on this list and dozens on others. The culprit is invariably Adobe RGB.  I don't favor that RGB definition anyway, but that's beside the point. Anyone who uses Adobe RGB absolutely has to take the responsibility to make sure that the next person knows, because if the next person doesn't know and doesn't take action it will be a flat, washed-out disaster. With other RGB definitions, the result wouldn't be nearly as bad.

We were first faced with this issue in, I think, 1999, when Richard Kenward was victimized by a somewhat similar turn of events. At that time there were some souls who thought that embedded profiling might catch on as a workflow, so some people blamed the service provider. Since then, the consensus on the list has been that most printers don't even know what an embedded profile is because they've set up their systems to ignore them completely.  Hence, those who've experienced the problem have found little sympathy here, because regardless of whether printers *should* respect the Adobe RGB setting, the fact is they *don't*, and people who expect that they will must accept the blame for the inevitable result.

Dan Margulis
________________________________________________________________________

Date: Sun, 29 Jun 2003 10:24:27 -0400
   From: John Castronovo
Subject: Re: Who's at fault?

Wait a second guys. Are we to understand that the printer sent the RGB file to press without converting to CMYK at all, or did the conversion turn out poorly because the profile was ignored?

john castronovo
tech photo & imaging
________________________________________________________________________

Date: Sun, 29 Jun 2003 22:59:49 +0100 (BST)
   From: Alain Corf
Subject: Re: Who's at fault?

 --- John Castronovo wrote:

Wait a second guys. Are we to understand that the printer sent the RGB file
to press without converting to CMYK at all, or did the conversion turn out
poorly because the profile was ignored?

Yes, the printer sent the RGB file to press without converting to CMYK at all.

I have now created a confirmation form to be sent out with my artwork which includes space for "special instructions". The form will be signed and faxed back to me so there can be no confusion about instructions.

Best regards,

Alain Corf
________________________________________________________________________

Date: Mon, 30 Jun 2003 01:00:02 +0100
   From: Martin Orpen
Subject: Re: Who's at fault?

On Sat, 28 Jun 2003 13:43:38 EDT, Dan Margulis wrote:

I would have to agree with others who have posted, that the blame is
primarily yours although the printer shares some of it.

Interesting logic...

The printer has been contracted to produce a piece of work. As part of the contract he has also *chosen* to accept:

1. An RGB file
2. A PSD file that was not ready for print
3. Data from a faulty computer

These are BIG warning signs to anybody who is prepared to give the project a millisecond of thought. Each of these points requires time (and therefore money) to deal with.

Accepting the job without agreeing additional fees would be unprofessional. Accepting the job without the expertise to deal with these issues would also be unprofessional.

But they did accept the job. And then they *allegedly* screwed it up.

The printer needs opprobrium not a sympathetic, minor *share* of the blame.
 
Hence, those who've experienced the problem have found little sympathy here, because
regardless of whether printers *should* respect the Adobe RGB setting, the fact is they
*don't*, and people who expect that they will must accept the blame for the
inevitable result.

If the printer does not have the expertise to handle the RGB data then they should *not* accept RGB data.

They had choices here: honour the profile, use their expertise to apply something better, call the client and ask them what to do - or ignore everything and screw it up.

The actual RGB space used is of no consequence whatsoever. Failure to respect an RGB profile means that the printer is *choosing* to *reinterpret* the image data.

Taking on all of this responsibility and then failing to get the client to sign off a proof before committing the job to press is an exceptionally unprofessional thing to do.

I see no reason for the commissioner of the work to accept responsibility for such unprofessional conduct from a supplier*.

If the printer will only accept plates, then you are being billed only for their press time and expertise at keeping the press running to acceptable standards. If they accept RGB PSD data then, you are being billed for their image manipulation, colour management, page layout, prepress and press expertise. If it ain't up to scratch then don't pay for it.
 
*Apart from the fact that both have opted to remove prepress professionals from the production process and therefore get exactly what they deserve...

--
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
 ________________________________________________________________________

Date: Sun, 29 Jun 2003 20:40:13 -0400
   From: John Castronovo
Subject: Re: Who's at fault?

Well, then he's an idiot or asleep at the wheel, especially since you told him that the file needed conversion. Something like this should be more than obvious on several levels in the work flow without you even having to say anything. It's almost as if he wanted your job to fail.

I'd 'politely' request that he take advantage of his errors and emissions insurance policy to cover the cost of reprinting the job at his expense. A law suit would get you at least half the money back for sure, but it's probably not worth going through the aggravation of the legal mill.

john castronovo
tech photo & imaging
 ________________________________________________________________________

Date: Mon, 30 Jun 2003 08:44:02 -0400
   From: Lee Clawson
Subject: Re: Who's at fault?

In our studio if I say "all design work is done" that means I signed off.
Ask for a proof before making (another) special instruction form.

Lee
________________________________________________________________________

Date: Sun, 29 Jun 2003 21:43:25 -0700
   From: Stuart Larson
Subject: Re: Who's at fault?

Who is at fault here?

This is a tough one, and the answer is, predictably...you probably both are to some extent. I've been on both sides of the coin, as co-owner of a design company and as a prepress guy for the last 13 years.

The printer should have a clearly defined policy of having clients sign off on proofs. They should also generate a proof that would be of good enough quality for you to feel comfortable signing off on it, and they should then be responsible for matching that proof. If you have any doubts or misgivings, you should ask for a press check to be sure you are satisfied before the job is run.

There isn't quite enough information in your post for us to know everything. Is this a printer that you've worked with before? Were your expectations realistic; were they based on past experience with this particular printer?

What you told him ("it just needs converting") could be interpreted many different ways. Converted from one color space to another? From one mode to another? From one file type to another? Obviously that's where the breakdown in communication occurred.

My expectation as a client, and what I do in practice as a service provider, is to call with any concerns to make sure what occurred with your job doesn't happen. It's all about expectations; color is subjective. What is acceptable to one client may not be to another--many clients are very happy with the old "pleasing color". Did you expect him to know what you wanted? Did he think he knew what you wanted? Or do you think he just didn't care enough or didn't have the expertise? Did you abdicate your control by agreeing somehow to bypass the proofing stage? I think these questions need to be answered before a conclusion can be drawn.

Bottom line is...if there wasn't a proof for him to match so he knew what you wanted, you can't fault him. If he didn't match the proof to your satisfaction, contact PIA (if he's a member)--they have arbitration services, so maybe you can get some relief there. If it's just a lousy looking job that you think he shouldn't have run looking that way, stand your ground. And find another printer.

As a printer, it doesn't do us any good if clients aren't happy. We eat jobs occasionally that we don't feel are our fault because it's good customer service...and good business in the long run.

There have been a few angry, or at least indignant, responses assigning responsibility to the printer, but the client does have responsibility in the workflow as well. Not every printer uses color management and many don't know how to handle profiles. Unfortunately, there isn't a clear or published standard--yet--regarding this. People using color management should know this as reality in the industry, and as a practical matter should probably assume the worst. It usually happens.
________________________________________________________________________

Date: Mon, 30 Jun 2003 08:26:21 -0500
   From: Dan Tesch
Subject: Re: Who's at fault?

Oh, come on. I have been watching this thread for a while now and wanted several times to chime in but this is it - to suggest legal action - give me a break!

THE GUY DID NOT SEE A PROOF - TIS BOTH THEIR FAULTS!

CLIENT FOR NOT ASKING, PRINTER FOR NOT INSISTING!

Let them both learn a lesson and go on - shame on you for suggesting legal action - and it is errors and omissions, not emissions.

I'd 'politely' request that he take advantage of his errors and emissions
insurance policy to cover the cost of reprinting the job at his expense. A
law suit would get you at least half the money back for sure, but it's
probably not worth going through the aggravation of the legal mill.

john castronovo
________________________________________________________________________

Date: Mon, 30 Jun 2003 09:47:58 -0400
   From: John Rawlins
Subject: Re: Who's at fault?

Martin Orpen wrote:

The printer has been contracted to produce a piece of work. As part of the
contract he has also *chosen* to accept: 1. An RGB file 2. A PSD file that
was not ready for print 3. Data from a faulty computer
Accepting the job without agreeing additional fees would be unprofessional.
Accepting the job without the expertise to deal with these issues would
also be unprofessional.

If the printer does not have the expertise to handle the RGB data then
they should *not* accept RGB data.

I see no reason for the commissioner of the work to accept responsibility
for such unprofessional conduct from a supplier*.

...If it ain't up to scratch then don't pay for it.

Well... &$?!!*@!!,  I agree with NONE of this. If a designer sends an RGB file with clear instructions to convert to CMYK, and the printers fails to do so, you have a legitimate complaint.  If you send a file with vague instructions to print it with no proof, end of story, you take what you get.

Additionally, there could be more to this story than we are being told. We get very few jobs saying "print my photoshop file". More than likely (just surmising here) it was an RGB image placed into some other layout program. When it was ripped, the RIP made its own conversion to CMYK, hence color that the creator was not happy with. Most RIPS do a "fair" CMYK conversion at best, since we all agree that is not the place it should be done. As a printer, we don't take a job apart piece by piece looking for RGB files, unless we are told to do so. I could have the RIP stop, and not proceed converting and RGB image, but we don't. It's only a problem if the client catches a color problem, IN THE PROOF, that we, or they go back and fix it.

IMHO, this is a clear example of someone wanting the printer to pay for a rookie's error. Most of the files we receive are clean, professionally prepared and rarely require any extra desktop time. We also get of lot of junk thrown at us that we are asked to, and do fix. Then there are the files from the "experts" that think they know everything. These are the ones that have the most problems, complaints, and don't want to pay for something. I had one tell me the other day that "he had been doing this for 2 years" and knew what he was doing. I didn't even respond.

John Rawlins
________________________________________________________________________

Date: Mon, 30 Jun 2003 16:34:19 +0100 (BST)
   From: Alain Corf
Subject: Re: Re: Who's at fault?

 John Rawlins wrote:

IMHO, this is a clear example of someone wanting the
printer to pay for a rookie's error.

On this particular occasion I had no choice but to give the printer an RGB PSD for reasons mentioned earlier in the thread. The images I normally send out are ready to go to press, hence the instruction to convert given to the printer.

I did not give the printer an RGB file thinking it would print OK on press! There was a comminication breakdown.

Alain "The Rookie" Corf
________________________________________________________________________

Date: Mon, 30 Jun 2003 10:37:12 -0500
   From: Al MacDonald
Subject: Re: Who's at fault?

Point, Match & Set!
Dan Tesch is absolutely correct and his point is well said.
Both are at fault.
There is far too much needless legal action taken in this country.
Blame can not fall to either party in this case.
Accept responsibility, resolve it & move on.
But, also let it be a lesson - send good files with a proof or take your chances!
--
Al MacDonald
Shaughnessy MacDonald, Inc.
Elk Grove Village, IL 60007
 ________________________________________________________________________

Date: Mon, 30 Jun 2003 11:45:02 -0400
   From: Michael O'Connor
Subject: Re: Who's at fault?

 It would be nice to know who initiated the communication cited, and whether there was ever any mention, by anyone, that the file was RGB.

It seems from the description of the communication that you were more concerned with the format conversion than the color conversion, leading the printer to associate the problems he was to call about with the format conversion.

Unless it was a very unprofessional shop I can't believe that RGB wasn't mentioned. In my experience, even shops that have established RGB workflows internally want CMYK files when supplied, so without any other specifics I would tend to think that somewhere in the communication you'd said something to the effect that "rgb will be fine", and maybe, "I do it all the time".

That there's no mention of asking to be provided a proof also seems telling, I certainly would never want to go to press with anything I cared at all about without seeing a proof I could trust or, at the very least, knowing the workflow intimately.

From the surrounding circumstances its reasonable to assume your hardware problems had you wait past the last minute, having the job become a rush. If that's true, your priorities would now look to have been time and file format.

With the provided information, I have a hard time seeing the developing situation as one in which the printer would be likely to have any real blame.

Michael O'Connor
 ________________________________________________________________________

Date: Mon, 30 Jun 2003 12:14:18 -0400
   From: "fred"
Subject: Re: Re: Who's at fault?

       As someone who is a printer and also usually just reads this list for the valuable information that appears here, I wanted to give some observations. It seems to me that there is not enough information here to properly decide the issue, however, having been in this business since 1981 and having been involved in digital workflows since the late '80s some things are usually true in printing.
       A customer who gives you a job based on no proofs in the quote, I assume no proofs in the quote because the originator of this thread was not surprised that he was not signing any proofs, means that the quality expectations of this project are usually very low. Any mid to high level quality project would never be quoted without proofs, in our case these include a final trimmed proof and a full color contract proof. Furthermore, most printers will not even quote any job, regardless of quality level without proofs being part of the quote. It is only when a customer tells us directly that they do not want to pay for proofing do we even consider it, and then only after the designer signs a disclaimer that they maintain sole responsibility for content and color. All we do at that point is RIP the file and print to SWOP ranges for color. The type of work that we do in this category is usually low-end direct mail crap that gets tossed out before making it into most people's homes (thank god for our disposable society providing us this employment).
       
Having said all that, our shop would probably reprint the job for this customer at paper cost only or possibly no cost if we felt we could salvage the relationship and move forward together as buyer and vendor. As a printer I would not have allowed this job to process through my shop without a proof once I knew the nature of the files I received. Our goal is to print your project and make you happy with the result, in this case that is not possible without first showing a proof. If the customer refused a proof at this point, I would insist they sign a waiver removing responsibility for quality. If I felt we would have a long relationship with this customer, as I want with all our customers, I would probably offer this proof free or reduced cost to show the importance of the proof, and our ability to match these proofs.
       
Thanks for letting me rant a little, I hope I did not offend anyone with these comments I just wanted to express a viewpoint from a vendor as opposed to a buyer of printing.
                                       Thank you,
                                           Fred Gamber
                                           MagnaniMedia
________________________________________________________________________

Date: Mon, 30 Jun 2003 12:39:00 -0400
   From: John Romano
Subject: Re: Re: Who's at fault?

Printer didnt have OKd proofs and client didnt request to see a proof knowing there would be a RGB to CMYK conversion.

Both are incredibly stupid acts on both behalfs.

So your mac went dowm and you had to send the file RGB, hogwash you could have done your own conversion to a standard CMYK space on any mac and you would have been better off.

The printer could have done a flightcheck to see if there were RGBs before assembly.

So many excuses for this but it comes down to seeing proofs, thats what there for. To catch stupid mistakes, Nothing goes to plate here without an CUSTOMER OKd proof.This eliminates most stupidity.

John
________________________________________________________________________

Date: Mon, 30 Jun 2003 10:03:55 -0700
   From: Ray Maxwell
Subject: RE: Re: Who's at fault?

Imagine the following:

A person sends a sketch of a house to a builder.  The sketch has dimensions in "units".  The designer does not specify whether the units are metric or English.  A unit is an arbitrary length chosen by the designer.  The sketch has no materials specified.  The builder does not talk with the person sending the sketch.  The builder builds the house facing the wrong way on the lot.  The builder builds the house out of very weak materials.  The door is so low the person cannot enter the house.  The first wind that comes along blows the house down.

Whose fault is this?

When we change the context to another industry this whole thing sounds bizarre.  But this is the current state of the art in the printing industry. We do this business by high craft and art.  Every job is custom.  We have not matured to a level where we use manufacturing methods that were developed more than two hundred years ago.  They started building guns with interchangeable parts at that time.  Before that time each gun was hand crafted.  Each job was custom.  Each one was done using high craft and art. No two guns were alike.  You could not move a part from one gun to another.

When will the printing industry move to specifying color on agreed international standards?  When will the printing industry move to tolerance color using agreed upon international standards?  When will we move to an objective way to specify what we want manufactured?  When will we agree on the language of our industry? It is possible to do this today.  The standards exist.  The measurement instruments exist.

"You must send a proof".  This tells the whole story.  We have to send a maquette.  An artists model.  We cannot just send numbers, because the numbers don't have any device independent meaning in our industry.

When was the last time you saw a house builder with a detailed model sitting on the construction site?

We will only change when we adopt manufacturing methods.  We need process control.  We need understanding of the physics of our process at a first principals level.  We need standards for specifying color and specifying the tolerance that is acceptable.  We need to agree on a page description language that covers all of the results that we need to complete the job.

We all are at fault for using methods that are more than 200 years out of date!!!

Now, I will wait for all of the replies that will tell me about the faults with CIELab.  I think these faults are small with respect to the situation described at the beginning of this thread.  I believe that if the designer had a calibrated proofing method in his shop (softproof or hardcopy) and had sent the data in a standard RGB space and had ask for the printer to print to a tolerance of delta E of 5, he would have gotten much closer than he did.

I hope all of us will be willing to spend some time working on standards committees like SWOP, GRACOL, SNAP, ICC, etc.

OK, I will put my soapbox away now and go back to getting some work done.

Ray

The opinions expressed here are solely those of the author and not of the company he works for.

Creo
Ray Maxwell  |  Senior Color Systems Engineer, Inkjet Printing
4225 Kincaid Street   |    Phone (604) 451-2700 ext. 2004
Burnaby, B.C.
Canada V5G 4P5
IMAGINE CREATE BELIEVE
________________________________________________________________________

Date: Mon, 30 Jun 2003 13:46:02 -0600
   From: Chris Murphy
Subject: Re: Who's at fault?
 
On Sunday, June 29, 2003, at 03:59  PM, Alain wrote:

Yes, the printer sent the RGB file to press without
converting to CMYK at all.

If that is really true, then the printer is 100% at fault (again, unless the contract/agreement you signed stipulates otherwise in a case like this). Relying on built-in, generic, in-RIP color management to properly separate a job is asking for disaster.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Mon, 30 Jun 2003 14:07:02 -0600
   From: Chris Murphy
Subject: Re: Who's at fault?

On Sunday, June 29, 2003, at 10:43  PM, Stuart Larson wrote:

Unfortunately, there isn't a
clear or published standard--yet--regarding this. People using color
management should know this as reality in the industry, and as a
practical matter should probably assume the worst. It usually happens.

There are standards committees working on this issue: CGATS (Committee for Graphic Arts Technologies Standards), and the ICC which deals pretty much exclusively with the format of profiles.

Also, the various flavors of PDF/X are published standards and all deal with the color management issue in different ways. PDF/X-1a is CMYK only, fonts and images must be embedded, and an OutputIntent specified. The OutputIntent is either an externally referenced output process (like SWOP/TR001), or an explicitly embedded profile. For PDF/X-3 is device color and device independent color (each object can be defined by a profile), fonts and images are to be embedded, and OutputIntent is to be specified.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Mon, 30 Jun 2003 13:37:53 -0600
   From: Chris Murphy
Subject: Re: Who's at fault?

Who is at fault here?

It's not a good idea for you to approve something for printing with an agreed upon proof, or agreed upon press behavior. You have no idea what you're going to get, and if you enter into a contract to purchase printed product without a proof, or defined press behavior that is very, very risky business.

For the printer to engage in printing without a proof is even riskier because he's usually not asking for upfront payment. It's not reasonable for a printer to acknowledge this is not a prepaid job, implying you will pay upon satisfaction of delivered product, and then say you must pay no matter what.

Conclusion: a failure to communicate by both parties. He probably heard "unless I think looks ugly, print it as is" and you heard "print it unless I think think it look ugly, otherwise call me." The sure form of communication is a proof. If you're going to cheap out on a proof, these are the kinds of things that happen.

Resolution: the fair thing would be for both of you to come to an agreement that neither of you are totally satisfied with. Whatever his actual materials cost is, if you were to pay 1/2 of that would hurt him and hurt you equally and hopefully there would be a lesson learned. In reality, he's mostly on the hook unless you've signed some kind of contract/agreement that stipulates otherwise, in cases like this.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Mon, 30 Jun 2003 13:44:07 -0600
   From: Chris Murphy
Subject: Re: Who's at fault?
 
On Saturday, June 28, 2003, at 11:43  AM, Dan Margulis wrote:

Since then, the consensus on the list
has been that most printers don't even know what an embedded profile is
because they've set up their systems to ignore them completely.  
Hence, those who've experienced the problem have found little sympathy here,
because regardless of whether printers *should* respect the Adobe RGB setting, the fact is they *don't*, and people who expect that they will must accept the blame
for the inevitable result.

It's a classic case of walking in a cross walk, assuming the Mack Truck is going to stop before smearing the pedestrian onto the pavement for the next two blocks. The customer can be in the right, and still get creamed. Don't make assumptions. (And that goes for printers too.)

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Mon, 30 Jun 2003 14:36:43 -0600
   From: Chris Murphy
Subject: Re: Re: Who's at fault?
 
On Monday, June 30, 2003, at 11:03  AM, Ray Maxwell wrote:

We all are at fault for using methods that are more than 200 years out
of date!!!

Pretty much, yeah. But even today there are root beers made in small batches by hand. Regular icky beer too. There will always be room for truly unique services and products.

Now, I will wait for all of the replies that will tell me about the faults
with CIELab.  I think these faults are small with respect to the situation
described at the beginning of this thread.  I believe that if the designer
had a calibrated proofing method in his shop (softproof or hardcopy) and had
sent the data in a standard RGB space and had ask for the printer to print
to a tolerance of delta E of 5, he would have gotten much closer than he
did.

This is asking a lot from a printer that apparently didn't even preflight the job, and ended up submitting it to the whim of PostScript color management in who knows what RIP doing who knows what kind of conversion. It might have taken him months to do delta E calculations on every unique color in a given image.

Also, we should be using delta Ecmc as a minimum. Delta E76 is really easy to use but it's also really flawed (underestimates the difference in neutrals and yellows, underestimates in red and blue.) I've been using delta E2000 lately, which is supposedly even better than cmc.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Tue, 01 Jul 2003 00:04:05 +0100
   From: Martin Orpen
Subject: Re: Who's at fault?

On Mon, 30 Jun 2003 09:47:58 -0400, John Rawlins wrote:

Well... &$?!!*@!!,  I agree with NONE of this. If a designer sends an RGB
file with clear instructions to convert to CMYK, and the printers fails to
do so, you have a legitimate complaint.

Glad we &$?!!*@! agree then.

The client sent an RGB Photoshop file. How are you going to print that without making two or more conversions to the file?

By choosing to accept this type of data you are implying that you have the expertise to offer the additional services of Photoshop work and RGB to CMYK separation.

If you have those expertise then you can usually spot the stuff which is totally out of gamut and make sure that borderline stuff is proofed and agreed before committing to a print run.

You don't just run it and then hold out your hand expecting to be paid.

If you send a file with vague
instructions to print it with no proof, end of story, you take what you get.

Dan's quoted version of the post didn't mention that the client had ordered the print job to go ahead *without* a proof - although the original poster is rather vague on this point.

If he did then that was a stupid thing to do. However, the supplier should have *insisted* on a proof being produced because of the dangers associated with accepting RGB PSD data - and the possibility that they won't get paid if it goes wrong.

Additionally, there could be more to this story than we are being told.

That's why I used the word *allegedly* in my post.

We get very few jobs saying "print my photoshop file". More than likely (just
surmising here) it was an RGB image placed into some other layout program.
When it was ripped, the RIP made its own conversion to CMYK, hence color
that the creator was not happy with. Most RIPS do a "fair" CMYK conversion
at best, since we all agree that is not the place it should be done. As a
printer, we don't take a job apart piece by piece looking for RGB files,
unless we are told to do so. I could have the RIP stop, and not proceed
converting and RGB image, but we don't. It's only a problem if the client
catches a color problem, IN THE PROOF, that we, or they go back and fix it.

IMHO in-rip separations are crap. And I'm dubious that you've got a RIP that can handle layered PSD data. RGB data should be caught during flight checking and the job stopped until the client either supplies new images or pays you to do the conversion. Proofing the job when you know it is incorrect is a waste of everyone's time and a waste of your client's money.

IMHO, this is a clear example of someone wanting the printer to pay for a
rookie's error. Most of the files we receive are clean, professionally
prepared and rarely require any extra desktop time. We also get of lot of
junk thrown at us that we are asked to, and do fix. Then there are the files
from the "experts" that think they know everything. These are the ones that
have the most problems, complaints, and don't want to pay for something. I
had one tell me the other day that "he had been doing this for 2 years" and
knew what he was doing. I didn't even respond.

We'll have to agree to disagree on this. The only things that are *clear* from this scenario are that there are plenty of businesses around that find it easier to proof and print rubbish than to pick up the phone and talk to a client.

If the guy is both a rookie and having problems with his system then one might think that:

1. he deserves a little helpful handholding, or
2. deserves being told politely to take his difficult job elsewhere

Instead, he gets a crap print job, a bill and a bunch of industry professionals snickering that he deserved everything that he got.

Maybe it's just me, but I don't see this working as a sustainable business model. And, as I mentioned in my post, as the prepress industry is disappearing one would assume that printers are going to be dealing with more and more *rookies* with iffy RGB Photoshop images.

You never know, being a little more helpful might be advantageous... Or maybe I'm just a pre-press guy who fancies being a customer relations consultant this week :-)
 --
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
 ________________________________________________________________________

Date: Tue, 1 Jul 2003 00:20:08 -0400
   From: John Castronovo
Subject: Re: Who's at fault?

From: "Dan Tesch"

times to chime in but this is it - to suggest legal action - give me a break!

Hey Dan! Who suggested legal action? I said it was an unwise thing to consider!

THE GUY DID NOT SEE A PROOF - TIS BOTH THEIR FAULTS!

A printer who runs a job and doesn't look at his output is a fool and a hack. Bells should've gone off in his head several times on this one. From the sound of it, the job was a total mess, and now he wants payment. If it were my shop, we'd have to eat the job because none of my customers would pay the bill for a crap. On the other hand, if it were my shop, this kind of thing doesn't happen because we preflight every image and we either tell the customer what's wrong or we charge to fix it. Turning out bad work and blaming the customer isn't an option. Especially when the customer flagged the job with a request for special attention.

CLIENT FOR NOT ASKING, PRINTER FOR NOT INSISTING!

The customer asked for the printer's help. The printer agreed but didn't follow through.

Let them both learn a lesson and go on - shame on you for suggesting
legal action - and it is errors and omissions, not emissions.

Again, who suggested legal action? Sorry for the typo. I do know better - I've been in this sorry business over 30 years. Should be sued for my error on omission?

john castronovo
________________________________________________________________________

Date: Mon, 30 Jun 2003 22:40:08 EDT
   From: Dan Margulis
Subject: Re: Re: Who's at fault?

Martin writes,

These are BIG warning signs to anybody who is prepared to give the project a
millisecond of thought. Each of these points requires time (and therefore
money) to deal with. Accepting the job without agreeing additional fees would
be unprofessional. Accepting the job without the expertise to deal with these
issues would also be unprofessional.

Very correct in principle. In real life, the printer was confronted with a young, inexperienced client (Alain has given me this info offline) who had an emergency over which he was no doubt very upset. The idea that under the circumstances the printer should want to try to help and not want to charge extra doesn't sound unprofessional to me.

And speaking of principle vs. real-life, permit me to quote myself, from "Professional Photoshop 6"

"Every few weeks, some color discussion group features wailing and gnashing of teeth on the part of a user of Adobe RGB who was foolish enough to pass an RGB file on to a service provider who had never heard of Adobe RGB and had all color management turned off, thus guaranteeing a nearly colorless result. The [Conventional Color Management Wisdom] waxes wroth when this occurs. The service provider is called all kinds of names, great sympathy is expressed for the victim, other service providers are warned that resistance is futile, and everyone waits for the next victim to fall into the trap so that the fun can begin again. The practical person, however, accepts the world the way it is. For better or worse most service providers have declined to learn much about this methodology."

This was written three years ago, by which time it had become apparent that the idea of honoring embedded profiles from strangers was going noplace. There was no need to include it in the most recent edition, because, except in the minds of most most extreme of color management zealots, the issue is dead--as many posts to the group have indicated, the topic is so far below the radar screen of most printers that they couldn't even tell you what a profile is.

It's pointless to argue right or wrong here. The facts are, far more so even they were in 2000, that if you expect the service provider to act on an embedded profile, you're asking for what happened to Alain to happen to you. If it *does* happen, then you can have the great fun of blaming the service provider on this list or elsewhere. Personally, I think it's more fun to have the job done right and forget about the politics.

Dan Margulis
________________________________________________________________________

Date: Tue, 1 Jul 2003 01:09:35 -0400
From: John Castronovo
Subject: Re: [colortheory] Who's at fault?

From: Chris Murphy

It's a classic case of walking in a cross walk, assuming the Mack Truck
is going to stop before smearing the pedestrian onto the pavement for
the next two blocks. The customer can be in the right, and still get
creamed. Don't make assumptions. (And that goes for printers too.)

Yes, I feel like I'm about to walk out in front of a truck whenever I deal with a printer too. It shouldn't be that way.

Of course, in this event the dead pedestrian's estate is entitled to an award for his untimely death. In our case, the client's file has been "smeared" and the driver of the Mack truck, our printer, still wants to get paid. Just because most printers don't respect profiles, it doesn't give them the right to print junk. Shouldn't they at least look and question if they're not going to go by the rules of the road?

Then Dan Margulis wrote:

It's pointless to argue right or wrong here. The facts are, far more so even
they were in 2000, that if you expect the service provider to act on an
embedded profile, you're asking for what happened to Alain to happen to you.

But we ARE arguing right and wrong here, aren't we? Why shouldn't we expect printers to respect profiles when they agree to accept RGB files? Everyone agrees that none of this would've happened if the profile had been observed, so why is it right to throw it out?

john castronovo
________________________________________________________________________

From: Loring Palleske
Date: Tue, 1 Jul 2003 09:13:55 -0400
Subject: Re: [colortheory] Who's at fault?
 
Personally it seems pretty simple to me. The printer agreed to do the  work and accepted a RGB file.

It wasn't reasonably close to customer expectation. I wouldn't pay (the  full amount anyways). Try to work out what is fair and if he isn't open  to finding an equitable solution go elsewhere. Don't pay dime one until  you receive acceptable output.

Regards,

Loring Palleske
Creative Imaging
  905.441.2661
 ________________________________________________________________________

From: Andrew Rodney
Date: Tue, 01 Jul 2003 07:16:41 -0600
Subject: Re: [colortheory] Who's at fault?
 
on 6/30/03 11:09 PM, john c. wrote:

But we ARE arguing right and wrong here, aren't we? Why shouldn't we expect
printers to respect profiles when they agree to accept RGB files? Everyone
agrees that none of this would've happened if the profile had been observed,
so why is it right to throw it out?

I agree! It1s a shame so many printers heads explode when they get an RGB file. Worse, so many haven1t a clue how to make a conversion. If they are going to accept RGB data, they better get hip to embedded profiles and how to make a good conversion. Otherwise don1t accept them.

More reason Photographers and the like should take the conversions out of the hands of these people. They know what color the file should be better than anyone pushing files though a system, they deserve to be paid to convert and edit if necessary the file in output space. They are the new group producing the RGB data anyway (the number of images from high end CMYK on the fly scanners is dropping like files and in fact film is becoming a less used media every day, replaced by cameras that only produce RGB data).

Provide the printer a file in output space and just have them crank it out. It1s been done this way for years and years in the photo labs by and large. They don1t open or mess with the files. They are not qualified to anyway. Pass the numbers to the output device and leave the data alone. OR if you are going to mess with the data, you better be dame sure you have a calibrated and profiled display and know how to deal with embedded profiles. Despite what Dan and company will tell us about the failure of embedded profiles, the fact remains that if you get a file which is nothing more than 11s and zero1s and have NO meaning without a profile, you simply have NO business messing or altering with those numbers unless you know what they mean. And they have no meaning without a profile. It1s as simple as that.

Andrew Rodney
http://www.imagingrevue.com
________________________________________________________________________

From: "don reisler "
Date: Tue, 1 Jul 2003 09:23:01 -0400
Subject: Re: [colortheory] Who's at fault?
 
I may be in a somewhat different situation from most of you in that I am a buyer of printing services and could not run a press with a manual.

However, I am in the business of buying things for resale (rare books) and have been doing that for 35 years. I use the same process for buying printing as I do for buying books. I try really hard to work with the seller so that they want to help me; have confidence that I will be a repeat costumer for a fair (not always the lowest) price, and that I will not try to get something extra or a bonus every time we do business. In return, I want them to care that the job be done well rather than tricking me to some obscure specification that gets them off the hook legally but is clearly no good.

I have been using the same printer for several years and most of the press and pre press and driver folks know us and their service has been wonderful. I pay my bills on time and in full. I have had many issues, problems, and technical questions over the past many years but they have been helpful, supportive, and willing to run tests and even send people to my office (home) to help calibrate. When there have been serious problems I have not screamed but asked for help and they have sent staff and management out here and we have gone over press sheets with loups, etc. to reach an understanding of what I want and what they can do.

The point being, think of your printer as part of your team not an adversary. On the other hand, if you are the printer, try to make the client an ally. So much of what I read about this very important thread sounds like war rather than collaborative efforts. In general, I try not to buy from adversaries since it ends up working poorly and whenever there is a problem there is a fight not an attempt to solve it to everybody's satisfaction. Of course, I want the press people to look at the run and if it is obviously wrong, then stop and call me. I always give a 24/7 telephone number and try really hard to convince them that I would rather be called than get a messed up job. I try always to tell folks that "better is better than quicker."

The impression I get is that the relationship between printer and printee is wholly adversarial and take no prisoners. Is that really how it is out there in the world and am I just lucky here in Virginia?

Don
--
Jo Ann Reisler, Ltd.
Fine Childrens and Illustrated Books
Original Illustrative Art
________________________________________________________________________

From: Martin Orpen
Date: Tue, 01 Jul 2003 14:24:34 +0100
Subject: [colortheory] Re: Who's at fault?

on Mon, 30 Jun 2003, Dan Margulis wrote:

This was written three years ago, by which time it had become apparent that
the idea of honoring embedded profiles from strangers was going noplace. There
was no need to include it in the most recent edition, because, except in the
minds of most most extreme of color management zealots, the issue is dead--as
many posts to the group have indicated, the topic is so far below the radar
screen of most printers that they couldn't even tell you what a profile is.

Dan, I'll make a point of reading all you write on Photoshop retouching because I have the utmost respect for your skills and knowledge.

But I take your views on colour management with a large pinch of salt.

Raw RGB data is meaningless. It's like supplying an orchestral score without a time signature or any of the above and below stave instructions that give the music it's feeling and emphasis. Sure, you can make out the tune - but nobody's going to pay to listen to it.

If you only have one output device and you are the final destination for the image then you can be in the comfortable position of printing it and telling the client to screw themselves if they don't like it.

We don't have that option. We output in RGB and/or CMYK and then move the data on to other suppliers.

Closed-loop is not an option for us. We wouldn't use it even if it were because the next supplier in the line will soon spot that we have adapted the image to suit our favourite device (no matter how awful it is) and thereby degraded it for anybody else who needs to use the data.

A simple colour tag means that I see pretty much what the client saw on their screen and get to use it as a basis for my separation and enhancements.

Ignoring it means that I have to re-invent the wheel every time I open a new image. I also needlessly risk getting it wrong and then wasting time, materials and getting into a needless argument at billing time.

What is the point?

My feelings are that many suppliers are reluctant to get into colour management because it requires them to:

1. educate themselves
2. purchase calibration devices
3. maintain strict production controls

Maybe we should meet up at the Heidelberg hall at DRUPA next year and you can introduce me to the press guys that recommend operating their presses without colour management?

You claim: "the topic is so far below the radar screen of most printers that they couldn't even tell you what a profile is".

So what's new? There are loads of idiots around who think that owning a good press makes them a good printer. Criticising and publicising piss-poor printing will get the business moving to people who can do a better job.

--
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
________________________________________________________________________

From: John Rawlins
Date: Tue, 1 Jul 2003 08:37:14 -0400
Subject: [colortheory] Re: Who's at fault?

 I guess since none of us were there as material witnesses to what exactly was said, or took place. Its one of those cases of don't get your eye poked out for the finger pointing ordeals. Sounds like a case for Judge Judy.

And... for all the sayers of "everyone makes proofs"... not always true. Hard proofs at least. We have some very high end 4-color real-estate work we do on a weekly basis. Job comes in at 5:00pm, and we're on press by 8:00pm. We run standard densities, and OK the jobs by the color bars only.  The job delivers to the mail house at 8: 00am the next morning. Usually 8 to 12 different card versions, 15 to 25M per card. There is no time for hard proofs, or time for client approvals. We make hi-res monitor proofs and go. We put a lot of trust in the file provider, and they put trust in us. The rules are simple, they are responsible for supplying CMYK, hi-res, images, bleeds, and limiting fonts to a pre-approved "pool" of about 50 different fonts. We are responsible for trapping, ripping the files, and imposing on the forms for the right card to back up the right card. In over two years of doing this particular job, on a weekly basis, printing over 1000 different cards, we have had 2 cards that had to be reprinted. Both due to RGB images, that the client let slip by, as did we. They paid to have them reprinted. To them, its far cheaper than the costs of contract proofs on everything, and the added time to make them, and get them approved. Certainly not recommended S.O.P. for all everyone, but we do have some clients we can do this with.

John Rawlins
________________________________________________________________________

From: Andrew Rodney
Date: Tue, 01 Jul 2003 08:08:09 -0600
Subject: Re: [colortheory] Re: Who's at fault?

on 7/1/03 7:24 AM, Martin Orpen wrote:

But I take your views on colour management with a large pinch of salt.

Thanks god Chris and I are not the only one1s on the list <g>. Off list, that1s a different story.

Raw RGB data is meaningless. It's like supplying an orchestral score without
a time signature or any of the above and below stave instructions that give
the music it's feeling and emphasis. Sure, you can make out the tune - but
nobody's going to pay to listen to it.

It1s like supplying the orchestral sore upside down and backwards.

The analogy I give to photographers is this. I1m going to supply you with an unlabeled box of 4x5 film in which you will find a single sheet of 4x5 film with no notches. Your job is to expose and process this sheet of film. That1s virtually impossible to do since the user has no idea if the film is Positive or Neg, or what ISO the film is. User has no idea how to expose or process the film. Now label the box 3VPS III2 (the profile) and you can do the job. The label on the box didn1t change the content of the film one lick. Neither do profiles. But they do provide critical information.

If you only have one output device and you are the final destination for the
image then you can be in the comfortable position of printing it and telling
the client to screw themselves if they don't like it.

You have to remember the background of so many who don1t understand profiles which is exactly what you describe. It goes back to that old fashion idea of working by the numbers only. Great if you KNOW the numbers for the device you plan to print to. And in the old days, that was a very viable workflow since there were so few devices (some press in house) that you were going to target. Today that1s simply not viable for a vast number of people. I want CMYK and RGB numbers from a file to output to a press, an Epson, a Lightjet, an Iris etc. No one can juggle all the correct numbers for so many devices. In fact, with the vast number of images coming from digital cameras and end user scanners in RGB, who can possibly control the input numbers without profiles?
 
We don't have that option. We output in RGB and/or CMYK and then move the
data on to other suppliers.
 
Yup. And that1s NOT going to change. If anyone thinks it is, they should also pray for type setting to come back too.
 
Closed-loop is not an option for us.
 
You and a huge number of users. And that grows all the time while closed loop shrinks.
 
A simple colour tag means that I see pretty much what the client saw on
their screen and get to use it as a basis for my separation and
enhancements.

Isn1t that a nice thing! Unless you like to work with totally ambiguous color that doesn1t appear correct anywhere. For those people that like control for themselves only and like to keep things closed loop and behind closed doors, this is scary stuff. For the rest of us, it1s a wonderful way to work.
 
Ignoring it means that I have to re-invent the wheel every time I open a new
image. I also needlessly risk getting it wrong and then wasting time,
materials and getting into a needless argument at billing time.
 
Look, when your job is to produce 2-3 proofs at a clients expense to nail the color, what you propose is heresy!
 
1. educate themselves

What, do something different? Shocking.

2. purchase calibration devices
 
I think that1s a lesser issue since most of these people have decent budgets (Presses and proofing devices are pretty expensive. CMS hardware and software is not in within the big picture.

3. maintain strict production controls
 
Oh boy, that1s a big one!
 
You claim: "the topic is so far below the radar screen of most printers that
they couldn't even tell you what a profile is".
 
You have to listen to this year after year with no statically data to back it up before you start to ignore this. You hear such statements as if they are fact but you never get any empirical data other than what appears to be a feeling or belief. Now, I can tell you that according to a recent report of the TWGA, 63% of graphic arts users are Applying Color Management. Did I pull this out of my-you-know what? No, here's a URL to check out:

http://members.whattheythink.com/news/newslink.cfm?id= 10627

Even if we discount the eye-balling part of the article, we have to consider the displays are calibrated in some fashion and that some descriptor of the numbers have to be in place or what all these guys are seeing is science fiction. The fact that 63& admit to using color management of some kind is telling in itself.

Even if we say that more users are not using CMS than are, the report leads me to believe that the scales are turning the other way, not the opposite as you would believe listening to some non backed-up, "ear to the ground" opinion.

Andrew Rodney
http://www.imagingrevue.com/
________________________________________________________________________

From: Chris Murphy
Date: Tue, 1 Jul 2003 08:51:29 -0600
Subject: Re: [colortheory] Re: Who's at fault?

On Monday, June 30, 2003, at 08:40  PM, Dan Margulis wrote:

Very correct in principle. In real life, the printer was confronted with a
young, inexperienced client (Alain has given me this info offline) who had an
emergency over which he was no doubt very upset. The idea that under the
circumstances the printer should want to try to help and not want to charge extra
doesn't sound unprofessional to me to me.

The printer did not stipulate any requirements to accepting the file. He accepted the file with the condition that it needed to be converted. Any monkey or house plant can do: Image>Mode>CMYK Color, but it takes someone with a certain level of competency slightly higher than a monkey or house plant to ensure a good separation is being made. Since he accepted the RGB file "convert to some crappy CMYK" is inherently not an option, yet this printer took it by submitting the file to the whim of color management in his RIP.

It was dereliction of duty for him to do this, and amounts to sabotage of the job. He could have done no worse had he intentionally or even "accidentally" dragged the whole job through a fifth color made of dog crap.

It's pointless to argue right or wrong here. The facts are, far more so even
they were in 2000, that if you expect the service provider to act on an
embedded profile, you're asking for what happened to Alain to happen
to you.

You know why? Because there is a culture of irresponsibility whenever those responsibilities are not clearly defined and enforced. If a printer were to ignore an embedded font, there would be a conniption fit on all sides and obviously the printer would be at fault. If the printer is going to accept RGB files, they are under obligation to honor an embedded profile. If the embedded profile is wrong for some strange reason, it's the customer's fault, just like if they'd embedded the wrong font.

And in Alain's case, it has nothing to do with embedded profiles. The printer not only ignored the embedded profile, he assigned and used a totally different profile by sending it off to his RIP - where the Adobe RGB profile was ignored, and a generic source and destination were used to convert his job. I don't see how you can blame this on embedded profiles because it's TOTALLY unrelated.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

From: Henry Segalini
Date: Tue, 01 Jul 2003 10:21:34 -0500
Subject: [colortheory] [color theory] - Who's at Fault?

A) One definition of quality is, "Meeting or exceeding customer expectations".  Were the customer expectations fully communicated and understood?

B) One of the participants in this thread referred to "rules of the road".  Does anyone know where can I get a copy of these rules?  How many people know these rules?

C) I feel sorry for those who believe that their relationship with their printer must be adversarial.  As a printer since 1959 (starting in that paragon of friendship, NYC), I can assure you that long-term relationships are not adversarial.  I cannot remember one customer/supplier relationship which has been both long-term and adversarial.  This includes both sub-contractor and customers who have a wide range of expectations and abilities (but excludes unavoidable relationships, as with a monopoly like the old Ma Bell).

Henry Segalini
St Louis  MO  USA
________________________________________________________________________

From: "Annette Murray"
Date: Tue, 1 Jul 2003 11:48:37 -0400
Subject: RE: [colortheory] Re: Who's at fault?

Do you realize how many Photoshop users and prepress personnel do not know how to use Image Mode Assign and Image Mode Convert?

Do you know how many do not even know it exists?!!?

A simple explanation needs to be developed and widely distributed on how to use Image Mode Assign and Image Mode Convert.

This would alleviate many of the "who's at fault issues".

Annette Murray
ANRO Inc.
________________________________________________________________________

From: Dennis Dunbar
Date: Tue, 1 Jul 2003 10:28:09 -0700
Subject: [colortheory] Re: Re: Who's at fault?

Dan wrote:

This was written three years ago, by which time it had become apparent that
the idea of honoring embedded profiles from strangers was going noplace. There
was no need to include it in the most recent edition, because, except in the
minds of most most extreme of color management zealots, the issue is dead--as
many posts to the group have indicated, the topic is so far below the radar
screen of most printers that they couldn't even tell you what a profile is.<<<

I am curious about something here. If accepting files in an RGB space, such as Adobe RGB, with the profile embedded is so far below the radar for printers why is there a very pronounced trend on the part of the magazine publishers to PREFER Adobe RGB?

I participated in the meetings of the DISC committee, a working group of the IDEAlliance - the folks behind the GRACOL specs. The work of this committee was to establish standards for accepting digital files for their editorial content. You can go to http://www.disc-info.org/ and see that Adobe RGB is specified as the preferred color space.

How does that fit with Dan's statements? Seems to me there is a disconnect here, can anyone shed some light on this?

Dennis Dunbar
________________________________________________________________________

From: Henry Segalini
Date: Tue, 01 Jul 2003 13:34:57 -0500
Subject: [colortheory] Who's at Fault?

Dennis Dunbar asked, "If accepting files in an RGB space is so far below the radar for printers...why is there a...trend on the part of magazine publishers to prefer Adobe RGB?"

One possibility:

The last count I heard was that there are about 36,000 printing establishments in the U.S.  Printers with the capabilities to print magazines and major publications are a relatively small number.

From what I've seen of the discussions here, my intuitive feeling is that most on this list will not be going to a printing company capable of printing major publications/magazines.  The "smaller" companies will have as wide a range of capabilities and operating philosophies  as the range of their customers' expectations.

Henry Segalini
________________________________________________________________________

From: Jim Rich
Date: Tue, 01 Jul 2003 15:29:45 -0700
Subject: Re: [colortheory] Re: Who's at fault?

All,

I have been reading these threads for a few days and there must be something in the water that makes everyone so tense.

Many of us on this list can and do make ICC based color management work everyday for photographic and pre-press workflows. And it is a plain fact that Color management works. It is  not fool proof and it is not perfect, but then again, neither is going by the numbers. Color calibration is hard, but profiling had made getting calibrated easier.

Oh yea (for those of you that might not know this), if you use any recent version of Photoshop you are using profiles.
 
Another reality of using profiles is that if you use them with a skilled by-the-numbers-person you get the color optimized quicker and more efficiently.  

A bottom line issue is that, if you get educated in using profiles and then get some experience you can make them work to save a business money.

Printing businesses are in financial straights these days, they should pay attention and learn how to save a buck using profiles.

Based on this thread, I dare say that if the client and printer had a better communication channel and the printer was paying attention to some fundamental profile issues each party would not be bickering over it.

Jim Rich
________________________________________________________________________

Date: Tue, 1 Jul 2003 16:02:55 -0700
   From: Dennis Dunbar
Subject: Re: Re: Who's at fault?

This hits the nail on the head. As an artist, and as the chair of the APA Digital Committee I've been seeing the same arguments and fears being expressed by both the photographer/artists and the printers.

In the "Real World" we're all part of a larger team. Images are not tangible things any more until they are printed. This means that photographer and printer MUST work together or disaster will happen followed very closely by finger pointing and the like.

When there is not time to dot the i's and cross the t's we need to be even more certain that there is clear communication. To what extent could this have prevented the problem in the first place? My bet is the current discussion would be unnecessary if the printer and the client had both taken the steps needed to be certain they knew what needed to be done and what was expected.

Dennis Dunbar
APA Digital Committee Chair

The point being, think of your printer as part of your team not an
adversary. On the other hand, if you are the printer, try to make the
client an ally. So much of what I read about this very important
thread sounds like war rather than collaborative efforts.
 ________________________________________________________________________

Date: Wed, 02 Jul 2003 00:55:00 +0100
   From: Martin Orpen
Subject: Re: Who's at fault?

On Tue, 01 Jul 2003 15:29:45, Jim Rich wrote:
 
I have been reading these threads for a few days and there must be something
in the water that makes everyone so tense.

Sorry if my post appeared *tense* - it wasn't supposed to.

I take on new working methods if they fulfil two simple criteria:

1. Give a better result than we were getting prior to adopting it
2. Make things easier for us to produce a good looking image

Using transparent colour management fulfils both criteria. This is no easy task when we are mixing and matching data from in-house digital cameras, drum and CCD scanners and/or whatever we are sent by the clients.

This stuff can then be proofed in RGB to Pictro or wide gamut inkjet or in CMYK to press-simulating inkjet queues or analogue film and Cromalin.

Colour management makes running all these devices and getting a match between them much easier than it was when we were trying to do this by eye.

Time saved can be spent making that image stand out from the page instead of only having just enough time to meet the minimum standard of a reasonable match.

Dan can call me a zealot if he wants. I seem to remember my boss calling me the same thing when I resigned from a highly paid job in typesetting to set up my own business with a Mac II (with Xpress 1 point something or other and Aldus FreeHand) and a LaserWriter back in 1990.

We really worked by the numbers back then - CORA on Apple II workstations, a display full of green text and waiting for the galleys to be imaged by a Lino 202. <m[x], f[x], h[x], l[x]>The quick brown fox jumped over the lazy dog...

WYSIWIG was treated like ICC Profiles - unnecessary for us real pro's that knew the numbers :-) Shame the clients were so pissed off about being charged for all those wasted galleys when their mark up was out by an em or two but the operator ran it regardless.

Sound familiar?

--
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
________________________________________________________________________

Date: Tue, 1 Jul 2003 19:38:52 EDT
   From: Dan Margulis
Subject: Re: Re: Who's at fault?

Martin writes,

A simple colour tag means that I see pretty much what the client saw on
their screen and get to use it as a basis for my separation and
enhancements. Ignoring it means that I have to re-invent the wheel every time
I open a new image. I also needlessly risk getting it wrong and then wasting
time, materials and getting into a needless argument at billing time. What is the
point?

What your personal workflow choices have to do with the subject of the present thread I don't know. If you're saying that you automatically honor all embedded profiles and execute conversions based on them without the client's knowledge, and then you eat the cost of the job if they don't like the result, that would be relevant.

My feelings are that many suppliers are reluctant to get into colour
management because it requires them to:
1. educate themselves
2. purchase calibration devices
3. maintain strict production controls

An interesting speculation that has nothing to do with the thread. Whether people are educated, or what equipment they own, or how strict their process control is, all have zero relation to the question of whether to respect embedded profiles from strangers. Service providers nearly unanimously have decided not to do so. One has to assume that at least *some* of them are well educated, own calibration devices, and have reasonable QC procedures.

Maybe we should meet up at the Heidelberg hall at DRUPA next year and you
can introduce me to the press guys that recommend operating their presses
without colour management?

To what end? "Color management" is irrelevant to this thread. All printers manage color. The question pertinent to this thread is: do you recommend making automatic conversions based on embedded profiles in files supplied by total strangers? And if you do, do you recommend eating the entire cost of the print run if it turns out that this was the wrong thing to do?

You claim: "the topic is so far below the radar screen of most printers that
they couldn't even tell you what a profile is". So what's new?

Nothing. The idea that embedded profiles are a viable way of exchanging documents between strangers has been a dead issue for at least three years.  The argument might have been made in 1998 and 1999 but the market has long since made its decision. Trying to hide this by saying that "ICC color management" has been widely adopted is to indulge in the Andrew Rodney school of extrapolation. It's as if we were to ask the printers at DRUPA whether they admired President Bush's decision to abstain from alcohol, in view of his past problems with it; and if they all answered yes, to come back and say that they unanimously supported every one of his actions in Iraq.

Responsible printers try to work with the limitations of their clients. Responsible clients try to work with the limitations of their printers. There is no need for this continual debate. The question is not how printer *should* operate but how they *do* operate. Those who are interested in having their jobs screwed up, and having adversarial relations with their printers, do things like surround their images with large black solid areas, or bust small three-color type out of a contrasting background, or stay away from GCR in areas where maintaining neutrality is important, or rely on printers to interpret embedded profiles correctly.

Those who are more interested in getting quality work done in a reasonable timeframe without undue aggravation avoid these practices.

Dan Margulis
________________________________________________________________________

Date: Tue, 1 Jul 2003 19:18:33 -0600
   From: Chris Murphy
Subject: Re: Re: Who's at fault?

  On Tuesday, July 1, 2003, at 05:38  PM, Dan Margulis wrote:

If you're saying that you automatically honor all
embedded profiles and execute conversions based on them without the client's
knowledge, and then you eat the cost of the job if they don't like the result, that
would be relevant.

I feel like we've entered the Twilight Zone. Why is it that absurd ideas like this are relevant, and logical ones like holding parties accountable for their respective responsibilities isn't? Dan, you're basically saying that because most printers don't know what they're doing in regards to RGB and embedded profiles, and even if a printer doesn't  preflight the job but accepts responsibility for printing it, that the customer is the party in the wrong if the job comes out bad because "they should have known better." That's lunacy, and you're only helping to perpetuate such lunacy.

To what end? "Color management" is irrelevant to this thread.

Oh really? YOU were the one who brought it up to begin with by blaming Alain's problem on Adobe RGB being embedded. Apparently embedded profiles are not a color management issue as far as you are concerned.

Nothing. The idea that embedded profiles are a viable way of exchanging
documents between strangers has been a dead issue for at least three
years.

That's insane. It's like saying because only 5% of the market uses PDF/X-1a that it's not a viable way of exchanging documents. You prefer a haphazard workflow where no one has any responsibility at all except to assume the printer is an idiot and the customer needs to have end all be all skill in order to prevent the printer from destroying a job; and if the customer can't do that, then it's their own damn fault because they should have known the printer was just itching to do something stupid. What an amazing set of assumptions and pessimism.

The question is not how printer *should*
operate but how they *do* operate.

Yeah, it's all clear now. You've basically given up on printers and printing. You assume they aren't ever going to change in any appreciable quantity, and yet there are ever increasing numbers of European printers that do follow things like standards and not only use embedded profiles but practically require them (that's where the major push for PDF/X-3 came from). No no, don't hold them responsible for anything they should do or even anything they say they will do; ultimately the printer's customer is totally responsible for how their printer actually behaves. That's simply brilliant.

Those who are interested in having their jobs
screwed up, and having adversarial relations with their printers, do things
like surround their images with large black solid areas, or bust small
three-color type out of a contrasting background, or stay away from GCR in areas where
maintaining neutrality is important, or rely on printers to interpret embedded
profiles correctly.

Everything is a bad habit except the last one by your own admission. You admit they use profiles incorrectly and yet you are unwilling to hold them accountable, instead you expect their customer to shoulder that burden entirely. Crystal clear, Dan.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Tue, 1 Jul 2003 19:32:50 -0600
   From: Chris Murphy
Subject: Re: Re: Re: Who's at fault?

On Tuesday, July 1, 2003, at 05:02  PM, Dennis Dunbar wrote:

My bet is the
current discussion would be unnecessary if the printer and the client
had both taken the steps needed to be certain they knew what needed to
be done and what was expected.

I mostly agree with this in concept, but in this particular case there is that one little detail. The customer qualified their file, although perhaps not fully. But the printer either didn't preflight the job, or did but decided to run it anyway; and the way they decided to run it is widely known on the printing side of things to cause problems.

The customer isn't in a position to tell the printer how to convert the file to CMYK, or the customer wouldn't be providing an RGB image in the first place. It was the printer's choice to use quite possibly the worst method known to convert a file. If the printer had said "oh yeah we can do that, the RIP takes care of them" that's *clear* communication but it's not something the customer is qualified to challenge. The job still would have been a mess. This is a case of a printer simply doing the wrong thing, as the story has been told.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Tue, 1 Jul 2003 23:24:58 -0400
   From: Loring Palleske
Subject: Re: Re: Re: Who's at fault?

There are no shortcuts. Take the time to do it right the first time or you will surely find the time to do right the second time.

A proof would've helped. A lot.

Regards,

Loring Palleske
Creative Imaging
  905.441.2661
________________________________________________________________________

Date: Wed, 2 Jul 2003 02:11:28 -0400
   From: John Castronovo
Subject: Re: Re: Who's at fault?

From: "Annette Murray"

A simple explanation needs to be developed and widely distributed on how
to use Image Mode Assign and Image Mode Convert.
This would alleviate many of the "who's at fault issues".

HA! This is the funniest reply yet.

Most of those who don't already know about it, don't want to. I've had people yell at me for merely mentioning profiles. These are pre-press "professionals" and printers who just want to dig their heels in. They're also the worst hacks that I've ever run into in this business.

You're right, though. Some education would alleviate these problems.

john castronovo
________________________________________________________________________

Date: Wed, 02 Jul 2003 02:22:48 -0400
   From: Michael O'Connor
Subject: Re: Re: Re: Who's at fault?

Chris Murphy wrote:

The customer isn't in a position to tell the printer how to convert the
file to CMYK, or the customer wouldn't be providing an RGB image in the
first place. It was the printer's choice to use quite possibly the
worst method known to convert a file. If the printer had said "oh yeah
we can do that, the RIP takes care of them" that's *clear*
communication but it's not something the customer is qualified to
challenge. The job still would have been a mess. This is a case of a
printer simply doing the wrong thing, as the story has been told.

You're certainly trashing this printer on a very skimpy one-sided explanation of circumstances. While its clear that the printer at least acquiesced to accepting the RGB, its not really clear that he "agreed" to, and its not clear what the full communication between Alain and the printer entailed.

It seems to me the printer most likely went out of his way to rush this job through to completion after Alain's hardware problems saw the job become a rush, and that's commendable.

You can make assumptions about the printer, but you don't know how he promotes himself, or just what level of professional he considers himself to be, he may not normally do any real pre-press at all.

Not knowing RGB workflow, and perhaps never having sent an rgb file to his rip before, doesn't mean he's an inept printer, if the first is true you probably wouldn't be working with him, if the latter is also true you might also be able to start forming a picture of his niche, but that's as far as you can go.

I still hear Alain desperate to get his job completed, expressing concern for the .psd format, and more than likely somehwere saying that the rgb will be fine, he does it all the time. The printer saw no problems when he converted the .psd to a tiff, so didn't see need for a call to Alain. Since it was a rush, Alain seemed unconcerned with the color conversion and didn't request a proof, the printer didn't pull a color proof either. Even if he thought the color was bad when he was on press, he had notihng to base it on, and for all he knew Alain may have been quite satisfied with crap. Telling someone in effect that their color or color sense stinks isn't going to promote your relationship much either.

Maybe he should have pulled a proof anyway, and thought of some diplomatic way of bringing the proof to Alain's attenion without directly impugning the color, but that doesn't open him to the degree of derision he's been receiving. And, if he had done that, Alain saw no problem with the proof, and he wound up missing Alain's time needs, where would that leave them both? If time seemed Alain's primary concern, as I'd infer it would have, the printer probably picked the odds-on right poison.

I do have to say though that your comment that the customer obviously didn't know cmyk or wouldn't have sent a rgb file in the first place seems to make Dan's argument more succinctly than he's made it himself. And if the customer doesn't know cmyk, how exactly will profile to profile conversions ensure that you'll meet that customers expectations?
________________________________________________________________________

Date: Wed, 2 Jul 2003 03:05:44 -0400
   From: John Castronovo
Subject: Re: Re: Re: Who's at fault?

From: "Dennis Dunbar"

current discussion would be unnecessary if the printer and the client
had both taken the steps needed to be certain they knew what needed to
be done and what was expected.

Isn't this what proofs and embedded profiles are for - communication? Ignoring a profile is like ignoring a proof or transparency. It's worse if it's all you have.

john castronovo
________________________________________________________________________

Date: Wed, 2 Jul 2003 03:29:56 -0400
   From: John Castronovo
Subject: Re: Re: Who's at fault?

----- Original Message -----
From: Dan Margulis
The idea that embedded profiles are a viable way of exchanging
documents between strangers has been a dead issue for at least three years.  The argument might have been made in 1998 and 1999 but the market has long since made its decision.

Dan,

I shouldn't give away my trade secrets, but color management has become the single most important edge (and secret weapon) that my shop has against our competition. Just today we landed an account many states away from us because we were able to accurately print digital photos using only the photographer's embedded profiles as a guide. We got it right the first time, too. Soft proof to the monitor, then print finals. Moreover, the job was hyper critical matching of leathers and fabrics.

The shop we took the account from was local to the photographer and our mutual client, but they were a closed loop shop and knew nothing about color management.

Please go right on believing that it's a dead issue, but the fact that we're able to do a great job from over 1,000 miles away and without a proof should tell everyone something.

john castronovo
tech photo & imaging  
________________________________________________________________________

From: John Romano
Date: Wed, 2 Jul 2003 07:54:23 -0400
Subject: Re: [colortheory] Re: Who's at fault?

When we recieve jobs that have supplied Hi res we usually have to open them in photoshop to check size and whatever.So with using embedded profiles this we can convert to our workspace will save atleast one round or two of proofs.
For supplied CMYK  we use thier profile to convert and if there isnt one we assign one that makes the file look good, visually. Then convert to our workspace.
Same goes for RGB files, if its got one we use it and if it doesnt we assign one.
With out Color management this would require rounds of Color correction and proofs.
So whats the big deal, Embedd the profiles when you send out your files.
There are alot more printers using CM than Dan thinks, and alot more using it right !!!
We all need to get on the same bus, communication is key

John
________________________________________________________________________

From: Martin Orpen
Date: Wed, 02 Jul 2003 13:45:36 +0100
Subject: [colortheory] Re: Who's at fault?    

On 1 Jul 2003 19:38:52 EDT, Dan Margulis wrote:
 
What your personal workflow choices have to do with the subject of the
present thread I don't know. If you're saying that you automatically honor all
embedded profiles and execute conversions based on them without the client's
knowledge, and then you eat the cost of the job if they don't like the result,
that would be relevant.

Well let me make it clear to you. My *personal* workflow choices are entirely relevant to this thread because they stop situations arising that cause people to post tales of woe like the one in this thread.

And don't try and put up straw men.

I do not: "automatically honor all embedded profiles and execute conversions based on them without the client's knowledge"

Our workflow is *transparent* so that everybody is aware of what we are doing and why. If a client sends untagged RGB to us then they get a call to discuss their Photoshop settings, image history and an email with a concise guide to Idea Colour Management.

Likewise, if they send tagged RGB that differs significantly from their supplied proofs they'll also get a call asking why and letting them know that we charge to fix this kind of problem. We'll also offer to profile their desktop ink jet and supply simulation profiles and techniques so that they can save time, money and heartache.

Chatting with people on the phone and sending a pre-prepared email takes no more than a few minutes of our time, impresses the hell out of the client and builds a good relationship.

If the printer that you are supporting with your argument did the same as us then then thread wouldn't exist would it?

An interesting speculation that has nothing to do with the thread. Whether
people are educated, or what equipment they own, or how strict their process
control is, all have zero relation to the question of whether to respectembedded
profiles from strangers. Service providers nearly unanimously have decided
not to do so. One has to assume that at least *some* of them are well
educated, own calibration devices, and have reasonable QC procedures.

To the contrary. If the printer in question could answer yes to all 3 points (especially point 3) then I doubt we'd have a thread.
 
To what end? "Color management" is irrelevant to this thread. All printers
manage color. The question pertinent to this thread is: do you recommend
making automatic conversions based on embedded profiles in files supplied by total
strangers? And if you do, do you recommend eating the entire cost of the print
run if it turns out that this was the wrong thing to do?

What a load of bollocks.

Colour management is entirely relevant to the thread. You were whining on about Adobe1998 and the evils of profiles before I came into the discussion.

And your term "automatic conversions" is another straw man. We don't *automatically* convert anything and we won't deal with data from total strangers (unless they agree to pay us additional fees to fix it up) - you can't get away with it in RGB like you can in CMYK.

The argument might have been made in 1998 and 1999 but the market has long since made its decision. Trying to hide this by saying that "ICC color management" has been widely adopted is to indulge in the Andrew Rodney school of extrapolation. It's as if we were to ask the printers at DRUPA whether they admired President Bush's decision to abstain from alcohol, in view of his past problems with it; and if they all answered yes, to come back and say that they unanimously supported every one of his actions in Iraq.

Rubbish and more straw...

A printer may decide that profiles are a complete waste of time. The fact that they own the final device in the chain and will only accept CMYK data allows them to reach and maintain this decision without a problem.

Many UK printers and publications choose this option and stipulate that images, PDFs whatever will be rejected if they are not CMYK and/or contain any of that pesky embedded profile data.

However, if a printer chooses to ignore profiles AND accept RGB data then the results are more interesting and the screw ups frequent - as was (probably) the case in this thread.

And there's no need for extrapolation when it comes to discussing the adoption of colour management. 100 per cent of Photoshop users have been forced to adopt "ICC Color Management" whether they like it or not. What's the alternative, editing pixel values from a CLI? Now that's what I'd call *working with the numbers*.

We choose to work it to mutual advantage. It doesn't appear that the supplier in this thread did.

Responsible printers try to work with the limitations of their clients.
Responsible clients try to work with the limitations of their printers. There
is no need for this continual debate. The question is not how printer *should*
operate but how they *do* operate. Those who are interested in having their
jobs screwed up, and having adversarial relations with their printers, do things
like surround their images with large black solid areas, or bust small
three-color type out of a contrasting background, or stay away from GCR in
areas where maintaining neutrality is important, or rely on printers to interpret embedded
profiles correctly.

Those who are more interested in getting quality work done in a reasonable
timeframe without undue aggravation avoid these practices.

Sound advice - perhaps you should re-read the first sentence and then tell me whether the printer was *responsible*?

I've got a sack of excellent quality spanners in my garage. If I set up business as a mechanic and service your car will you work with my limitations?

Caveat emptor...

--
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
 ________________________________________________________________________

From: Stephen Marsh
Date: Wed, 02 Jul 2003 13:28:46 -0000
Subject: [colortheory] Re: Who's at fault?

John Romano wrote:

For supplied CMYK  we use thier profile to convert and if there isnt one we
assign one that makes the file look good, visually. Then convert to our
workspace.

I was doing a good job of staying out of this thread, until John wrote the above. <g>

John, this may have come up before between us (I seem to recall a similar post in the past, but not the author) - but have you ever considered the possibility that the supplied separation is what _is_ wanted, and that changing the numbers in the file may not be what the supplier wanted? Has a customer ever wondered why their custom black linework plate data is now spread over all four channels and is out of register and fuzzy looking? Do they wonder why they now have black in a certain colour where none was in their original separation?

This is the flipside to the tagged RGB issue. The standard RGB workflow (does such an animal exist?) is to require a description of the files numbers to be kept with the file. In this case both the numbers in the file and the correct profile are needed for accurate use down the production line.

CMYK is a little different, due to it being a final output space for a specific condition. When handing off a _final_ file for output - it might be my task to meet the printers specifications and I may have a special game plan with how I craft a sep. Profiles are being used, but the final file being handed off may not end up with a tagged profile if this is a final layout file. Only the files numbers matter now, they simply get passed on unaltered to the RIP for plotting.

Any digital proofing can simply presume the final output aimpoint CMYK as the definition of my files numbers - the profile is not needed for film or plate imaging in the standard prepress workflow.

Mystery meat CMYK does serve a purpose here, as the file is in output space for this particular setting, the tag does not add extra benefit (the tag or aimpoint is presumed and known by all parties in the loop).

I would not want someone to convert my final CMYK crafted sep based on my tag to thier house profile (which was my aimpoint anyway) - and in the process re-separate my tweaks to the K and other plates. Yes the tag would describe the file, but the SP knows what this description is anyway! By not having a profile, it can be less tempting to change the files numbers. With the profile, it can happen with more ease - but with no profile the conversion source has to be presumed or manually assigned...which is not something that _most_ prepress and printers do, as they traditionally treat a CMYK sep final data and not something to be changed without explicit approval.

In my little world, CMYK is considered final and hands off unless otherwise noted. Unlike RGB, it is final and output ready and not open to further interpretation - the numbers in the file are that way for a good reason.

So we have had a long thread chewing over old ground once more, on the same old RGB workflow issues, which has sidetracked the original issue somewhat, in that RGB would not normally be supplied for this work, but due to unforeseen issues RGB was supplied. I hope I never have to start a thread of my own which laments that I handed off a hand crafted separation which I did not want altered, which was then mangled through an unwanted conversion. I can see just as many issues with this type of workflow as with the RGB tag being ignored.

So whats the big deal, Embedd the profiles when you send out your files.

See above - I do not want you [anyone, nothing pesonal] messing with my CMYK values!!!

I find what the next party requires, and supply that.

If the numbers in my file do not suit the conditions - then, why was I given an unsuitable aimpoint to separate to?

The service provider does not need a profile in the fianal CMYK image - as the SP knows what profile describes the setting that will produce the output - which is what I am working towards as my aimpoint - which was given out as part of the 'material specifications' that I am working toward.

We all need to get on the same bus, communication is key

Agreed, communication is the key - all we need is an _agreed_ method of communication between both parties.

John, the bus you drive and the one I drive are very different - I don't think I even want a test drive [bus being chosen way to handle incoming data].

Stephen Marsh.
________________________________________________________________________

From: Andrew Rodney
Date: Wed, 02 Jul 2003 07:23:51 -0600
Subject: Re: [colortheory] Re: Who's at fault?

on 7/2/03 6:45 AM, Martin Orpen wrote:

Chatting with people on the phone and sending a pre-prepared email takes no
more than a few minutes of our time, impresses the hell out of the client
and builds a good relationship.

Amen to that. That1s what service means. It1s pretty clear the printer that hosed the job discussed in this post isn1t getting much repeat business from the poster.

If the printer that you are supporting with your argument did the same as us
then then thread wouldn't exist would it?

Exactly. But god forbid a profile or CMS workflow actually work or insure color communications. That would mean someone is a 3calibrationist2 which is worse in some people1s minds than being un-American!

Colour management is entirely relevant to the thread. You were whining on
about Adobe1998 and the evils of profiles before I came into the discussion.

I1m not sure how long you1ve been on this list but this has been preached for so many years and is so bogus it1s getting tiring. If Dan keeps saying such things, with no backup of any kind of stat1s other than his partially exposed ear to the ground, eventually someone other than Dan will believe it.

A printer may decide that profiles are a complete waste of time. The fact
that they own the final device in the chain and will only accept CMYK data
allows them to reach and maintain this decision without a problem.

However, if a printer chooses to ignore profiles AND accept RGB data then
the results are more interesting and the screw ups frequent - as was
(probably) the case in this thread.

Exactly. IF said printer can1t deal with RGB to CMYK conversions, simply don1t except RGB. Watch business go down the tubes but that1s not a color management or technical issue at all. Printer in question was out of his league, didn1t proof and didn1t know how to make a conversion. As soon as they opened the file a big flag needed to come up saying 3Game over2.

It1s funny how this list lumps along a few posts a month until some color profile issue is discussed and then it really gets going. Only to have a few dinosaurs claim that profiles are the evil of the world or that no one for years has accepted that profiles are useful for anything other than hosing a job. All kinds of self proclaimed statements like profiles haven1t been on printers radar or that 3the industry2 has accepted this or that. It1s nonsense. It1s nice to see for a change others that are calling this straw man as you call it, exactly what it is. Then silences for a few days so the challenges go unnoticed and back to a limp for a few weeks.

Andrew Rodney
http://www.imagingrevue.com/    ________________________________________________________________________

From: Dan Margulis
Date: Wed, 2 Jul 2003 09:31:11 EDT
Subject: Re: [colortheory] Who's at fault?

Dennis writes,

I am curious about something here. If accepting files in an RGB space, such
as Adobe RGB, with the profile embedded is so far below the radar for
printers why is there a very pronounced trend on the part of the magazine publishers
to PREFER Adobe RGB?

It's certainly true that the industry has been scrambling since 1998 to try to get back to a point where RGB files carried their own meaning and weren't susceptible to disastrous misinterpretation. The link Dennis provides compensates by demanding that incoming files be in Adobe RGB.

That, in itself, is a strong vindication of everything I've been saying over the past half decade.  As Andrew Rodney and others were so fond of saying in 1998, the beauty--nay, the genius--of the ICC revolution is that such fiats are no longer necessary. Every user can have his own version of RGB, for they will all carry their own unambiguous profile. Let a thousand RGBs bloom!

In this paradise, the magazine publisher has no reason to ask us to use Adobe RGB as opposed to TomRGB, DickRGB, or HarryRGB. They're all treated the same, with an unambiguous, colorimetrically correct conversion to whatever the unambiguous, colorimetrically correct output space is.

Back on this planet, such every-man-for-himself workflows have been a resounding flop. So, just about every exchange of RGB files now has to be negotiated. It's a shame that it's necessary just to get us back to where we were pre-1998, but it is, and that's what Dennis's group is doing, to its credit.  

So, instead of giving us the open-format paradise that Andrew and his friends promised us, they *dictate* the input, saying, "We do not accept sRGB or CMYK, leave all color management to the prepress professional since they do this job best." IOW, they are as hostile to embedded profiles, in a different way, as Alain's printer. Their refusal to accept files with different embedded profiles is about as clear an indication of the complete failure of this the-profile-speaks-for-itself workflow as one could ask for.

But, of course, all this has nothing to do with the thread. Handing off an Adobe RGB file to somebody who has specifically told us that Adobe RGB is what they want is one thing, and if anything goes wrong, we can certainly blame Dennis's group. Handing off an Adobe RGB file to somebody who hasn't indicated they know what it is, is quite another. In that case, the chances are the job will get hosed.

Dan Margulis
________________________________________________________________________

From: David Chusid
Date: Wed, 02 Jul 2003 10:24:34 -0400
Subject: Re: [colortheory] Re: Who's at fault?

Every one of Dan's comments about printers respecting embedded profiles has included the word "strangers".  If the printer and the client talk, and develop some mutual respect regarding their respective use of color management techniques, and decide to give embedded profiles a try, and run some tests, and check the results, and decided that all is OK, then they aren't strangers any more.

When I ran a service bureau we couldn't even trust a "stranger's" fonts.  How on earth are we going to trust software as complicated and critical as color profiles, when it comes from sources we haven't verified?

David Chusid
________________________________________________________________________

From: Andrew Rodney
Date: Wed, 02 Jul 2003 08:02:47 -0600
Subject: Re: [colortheory] Re: Re: Who's at fault?

on 7/2/03 7:31 AM, Dan Margulis wrote:

It's certainly true that the industry has been scrambling since 1998 to try
to get back to a point where RGB files carried their own meaning and weren't
susceptible to disastrous misinterpretation.

The industry? Or Dan? Again, I ask that you provide some kind of data to back up such a claim.

That, in itself, is a strong vindication of everything I've been saying over
the past half decade.  As Andrew Rodney and others were so fond of saying in
1998, the beauty--nay, the genius--of the ICC revolution is that such fiats are
no longer necessary. Every user can have his own version of RGB, for they
will all carry their own unambiguous profile. Let a thousand RGBs bloom!

First, there ARE a thousand RGB1s that could bloom from every capture device on this planet. Getting that into a few sets of well behaved RGB editing spaces is a brilliant idea of Adobe1s (and before them, people like Radius for PressView users).
 
In this paradise, the magazine publisher has no reason to ask us to use Adobe
RGB as opposed to TomRGB, DickRGB, or HarryRGB. They're all treated the same,
with an unambiguous, colorimetrically correct conversion to whatever the
unambiguous, colorimetrically correct output space is.

I can take TomRGB, DickRGB, or HarryRGB or for that matter ANY tagged RGB
and get to Adobe RGB or where ever I want. I1m not saying this is a great
idea but it1s certainly far better than unknown RGB which is simply a set of
numbers with no meaning. I still haven1t understood (nor have you ever
clearly or for that matter unclearly) explained why there is any benefit to
mystery meat digital files.

Files can either have a meaning or not. They either have a profile or they don1t. You can either read English or you can1t so why would I send you a letter in German knowing you can1t possibly read it?

Back on this planet, such every-man-for-himself workflows have been a
resounding flop.

Says who? Everyman has been scanning and capturing data for at least a decade expect for that dinosaur model whereby user is supposed to pay for some on the fly CMYK scan to size and do this every time they wish to reproduce the file. That1s a flop Dan. I have the stat1s to back it up too. Do you really think the number of images produced TODAY from RGB producing digital cameras is less than what it was when you were scanning transparencies on a scanner that is likely unable to sale today on ebay?

So, just about every exchange of RGB files now has to be negotiated.
It's a shame that it's necessary just to get us back to where we were
pre-1998, but it is, and that's what Dennis's group is doing, to its credit.

RGB has to make it to CMYK yes? Without any descriptor of RGB how do we get there correctly with visual back up? And what does 3Negotiated2 mean? Pre 1998, everyone1s RGB was totally ambiguous and based on an inherently unstable device (their display) which a fraction of users even calibrated (god forbid we calibrate a device or be known as calibrationists. Let1s not even calibrate those Imagesetter1s or film processors).

So, instead of giving us the open-format paradise that Andrew and his friends
promised us, they *dictate* the input, saying, "We do not accept sRGB or
CMYK, leave all color management to the prepress professional since they do
this job best."

I don1t know who1s saying this but clearly they don1t IF they haven1t a clue how profiles and Photoshop works. I1d far prefer sRGB than mystery meat RGB. I1d rather have Adobe RGB than sRGB but I'll take sRGB if indeed the file is in sRGB and tagged as such. And if it1s NOT in sRGB, I'll see so OR I'll produce a visual match the file as tagged in sRGB (what more can a customer ask for?).

Andrew Rodney
http://www.imagingrevue.com/  
________________________________________________________________________

From: Andrew Rodney
Date: Wed, 02 Jul 2003 07:44:26 -0600
Subject: Re: [colortheory] Re: Who's at fault?
 
on 7/2/03 7:28 AM, Stephen Marsh wrote:

Mystery meat CMYK does serve a purpose here, as the file is in output
space for this particular setting, the tag does not add extra benefit
(the tag or aimpoint is presumed and known by all parties in the
loop).

My question would be, why make it mystery meat and untagged it? Seems like this would make more people who have a clue about profiles suspect the end user is of questionable skill. IF you1ve made a screaming CMYK conversion for my process (you made a custom profile), the only reason I1d NOT embed is if I thought you1d have a hissy fit and hose my job. Again this all boils down to communications. I1d prefer that someone take my CMYK file (in as you point out output space by nature), not mess with it and simply print the job. IF they suspect something is up, make a phone call. Or I as a custom could simply instruct the printer to print the file and not touch it. Seems pretty easy to me. The question is then, why not have the profile?

Andrew Rodney
http://www.imagingrevue.com
________________________________________________________________________

From: Stephen Marsh
Date: Thu, 3 Jul 2003 01:03:41 +1000
Subject: Re: [colortheory] Re: Who's at fault?

on 7/2/03 7:28 AM, Stephen Marsh wrote:

Mystery meat CMYK does serve a purpose here, as the file is in output
space for this particular setting, the tag does not add extra benefit
(the tag or aimpoint is presumed and known by all parties in the
loop).

===

Andrew Rodney replies:

My question would be, why make it mystery meat and untagged it? Seems like
this would make more people who have a clue about profiles suspect the end
user is of questionable skill. IF you've made a screaming CMYK conversion
for my process (you made a custom profile), the only reason I'd NOT embed is
if I thought you'd have a hissy fit and hose my job. Again this all boils
down to communications. I'd prefer that someone take my CMYK file (in as you
point out output space by nature), not mess with it and simply print the
job. IF they suspect something is up, make a phone call. Or I as a custom
could simply instruct the printer to print the file and not touch it. Seems
pretty easy to me. The question is then, why not have the profile?

Andrew, it may end up in the hands of someone who thinks it is a good thing to convert to their output space, based off my profile (even though my aimpoint is the same as the one they are separating to). You know who you are. <g>

As stated, all parties concerned know the separations aimpoint. The output folk dictate it - and I meet it. Call me old fashioned, but I still believe in working with the next step in mind, rather than forcing my own way on others down the line (which will not work).

To convert my final CMYK data would then either take a conscious step via assign profile, or the source would be presumed as I chose not to describe it.

True, by not describing the CMYK values, I have probably hosed the unwanted conversion and colour will be very different due to the presumed CMYK source being way off.

By tagging the CMYK profile,  the unwanted conversion would be a very close or same match for colour - but the actual colour builds of the plates and the GCR ratio and special K plate tweaks would be lost.

Either way an an unwanted conversion has taken place, hosing my files wanted values.

In my experience, prepress and printers who receive tagged CMYK shrug their shoulders and wonder why? Tagging the CMYK would make many of these dinosaurs suspect that the originator did not know what they were doing, the opposite of what you propose. Wether or not that is for right/wrong/true/false is not the point - just some of the attitudes I have seen. I have also seen those who understand ICC colour management for the tool that it is and who have a knowledgeable and balanced view. I have also seen the ignorant and the confused well meaning users with crazy settings and methods.

Let me ask this question in return (not directed at Andrew, unless he wants to have a shot at it), although Chris Murphy has stated on this list one answer to this question in the past...

Why did the authors of Real World Colour Management choose not to tag thier CMYK press ready files for their book? If anyone was going to champion this effort, one would think it would have been them. What better way to show those who see no need for a description which is known to all parties to be tagged to the file, than to do it on the live print production of the book devoted to real world use of colour management.

Could it be in the real world, that real, real world colour management practitioners see no need for tagging every image with the same output description, for a output process which does not require said description - and even if it is needed at some interim point (proofing), that description is known by all parties in the production loop anyway?

Stephen Marsh.
________________________________________________________________________

From: Les De Moss
Subject: [colortheory] Re: Who's at fault?

Michael O'Conner Writes:

Even if he thought the color was
bad when he was on press, he had nothing to base it on

...except perhaps years of experience. In our commercial photo lab, we've been printing color negatives for 30 years with little basis for quantifying 'correct color' except experience viewing printed images. Grossly bad color hits you in the face; An untrained eye can usually spot it, a trained eye can define and correct it.

If we were to dismiss color we thought was 'bad' because we lacked a reference, we'd have been out of business years ago. Color professionals have a reference; it's called experience... the stuff we used to rely on before digital came along.

That's not to say that this particular job was grossly misprinted, for all we know the two parties are splitting hairs over minutiae... some clients require valium over 2 points of cyan. But from the sound of it, what came off the press should have raised an eyebrow, regardless of the whole set of circumstances including turnaround, profiles, RGB... etc.

You're certainly trashing this printer on a very skimpy one-sided explanation of circumstances.

I totally agree. The initial question "Who's at fault" cannot possible be answered without input from both parties and a sample of the printed job (Alain, if you have a website, why not scan and post?). The benefit of this entire thread is in uncovering and understanding the dynamic between buyer and vendor when things go wrong.

Les De Moss
DigiGraphics
________________________________________________________________________

From: Henry Davis
Date: Wed, 02 Jul 2003 11:21:12 -0400
Subject: Re: [colortheory] Re: Re: Who's at fault?

Dan Margulis wrote:

"In this paradise, the magazine publisher has no reason to ask us to use Adobe
RGB as opposed to TomRGB, DickRGB, or HarryRGB. They're all treated the same,
with an unambiguous, colorimetrically correct conversion to whatever the
unambiguous, colorimetrically correct output space is."
 
I would like to add, that by calibrationist's admission, there are a plethora of possible 4-color profile options to consider.  Possibly one for every substrate/inkset combination and finishing option.  Add to this profile "tweaking" to suit the unknown and subjective tastes of print buyers who are all way too behind schedule to be concerned with all of the niceties of print paradise.

On the other hand, there may be nothing but a lack of will, to stop print customers from contracting the services of color management consultants and obtaining a profile made for a particular ink/stock combination, and tweaked according to their taste. Then miscommunications about color expectations can be discussed between them.  Meanwhile, the print shop comes up to density and prints more jobs.

Henry Davis
Former Stripper
________________________________________________________________________

From: Stephen Marsh
Date: Wed, 02 Jul 2003 15:25:40 -0000
Subject: [colortheory] Re: Who's at fault?

Andrew Rodney wrote:

IF you1ve made a screaming CMYK conversion
for my process (you made a custom profile),

A custom or more commonly, an Adobe v2 profile used as a 'standard' aimpoint is the base level to getting a screaming CMYK conversion.

It is what one does post separation that counts - using all those wonderful tools and methods available in Photoshop (which is what I would prefer this list concentrate on rather than finger pointing).

Stephen Marsh.
________________________________________________________________________

From: Dan Margulis
Date: Wed, 2 Jul 2003 10:59:58 EDT
Subject: Re: [colortheory] Re: Who's at fault?

Andrew Rodney writes,

It1s funny how this list lumps along a few posts a month until some color
profile issue is discussed and then it really gets going.

It is funny, at that.  I much prefer the lower volume of messages, myself, and as nearly as I can tell most of the group does as well.

Dan Margulis
________________________________________________________________________

Date: Wed, 2 Jul 2003 10:34:57 -0500
From: Bob Smith
Subject: Re: [colortheory] Re: Who's at fault?

Andrew Rodney wrote:

My question would be, why make it mystery meat and untagged it? Seems like
this would make more people who have a clue about profiles suspect the end
user is of questionable skill.

Because if it's tagged its far more likely that someone will make an unintended conversion.  For example, I may have done a separation with a custom black build that will print beautifully on SWOP.  I may send that file out with a profile for my particular separation parameters to a shop that prints SWOP and could print my file as is just fine.  However if they have some house profile that describes SWOP with a different type of black generation and they happen to convert my file from my SWOP to their SWOP, they've just hosed my custom separation and altered the intended look of my file.

With CMYK its possible to have any number of profiles using different recipes for building the separation that all describe the same press characteristics.  The choice of which one to use is based on image content, not press characteristics.  If an unintended conversion takes place between those profiles, the carefully crafted custom seps are hosed.  The image will likely still print reasonably well, not horribly, but not like the person who created the seps intended.

I used to send out everything tagged figuring what harm could it possibly do. Then I got burned a couple of times in exactly the situation described above.

I'm right with you on your arguments for embedded profiles in RGB files where conversions down the line are inevitable and wanted.  I want that tag there so any conversion will be made properly.  But if I'm sending out CMYK files, ready to print, including a profile just adds to the probability that someone will act on it improperly.  A CMYK file with no profile will generally be left alone.

Bob Smith
________________________________________________________________________

From: Alain Corf
Date: Wed, 2 Jul 2003 17:08:25 +0100 (BST)
Subject: [colortheory] When good jobs go bad

Michael O'Conner Writes:

(Alain, if you have a website, why not scan and
post?).

I wish I could but unfortunately the material is copyrighted and I cannot reproduce in any way, shape or form... sorry! This is not BS, I would quite simply lose my job.

The colour was MILES off! The reprint is acceptable. We paid half the cost of the job again. I think this is fair since it was a communication error (amongst other things) and it is/was both of our jobs to communicate effectively.

BTW, the printer had PS set up to ignore profiles completely and didn't get a warning when he opened the file. I have looked at the colour settings in his PS, he didnt seem aware of the fact that these settings existed. He did say that he normally uses PDF to exchange files and they do not normally EVER use proofs to preview a job for the client.

Alain Corf
________________________________________________________________________

From: Dan Margulis
Date: Wed, 2 Jul 2003 11:59:35 EDT
Subject: Re: [colortheory] Re: Who's at fault?

Martin writes,

Well let me make it clear to you. My *personal* workflow choices are
entirely relevant to this thread because they stop situations arising that
cause people to post tales of woe like the one in this thread.

Your personal workflow choices will answer,  among other things, the following:

  1) Can you be relied on to process a QXP/6 file from a PC?
  2) How about an InDesign 2 file containing transparency, any platform?
  3) How about a file that requires you to make an accurate conversion based on an embedded RGB profile?
  4) How about a file that contains lots of instances of a particular Multiple Master font?
  5) How about a file prepared in CorelDraw 8 with lots of nested gradients containing fonts (provided) from an unknown foundry?
  6) What about if any one of these jobs comes in on an 88mb Syquest disk?

We know, because you've told us, that you definitely can deal with #3. Maybe you can deal with all five of the others as well. But suppose we have such a job, in one of the other five categories, that's due tomorrow morning.

Anyone sending such a job to you without first inquiring about your fitness to handle it, needs a checkup from the neckup.

All six of these items are nonstandard services--most service providers are going to answer yes to some of them, but for any one of the six, well over 50% will answer no. People who simply assume that a provider is going to be able to cope with any particular one of them are asking for the result Alain got.

You've made your choice about which services to support and how seriously, and I hope the choice works well for you. However, different firms make different decisions, ordinarily based on what their clients need. The printers I speak to here are becoming seriously concerned about their ability to handle InDesign files because they are seeing increased demand from clients, therefore they are diverting resources to train personnel to be able to handle the problem. They see zero client demand to deal with any profile-related issue, therefore they are disinclined to learn anything about it.

If the printer that you are supporting with your argument did the same as us
then then thread wouldn't exist would it?

Of course not. It would be put off for another three months until the next person who doesn't know that printers and profiles don't mix runs into a similar problem and writes to tell us about it.

What a load of bollocks. Colour management is entirely relevant to the
thread. You were whining on about Adobe1998 and the evils of profiles before I
came into the discussion.

You are apparently mistaking someone else's posts for mine. As anyone familiar with my writings will tell you, and as the archived posts will confirm, I have nothing against Adobe RGB, and recommend it in certain settings in my book. However, I do issue the strong warning that Adobe RGB is unique among the four majors in that if it is misinterpreted the result is likely to be really, really bad. And I say, even in the latest edition of my book, "For work primarily aimed at non-Web RGB: If you are certain that your workflow won't let anyone convert (or fail to convert) it improperly later, use Adobe RGB. If not, use ColorMatch RGB."

As for the evils-of-profiles drivel, I have always advocated that service providers take account of incoming profiles and make intelligent decisions of whether to use them or not. Thus, if the job had come to a company I was running, it wouldn't have been screwed up, either.

But, as we all know, service providers have by and large not taken that advice.

Again: this thread has nothing to do with color management. It has to do with common sense. Forget the way the world *ought* to work. Concentrate on the way the world *does* work.

In the archives of this list, the ColorSync list, in my book, and in various other publications, you will see warning after warning after warning that giving a job that depends on correct interpretation of a profile to an unknown service provider is a recipe for disaster. It's gotten to such a point that even Andrew Rodney, as CMYK-phobic a person as there is on this planet, advocates converting files to CMYK rather than depending upon an unknown service provider to do it. Me, I always give either untagged CMYK or LAB and recommend that others do so as well.

But some seem to feel as a matter of principle that they must court disaster. They declare, on principle, that it is absolutely the service provider's obligation to know what to do with a tagged file, and that there is no need to take such an easy insurance policy against his not knowing. Thus, the parade of principled people with ruined jobs complaining to this and other lists.

I've said enough on this topic now, so before leaving the thread, I offer the following to those who insist that strangers convert their files properly:

"He was right, dead right,
   As he sailed along,
And he's just as dead now,
   As if he'd been wrong."

Dan Margulis
________________________________________________________________________

From: Chris Murphy
Date: Wed, 2 Jul 2003 10:04:07 -0600
Subject: Re: [colortheory] Re: Who's at fault?

On Wednesday, July 2, 2003, at 07:28  AM, Stephen Marsh wrote:

John, this may have come up before between us (I seem to recall a
similar post in the past, but not the author) - but have you ever
considered the possibility that the supplied separation is what _is_
wanted, and that changing the numbers in the file may not be what the
supplier wanted? Has a customer ever wondered why their custom black
linework plate data is now spread over all four channels and is out
of register and fuzzy looking? Do they wonder why they now have black
in a certain colour where none was in their original separation?

This is why repurposing CMYK content needs a DeviceLink profile (a class of ICC profile) instead of two Output Device profiles. Black only can be retained, or even scaled (compensating for dot gain between the two black channels), preserve channel purity, etc.

Profiles are being used,
but the final file being handed off may not end up with a tagged
profile if this is a final layout file. Only the files numbers matter
now, they simply get passed on unaltered to the RIP for plotting.

Not quite. BOTH the file numbers and color appearance matter.

See above - I do not want you [anyone, nothing pesonal] messing with
my CMYK values!!!

If the job wasn't separated for the specific print condition, yes you do. You just don't want certain aspects of the data changed, such as black only drop shadows, or text made out of 100% yellow, or 80% black. But if it's necessary to print 65% black to get the same *appearance* you had at your value, what do you care so long as it's still black only? You'd most likely prefer the result of the scaled black.

If the numbers in my file do not suit the conditions - then, why was
I given an unsuitable aimpoint to separate to?

If you're using their profile, and it gets embedded, there will be no repurposing.

The service provider does not need a profile in the fianal CMYK
image - as the SP knows what profile describes the setting that will
produce the output - which is what I am working towards as my
aimpoint - which was given out as part of the 'material
specifications' that I am working toward.

If the job order stipulates that you are responsible for having converted everything to a specified profile, provided from the printer - that may be an adequate substitute for some printers. For others,
they'll want to see it embedded. What do you care whether it's embedded or not, so long as you know they're going to treat it correctly?

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

From: Chris Murphy
Date: Wed, 2 Jul 2003 10:07:41 -0600
Subject: Re: [colortheory] Re: Re: Who's at fault?

On Wednesday, July 2, 2003, at 12:22  AM, Michael O'Connor wrote:

You're certainly trashing this printer on a very skimpy one-sided
explanation of circumstances.

Please try not to exaggerate my position. For all practical purposes, it's a hypothetical situation because we don't have both sides of the story, and we don't necessarily have the full story from even one side. But as it was presented, which the discussion has already been qualified as being in the context of, and as I also reiterated "as the story has been told", criticism of the actions of this hypothetical printer is not without merit. There are all kinds of questions I have about the actual event, but that information simply isn't available. And it is absolutely possible the printer handled this logically in all other faucets, except where it comes to insisting on a proof.

While its clear that the printer at least acquiesced to accepting
the RGB, its not really clear that he "agreed" to, and its not clear what the
full communication between Alain and the printer entailed.

His agreement is implicit by not having rejected the job prior to printing it.

Not knowing RGB workflow, and perhaps never having sent an rgb file to his rip
before, doesn't mean he's an inept printer,

Having never done so, and subsequently going to press without a proof on top of it does demonstrate a certain lack of understanding of the potential consequences.

I still hear Alain desperate to get his job completed, expressing concern for
the .psd format, and more than likely somehwere saying that the rgb will be
fine, he does it all the time. The printer saw no problems when he converted the
.psd to a tiff, so didn't see need for a call to Alain.

Why would there be a problem converting PSD to TIFF? The concern should be with RGB to CMYK, not PSD to TIFF.
Since it was a rush,
Alain seemed unconcerned with the color conversion and didn't request a proof,
the printer didn't pull a color proof either. Even if he thought the color was
bad when he was on press, he had notihng to base it on, and for all he knew
Alain may have been quite satisfied with crap. Telling someone in effect that
their color or color sense stinks isn't going to promote your relationship much
either.

Maybe that happened, but the logic fails in numerous places because the printer is making all kinds of assumptions in your story that are known to cause problems between printer and customer.

Maybe he should have pulled a proof anyway, and thought of some diplomatic way
of bringing the proof to Alain's attenion without directly impugning the color,
but that doesn't open him to the degree of derision he's been receiving.

If he *really* sent an RGB file to a PostScript RIP, which subsequently used a generic source and generic destination profile, he really does, but hopefully he learns from the mistake. And it is a mistake, not proofing the job, and then insisting the customer pay for the press run only makes it a bigger one.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

From: John Romano
Date: Wed, 2 Jul 2003 12:59:56 -0400
Subject: Re: [colortheory] Re: Who's at fault?

Stephen, and Chris

I agree that device links are good when it comes to managing plate ready files.

You also need an application that can use Device links and one to create a good one that doesnt reseparate through lab.

But for loose color that the client wants to see proofs on that are supplied from god knows where I prefer to manage them in Photoshop. If I get a sep that was prepared for 20% dot gain it would be flat and washed out, if  i have a profile to convert to our specs it  gets the right numbers for our press conditions.

Our gain is much less even on our web presses.

Steven you know better than to think you know how every printer is going to print.

Every place is different and everyone has a different flavor of swop, different ink sets and so on. So letting a printer convert your file would be a good thing, my opinion.

Our customers like the way we are handling the files and cant believe the difference form where they did the job the last time.

We have seen so many times if a job goes through not colormanaged the proofs get marked up so bad thet you end up starting over by assigning a profile and converting to our workspace and it gets you 90% if not 100% there.

Color management can be good for everyone, you just cant ignore it and pretend it doesnt work, like dan would like you to believe.

John
________________________________________________________________________

Date: Wed, 02 Jul 2003 09:57:02 -0600
   From: Andrew Rodney
Subject: Re: Re: Who's at fault?

on 7/2/03 9:03 AM, Stephen Marsh wrote:

Let me ask this question in return (not directed at Andrew, unless he wants
to have a shot at it), although Chris Murphy has stated on this list one
answer to this question in the past... Why did the authors of Real World
Colour Management choose not to tag their
CMYK press ready files for their book?

From what I know from Bruce, the number of images and the size of each profile, the resulting storage of the book would have been much larger.

I this situation, I have no problem NOT embedding a profile. In fact, if I KNOW a printer is simply going to send the numbers in my file to their output device without messing with it, there1s no reason to embed a profile what so ever. I1m not against this one bit. The file is in the correct output space and the profile (descriptor of the data) doesn1t serve any purposes. My question would be in a case where you do NOT know if the printer will honor the numbers, is there a reason not to embed the profile? I can1t think of one off hand other than saving 500K to 1mb per file. That1s viable but dangerous if you suspect the printer might try to open and worse convert the data.

In the case of the book, there was no question that the authors had control over the process whereby they knew the numbers were correct for print and going to that output device with no other alterations of the data. No descriptor was needed because the only people who would open, view, correct or convert the data knew the right answer (what the profile should be) and they knew those numbers were not going to altered.

Could it be in the real world, that real, real world colour management
practitioners see no need for tagging every image with the same output
description, for a output process which does not require said description -
and even if it is needed at some interim point (proofing), that description
is known by all parties in the production loop anyway?

Yes. The profile as we know is only a label. So it serves no purpose in this situation. In fact, if I were 3king of the world2 and mandated that every printer would simply send all the numbers in a file to the output device it was intended for, we would not need to embed a profile. However, some printers feel they need to view the file or do some kind of conversion to the existing numbers. In such a case, unless provided with a profile, they simply can1t do this without guessing. Why guess?

One could embed one profile in one file (say a representative image among hundred of others) and leave the others untagged for space and at least the person on the receiving end could open the file, view it, extract the profile if necessary and use it for the other files. But again, if the numbers are simply going to the device with no other human intervention, the profile serves no use in this case.

Andrew Rodney
http://www.imagingrevue.com/
________________________________________________________________________

Date: Wed, 2 Jul 2003 10:42:34 -0600
   From: Chris Murphy
Subject: Re: Re: Who's at fault?

On Wednesday, July 2, 2003, at 09:34  AM, Bob Smith wrote:

Because if its tagged its far more likely that someone will make an
unintended conversion.

Only by a shop that doesn't know what it's doing. I'd like to hear anyone demonstrate how unintended conversions occur otherwise. How is it more likely, let alone far more likely, that a shop will convert ONLY BECAUSE there is an embedded profile? Why isn't it just as likely that an untagged image will have a source profile assumed, and then converted to some other destination?

I do see a problem with *intended but undesired* conversions due to embedded profiles - that's a discussion about automation that isn't well suited to this forum.

However if they have some house profile that describes SWOP with a
different type of black generation and they happen to convert my file
from my SWOP to their SWOP, they've just hosed my custom separation and
altered the intended look of my file.

How, specifically, would this occur *unintentionally*, and do you have any data to show how often this happens? I have heard only 3rd hand stories as though they were legend, and most of them were from Dan Margulis, yet he hasn't ever described how this happens or shown data with a sufficient sample size to support the contention.

If an unintended conversion takes
place between those profiles, the carefully crafted custom seps are
hosed.

Not with DeviceLinks, that's the point of using them and why their use is on the increase.

I used to send out everything tagged figuring what harm could it
possibly do. Then I got burned a couple of times in exactly the
situation described above.

And you called them on this, and they acknowledged there was an unintended conversion of your file, and acknowledged responsibility? What else happened?

I'm right with you on your arguments for embedded profiles in RGB files
where conversions down the line are inevitable and wanted.  I want that
tag there so any conversion will be made properly.  But if I'm sending
out CMYK files, ready to print, including a profile just adds to the
probability that someone will act on it improperly.  A CMYK file with
no profile will generally be left alone.

I disagree. It's just as possible something will be assumed as source, as someone erroneously using your profile as unintended permission to convert your file.

Summary: It's logical to not embed a profile in a CMYK image, if you assume that whoever you send it to won't know what they're doing when it comes to color management, and has improper settings that would lead them to make an unintended conversion of your document only because it has a profile embedded in it.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Wed, 2 Jul 2003 10:51:24 -0600
   From: Chris Murphy
Subject: Re: Re: Re: Who's at fault?

On Wednesday, July 2, 2003, at 07:31  AM, Dan Margulis wrote:

 The link Dennis provides
compensates by demanding that incoming files be in Adobe RGB.

If you look at the members, it's pretty easy to see why they might do this. They are very likely looking to engage in a streamlined, normalized, single assumed profile workflow. It's probably so they can automate separations.

You're taking one example, and blowing it out of proportion as though it's representative of a larger market. You cannot prove this.

Their refusal to accept files with different embedded
profiles is about as clear an indication of the complete failure of
this the-profile-speaks-for-itself workflow as one could ask for.

Under "What is DISC?" they say: "We are recommending the Adobe RGB 1998 color space because it is large enough to encompass most digital capture, display, and output devices." There it says recommended, elsewhere it sounds like a requirement. I'll see if one of their contacts can clear up the apparent contradiction, and why they would require a particular space rather than merely highly recommending a certain space. (It's probably for ease of workflow, to avoid questions, to avoid having to train potentially tons of people, to avoid getting potentially marginal or jacked up source profiles, on and on...)

But, Dan, you love to make assumptions. You're selectively skeptical, Dan. The refusal to accept files other than Adobe RGB files *including CMYK* is a mostly clear indication of an intent to automate some aspect of workflow regarding *specifically* digital images. Their documentation even makes it more specific that they are referring to digital camera images vs. stock photography which is also digital.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Wed, 2 Jul 2003 13:36:47 -0400
   From: John Romano
Subject: Re: Re: Who's at fault?

Stephen

When you call a printer and ask for their specs wouldnt it be easier if they just gave you their profile ? Maybe Im too optomistic here but there are more printers using profiles than not atleast in high end printing. Like someone said ealier that CM works and helps to save money so I would figure that all of the Major printers are into it in the US anyway not sure overseas yet.
 
Sure you can color correct and image with out using a profile to profile conversion and do a good job but somewhere you need to see proofs, one to see where you started and one to see where you are.Proofs arent cheep right? Just that alone with the time it takes to make one is enough to get you thinking about it anyway.

 m trying to get Gretags IQ140 so we can start managing plate ready files, it has helped so much we want to take everything into a Color managed workflow.

John
________________________________________________________________________

Date: Wed, 2 Jul 2003 11:44:41 -0600
   From: Chris Murphy
Subject: Re: Re: Who's at fault?

On Wednesday, July 2, 2003, at 09:03  AM, Stephen Marsh wrote:

Why did the authors of Real World Colour Management choose not to tag their
CMYK press ready files for their book?

It had to do with size of profiles magnified by the number of images, not due to some neurotic concern that the printer was going to convert all of our files because a profile was embedded in them. Plus the specific destination (press, ink, paper, etc.) was known and was profiled. Since that was the only possible destination, and because the submission of final press ready material was our responsibility per the contract, it simply was not necessary.

Embedded profiles are not a universal requirement of color management in all situations. They are an aspect of it, and a tool in the arsenal.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Wed, 2 Jul 2003 11:21:06 -0600
   From: Chris Murphy
Subject: Re: Re: Who's at fault?

On Wednesday, July 2, 2003, at 09:59  AM, Dan Margulis wrote:

People who simply assume that a provider is going to be able
to cope with any particular one of them are asking for the result Alain got.

He didn't assume! He told the printer it was RGB from the get go and the printer accepted the job. The assumption as on the part of the printer who took the job. Assuming it would be OK, or assuming Alain would accept color miles off the mark.

They see zero client demand to deal with any profile-related issue,
therefore they are disinclined to learn anything about it.

They probably don't know the first thing about it, and if they do ask presumably the printer tells them its all about the proof and if there are any problems, that's when everyone finds out and then gets them fixed. Customers are used to these kinds of iterative proof-fix-proof-fix workflows and don't know any better to be able to say anything about an alternative let alone complain about it. It's
merely part of a traditional workflow.

You are apparently mistaking someone else's posts for mine. As anyone
familiar with my writings will tell you, and as the archived posts will confirm, I
have nothing against Adobe RGB, and recommend it in certain settings
in my book.

Dan, you are the Iraqi Information Minister!!!! You wrote, barely two days ago:

"The culprit is invariably Adobe RGB. I don't favor that RGB definition anyway, but that's beside the point."

So you have nothing against Adobe RGB, but you don't favor it, and it's invariably the culprit? Tell me, will our stomachs roast in hell at the hands of Adobe RGB, or just embedded profiles in RGB images?

Again: this thread has nothing to do with color management. It has
to do with common sense. Forget the way the world *ought* to work. Concentrate
on the way the world *does* work.

Of course we should. And then nothing ever changes. Nothing ever gets better. People can have the same old arguments, and you can pull your responses from archive. They can do the age old, it works perfectly well workflow of iterative fix and proof instead of improving workflow efficiency. You don't have to update your book as often, and you can keep on telling people the same things year after year, status quo, day in day out, on and on.

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Wed, 02 Jul 2003 19:52:28 +0100
   From: Martin Orpen
Subject: Re: Who's at fault?

Stephen Marsh wrote:

Why did the authors of Real World Colour Management choose not to tag their
CMYK press ready files for their book? If anyone was going to champion this
effort, one would think it would have been them. What better way to show
those who see no need for a description which is known to all parties to be
tagged to the file, than to do it on the live print production of the book
devoted to real world use of colour management.

Perhaps they used a printer that told them that images with profiles would be rejected because they have no interest in colour management?

As has already been pointed out - the owner of the final device in the chain has this option open to them without it affecting the image quality in any way whatsoever.

They don't need to make any conversion. They just need to ensure that correct percentage of cyan, magenta, yellow and black appears consistently on every page.

They don't need tags on the images to create proofs either. The proofer is adapted to match the response of the press - the image data does not need to be converted.

The main beneficiaries of colour management are the originators of the images and the intermediaries who have to work on them, repurpose them and then pass them on to who-knows-where. In other words *everybody apart from the owners of the printing press.

In Dan's *real world* the printers could sit about demanding CMYK data only and bitching about colour management zealots all day without any repercussions.

The problem is that they need to get work to pay for the sales reps Mercedes and are drawn like flies to a turd to the source of some major image problems: PHOTOGRAPHERS  and DESIGNERS.

You know the result when two tectonic plates meet? The creatives don't know the difference between UCR and HRT (yeah, that's hormone replacement therapy). The producers don't know ICC from MCC (Marylebone Cricket Club).

The really funny thing about the situation is that they don't need to do much more than RTFM that came with Photoshop, agree to take responsibility for profiling and maintaining their own equipment and then sit about discussing what makes a good image.

Instead they argue with each other and post their sorry tales to this and many other forums.

Still, I shouldn't grumble. It's good business being an intermediary between the warring factions :-)

--
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
________________________________________________________________________

Date: Wed, 2 Jul 2003 11:57:07 -0700
   From: Rick Gordon
Subject: Re: Re: Who's at fault?

On 7/2/03 at 11:59 AM -0400, Dan Margulis wrote in a message entitled "Re: [colortheory] Re: Who's at fault?":

However, I do issue the strong warning that Adobe RGB is unique among the four majors in that if it is misinterpreted the result is likely to be really, really bad. And I say, even in the latest edition of my book, "For work primarily aimed at non-Web RGB: If you are certain that your workflow won't let anyone convert (or fail to convert) it improperly later, use Adobe RGB. If not, use ColorMatch RGB."

But what then, if the client sends the printer an untagged RGB file that was created in a ColorMatch RGB working space? If it is opened under the assumption that it's untagged Adobe RGB, the result before CMYK conversion will be too dark and too strongly red.

With the converse instance -- an untagged RGB file in was created in an Adobe RGB working space that is opened under the assumption that it's untagged ColorMatch RGB -- you'll have the result you disparaged in the last Makeready article: an overly light file that's weak in saturation.

While neither is preferable, a case could be made that problem #1 (too dark and red-cast) is more damaging that problem #2 (too light and unsaturated).

This just seems to further demonstrate why not embedding a profile in an RGB file is a invitation for problems regardless what defensive strategy you employ.

Rick Gordon
___________________________________________________

RICK GORDON
EMERALD VALLEY GRAPHICS AND CONSULTING
___________________________________________________

WWW:   http://www.shelterpub.com
________________________________________________________________________

Date: Wed, 02 Jul 2003 20:50:20 +0100
   From: Martin Orpen
Subject: Re: Who's at fault?

On Wed, 2 Jul 2003 11:59:35 EDT, Dan Margulis
 
We know, because you've told us, that you definitely can deal with #3. Maybe
you can deal with all five of the others as well. But suppose we have such a
job, in one of the other five categories, that's due tomorrow morning.

Anyone sending such a job to you without first inquiring about your fitness
to handle it, needs a checkup from the neckup.

It may surprise you, but my colour managed workflow also doesn't:

7) cure cancer,
8) remove unsightly nasal hair,
9) add a *genuine* 3 inches to the length of my penis

But the important thing is that I don't have you bitching about my service because I didn't claim to be able solve any of the problems that you have cited - or imply that I could by accepting the job with the problems that you've warned me about.

But some seem to feel as a matter of principle that they must court disaster.
They declare, on principle, that it is absolutely the service provider's
obligation to know what to do with a tagged file, and that there is no need to take such
an easy insurance policy against his not knowing. Thus, the parade of
principled people with ruined jobs complaining to this and other lists.

ICC profiling is the only insurance policy that I have!

We scan and send out RGB data, we retouch both RGB and CMYK and proof both RGB and CMYK. You advocate a non-ICC insurance policy because you have one open to you. We don't and neither do the photographers who are creating most of the images that we work on.

I've said enough on this topic now, so before leaving the thread, I offer the
following to those who insist that strangers convert their files properly:

"He was right, dead right,
 As he sailed along,
And he's just as dead now,
 As if he'd been wrong."

Thanks.

In the meantime, if anybody needs a bunch of strangers in London to convert their files sympathetically you can get our number from the web site. Alternatively, you can go to a *real world printer* and get an in-RIP separation, crap print job, a bill and that warm feeling from perpetuating the myth that colour management is a waste of time.

--
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
________________________________________________________________________

Date: Wed, 2 Jul 2003 14:56:28 -0400
   From: Annette Murray
Subject: RE: Re: Who's at fault?

You don't have to update your book as often, and you can
keep on telling people the same things year after year, status quo, day
in day out, on and on.

The "book" is overwhelming to the majority of Photoshop users.

Most users need a simple explanation of how Image Mode Assign and Image Mode Convert.

Most users need a simple explanation of how helpful and dangerous Image Mode Assign and Image Mode Convert are.

Annette Murray
anro.com
________________________________________________________________________

Date: Wed, 02 Jul 2003 12:33:56 -0700
   From: Don Wiechec
Subject: Re: Re: Re: Who's at fault?

on 7/1/03 11:22 PM, Michael O'Connor wrote:

"You're certainly trashing this printer on a very skimpy one-sided explanation of
circumstances. While its clear that the printer at least acquiesced to accepting
the RGB, its not really clear that he "agreed" to, and its not clear what the
full communication between Alain and the printer entailed."
 
"but that doesn't open him to the degree of derision he's been receiving."

I haven't read derision anywhere in this thread.

I don't agree anybody is "trashing this printer."  Why are you are writing so pronouncedly one-sided for the printer? None of us needs a flame.  As far as your use of 'agreed to'  regarding the RGB, that is implicit in the printer's acceptance of the work. No customer needs to be an expert or anticipate the supplier's mind.

Don Wiechec
Traveling Camera,Inc.
________________________________________________________________________

Date: Wed, 2 Jul 2003 12:38:43 -0700
   From: Rick Gordon
Subject: Re: Re: Who's at fault?

On 7/2/03 at 12:59 PM -0400, John Romano wrote in a message entitled "Re: [colortheory] Re: Who's at fault?":

You also need an application that can use Device links and one to create a good one that doesnt reseparate through lab. But for loose color that the client wants to see proofs on that are supplied from god knows where I prefer to manage them in Photoshop. If I get a sep that was prepared for 20% dot gain it would be flat and washed out, if  i have a profile to convert to our specs it  gets the right numbers for our press conditions.

John,

What are the considerations between doing that and applying a transfer curve to match the dot gain?

Rick Gordon

------------------
___________________________________________________

RICK GORDON
EMERALD VALLEY GRAPHICS AND CONSULTING
___________________________________________________
 ________________________________________________________________________

Date: Wed, 02 Jul 2003 17:41:14 -0400
   From: Michael O'Connor
Subject: Re: Re: Re: Who's at fault?

Sorry, but I've seen lots of derision in the thread, and I wasn't really looking just to take the printer's side, and certainly didn't intend to flame anyone, I was just trying to imagine the actual situation and what the printer may have been faced with.

While there needs to be accountability on the supplier end, a customer has to accept some responsibility as well. If a customer brings a one-color hardcover book to a full color ad shop or a critical ad to a stationery shop, the customer has made a mistake. The printer has to let the customer know that its not their niche, but the customer has to make sure the printer has enough information about the job to realize that before the project snowballs. I don't see why a supplier should be required to anticipate the customer's mind either.

As to rgb, we still don't know the full circumstances, if the printer had agreed beforehand to accept rgb that's one thing, if the printer was under a misconception about the job and was later faced with rgb that he thought he was told not to worry about, that's another. From Alain's latest post it would seem the printer had rarely ever opened photoshop, though there are other indications that would seem to say that couldn't be true, Alain and/or the buyer should have have had enough communication with the printer to have determined this beforehand for themselves, they're serving as professionals too.

Michael O'Connor
________________________________________________________________________

Date: Wed, 02 Jul 2003 14:43:01 -0700
   From: Kevin Connery
Subject: Re: Re: Who's at fault?

While I understand why the participants in the "profile issue" get emotional, I've also noticed one very significant "market segment" that's not included in any of the discussions: photographers.

Not commercial photography, where the images will be retouched and printed in books or magazines by the thousands, but portrait, wedding, and school photographers. In those cases, "running a proof" to identify the colors desired is just another way to double the costs of every single image.

Some form of describing how a digital image's colors should look is needed.

(Well, or have the lab correct the way they had been doing for the last 30 years--by eye--which, unfortunately for most photographers, very few labs will do.)

Are profiles "evil" in that situation? Can anyone suggest an alternative?

I help a lot of digital  photographers get their colors to match, and without having to send a physical print along with every file they send to their lab.

Maybe CMYK IS different--I don't do that much printing to CMYK as such--but the heat generated in this discussion is making it hard for the sparce facts to be convincing either way.

--kdc
 ________________________________________________________________________

Date: Wed, 2 Jul 2003 15:34:24 -0500
   From: Bob Smith
Subject: Re: Re: Who's at fault?

On Wednesday, July 2, 2003, at 11:42  AM, Chris Murphy wrote:

Only by a shop that doesn't know what it's doing. I'd like to hear
anyone demonstrate how unintended conversions occur otherwise.

agreed, its a major mistake, but in my experience its a mistake made more likely with embedded profiles in CMYK files.  Again, I'm ONLY talking about files already prepped for output where no further conversion is needed.

How is
it more likely, let alone far more likely, that a shop will convert
ONLY BECAUSE there is an embedded profile?

Because a profile mismatch is more likely to trigger a warning that gives a user an opportunity to make a mistake than a missing profile.  An image with a missing profile will almost always be treated first as though its in the shop's standard CMYK space. The reality is that there is no profile mismatch in the situation I was describing.  My file will print fine; and most likely soft proof just fine, it just wasn't made with a profile with identical characteristics as they one they use as their working CMYK space...most likely different black generation.  Hell, we could even be using identical profiles that just have different names.  That could trigger an unwanted conversion that will destroy any custom channel editing I may have done as the file is re-separated. Maybe I airbrushed in black only drop shadows and the re-separation turns them into four color shadows. I'm not saying this is a likely scenario at all, but it has happened to me more than once.  I've never had any similar problem with an untagged, print ready file.  I realize that yes, unintended conversions can happen with untagged files as well.  I just don't see it as likely and my admitedly very unscientific sampling of my experience backs this up.  If (very big if) the file I give them was properly prepped for their conditions, it will look fine on their monitor and print fine on a composite proof even though they may be assuming some other profile as the source.  Their assumed profile, their CMYK working space, describes the same conditions that my file is targeting.  We just may have used different black recipes or for that matter may have just named our profiles differently.  If my file looks fine on their monitor there's not going to be any urge to waste time as they look for a way to assign a profile to this untagged file and convert it.  They'll just print it.

How, specifically, would this occur *unintentionally*, and do you have
any data to show how often this happens?

unintentional may have been a poor choice of words.  Unintended on my part... I didn't want it.  The print shop may well have actively (correctly or not) taken steps to cause the conversion.  Or it may be one of the automation issues that you referenced.

I have heard only 3rd hand
stories as though they were legend

It only has to happen a couple of times to worry me and it has.  In the worst case (meaning affecting many images on a large job and the blame fingers were pointed squarely at me) the problem was discovered in one of the early proofing rounds and easily corrected. This was a case where an unwanted conversion was causing a color problem.  What was causing the conversion, I never was told exactly, but as soon as I explained to the foreman how I had prepped the files and had included an embedded profile he knew exactly what his production artist had done and corrected the job quickly. The next batch of proofs were right on.

Bob Smith
________________________________________________________________________

Date: Wed, 2 Jul 2003 19:23:00 -0500
   From: Ron Bean
Subject: Re: Re: Who's at fault?

David Chusid writes:

Every one of Dan's comments about printers respecting embedded
profiles has included the word "strangers".  If the printer and
the client talk, and develop some mutual respect regarding their
respective use of color management techniques, and decide to give
embedded profiles a try, and run some tests, and check the
results, and decided that all is OK, then they aren't strangers
any more.

IMHO this hits the nail on the head. What's described above is no different from two modems negotiating protocols before they exchange data. This is the 21st century, we have the internet, cellphones, etc. We should be communicating more, not less.

I wouldn't mind these color management discussions, if they didn't cause the S/N ratio to drop like a rock. If a [printer|customer] doesn't understand color management, you can either:

1: Call him names, or

2: Work together to find a solution, or

3: Find another [printer|customer]

I don't need a mailing list to tell me how to do #1.

[Hint: If you need to vent, write a response and *don't send it*. After you've calmed down, if you still have something to add, write a different one and send that instead.]  ________________________________________________________________________

Date: Wed, 2 Jul 2003 21:39:47 -0400
   From: John Romano
Subject: Re: Re: Who's at fault?
 
Rick

Getting the files close up front will let you still use a transfer/ plate curve later on. The only time I like to go for plates curves is if the press is having problems with paper or what ever, sort of a last defense.We run linear plates 95% of the time. Dont get me wrong we still use plate curves on rare occaisions, Our profiles are an average of multiple presses and like I said it will get you very close.
But in the rare occurance you still have plate curves in the end.

One other thing to mention, on supplied files. Some can be UCR seps and some can be GCR or nothing at all. How do you think this will print ? different set ups across the sheet. Sounds like a fun day on press no ? Converting all supplied files is they way to go . If they cant give you a profile, send them a target and make one yourself. Assign and convert and your there. Im not a consultant, not that its a bad thing here but I make profiles and use them in day to day production. If they didnt work I wouldnt feel so strongly for them. Im done nothing left to add here, Eveyone have a nice 4th and im sure this thread will die soon.

Regards

John
________________________________________________________________________

Date: Thu, 03 Jul 2003 08:45:53 -0000
   From: Stephen Marsh
Subject: Re: Who's at fault?

Chris Murphy wrote:

This is why repurposing CMYK content needs a DeviceLink profile (a
class of ICC profile) instead of two Output Device profiles. Black only
can be retained, or even scaled (compensating for dot gain between the
two black channels), preserve channel purity, etc.

Yes, DLP's sound good and we have briefly touched on them before.

It's a pity that APS can't use them, though.

If the 'average' printer does not know what a ICC profile is - what chance have we got that they will know about a DLP? <g>
 
Not quite. BOTH the file numbers and color appearance matter.

It is right, if all the printer does is open up the layout and print the linked images with no colour conversion - the CMYK values are simply passed on, as has traditionally been the case.

See above - I do not want you [anyone, nothing personal] messing  with
my CMYK values!!!

If the job wasn't separated for the specific print condition, yes you
do.

I was talking of sending a known sep based on the SP's guidelines.

Changing the supplied numerical colour data would be as bad as changing a font.

If you're using their profile, and it gets embedded, there will be no repurposing.

It is rare that a profile is given, more often aimpoints or an Adobe profile preference is stated. It all depends on the shop.

What do you care whether it's embedded
or not, so long as you know they're going to treat it correctly?

I tag when the job is to supply a scan.

I often do not tag for a final layout image which is in output.

By treating the CMYK tag correctly - no conversion would be made of the data for film/plate setting. It is only for preview if needed. I do not want a conversion, otherwise I would supply tagged RGB (if it was accepted). The other use would be for digiproofing, where a conversion from the CMYK source to the inkjet space is needed and a good thing - but that is not the same as film/plate.

Since every image does not need a tag (the same one) - it would be more beneficial to simply provide the single profile along with the files (license restrictions may apply).
 
Stephen Marsh
________________________________________________________________________

Date: Thu, 3 Jul 2003 11:20:57 -0600
   From: Chris Murphy
Subject: Re: Re: Who's at fault?
 
On Thursday, July 3, 2003, at 02:45  AM, Stephen Marsh wrote:

It is right, if all the printer does is open up the layout and print
the linked images with no colour conversion - the CMYK values are
simply passed on, as has traditionally been the case.

With a DeviceLink profile in place there is still a conversion, but it's also possible to preserve certain things UNLIKE regular Output Device profile to Output Device profile conversions; such as black only drop shadows, text, preservation of channel purity, etc.

Changing the supplied numerical colour data would be as bad as
changing a font.

It happens all the time, I think people need to get over this obsession with device values. For all you know, the printer has a bunch of presses and only gives you a general profile because they have no idea what specific press it's going to be on until perhaps an hour or even minutes before it goes on press. And what happens is your file ends up going through either a DeviceLink or set of transfer curves to change your numerical values anyway.

It is rare that a profile is given, more often aimpoints or an Adobe
profile preference is stated. It all depends on the shop.

Yes and just imagine if people treated fonts the way some people on this propose we treat color management. If the industry were so cavalier about fonts we'd have regular font disasters. Ooops! We do have regular color disasters. I wonder why? Maybe it's because people are unwilling to define responsibilities and stick to them, like we do with font handling.

Since every image does not need a tag (the same one) - it would be
more beneficial to simply provide the single profile along with the
files (license restrictions may apply).

That's the point of PDF/X-1a. OutputIntent is required. That is the source and destination for everything in the file. It's either an externally referenced colorimetric definition or it can be an explicitly embedded ICC profile. But it is required. You can't NOT have it, and have a PDF/X-1a file. And PDF/X-1a is the press ready delivery format. All colors are CMYK (or black), images must be embedded, fonts must be embedded. And the point of the OutputIntent being both source and destination is that it defines the data as being for a specific destination which means a null transform. Only when the destination is different (which should be only for proofing) is there a transform.

The reason why PDF/X-1a and PDF/X-3 work is because there are clearly defined, and required responsibilities for the two parties (print buyer and print supplier).

Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________

Date: Fri, 04 Jul 2003 01:04:01 -0700
    From: Dennis Dunbar
 Subject: Re: Re: Who's at fault?
 
 Chris Murphy wrote:
 
 Under "What is DISC?" they say: "We are recommending the Adobe RGB 1998
 color space because it is large enough to encompass most digital
 capture, display, and output devices." There it says
 recommended, elsewhere it sounds like a requirement. I'll see if one of
 their contacts can clear up the apparent contradiction<<<
 
There was a great deal of discussion about whether it's better to "require" or to "recommend". As I understand it the practice comes down to practicality, if a publisher has a strong relationship with a photographer then there is little need for these specs. But for new relationships, and to "guide" existing ones towards some sense of simplicity workflow wise they thought it best to create the specs. These are not meant as a threshold, as sole cause for the rejection of an image. They are meant as a basis for communication, for the exchange of files that cannot happen in a vacuum.
 
It's amazing to see how this thread dances with an obvious point but just doesn't seem to ever get close enough to it. In this line of work communication is essential. Without it disaster is inevitable. Both client and vendor need to do everything necessary to be certain that the expectations are set and met properly.
 
Whether it's best to embed profiles or not is just a part of the communication. If a vendor assumes all images are in a given space and has a workflow where inappropriate conversions may happen they need to be very upfront with the clients on this. If they honor profiles and expect them to be embedded then they need to be upfront about this too. One way or the other it comes down to giving the other side the information necessary to do the job right. If we substituted another subject for profiles the need for communication would be obvious. Suppose I needed my job printed on cardboard, but the printer assumed everything was going to be printed on card stock. If I didn't tell him what to print on and he didn't ask who'd be responsible? In rush circumstances these things happen more often than we care to admit.
 
In the end if we want to stay in business we will learn to communicate.
 
Dennis Dunbar
APA Digital Committee Chair
________________________________________________________________________

Date: Fri, 04 Jul 2003 01:26:28 -0700
    From: Dennis Dunbar
 Subject: Re: Re: Who's at fault?
 
 Martin Orpen wrote:
 
 The problem is that they need to get work to pay for the sales reps Mercedes
 and are drawn like flies to a turd to the source of some major image
 problems: PHOTOGRAPHERS  and DESIGNERS
 
Sure, life is always easier without the darn clients and content providers. Of course without them there'd be no need for people who express this insulting attitude either.
 
 Still, I shouldn't grumble. It's good business being an intermediary between
 the warring factions  [:-)]
 
This post sounds a bit like the arms merchant who sells guns to both sides, laughing at what he thinks is their folly. Laughing that is until he notices his whole world is burning as a result of the war he's enabling.
 
 Isn't business tough enough? Shouldn't we seek out ways of working WITH each other instead of finding new ways of squawking about how bad the other guys are? Where is the future in this attitude?
 
Dennis Dunbar
________________________________________________________________________

From: Terry Britton
Subject: [colortheory] Re: Who's at fault?

Dennis Dunbar wrote:

Whether it's best to embed profiles or not is just a part of the communication. If a vendor assumes all images are in a given space and has a workflow where inappropriate conversions may happen they need to be very upfront with the clients on this. If they honor profiles and expect them to be embedded then they need to be upfront about this too. One way or the other it comes down to giving the other side the information necessary to do the job right.

This seems like you are spelling out a list of items that should appear on any preflight checklist or printed communique or job ticket. Of course the color management details belong there - and there should be entries for it whether you use it or not. Why depend on remembering to do this via word of mouth? The checklist approach means you will be going over the list verbally, with both parties holding the same information in their hands, when doing the usual phone conversation followup.

To some people, communication aside from their company's job order or checklist simply means, "So, how are the wife and kids?" or "Did you see those Allstars?!?" Lists help a lot to make certain the 'other' important stuff is covered before the friendly schmoozing stuff takes over the left brain's activities! (Right brain *rules* around here! <bg> )

Terry Britton
________________________________________________________________________

From: Dan Margulis
Date: Sat, 5 Jul 2003 14:03:05 EDT
Subject: Re: [colortheory] Re: Who's at fault?

Rick Gordon writes,

But what then, if the client sends the printer an untagged RGB file that
was created in a ColorMatch RGB working space? If it is opened under the
assumption that it's untagged Adobe RGB, the result before CMYK conversion will be
too dark and too strongly red.

Who said anything about giving untagged RGB? The topic of this thread is, if you *do* give tagged RGB to a stranger, is there some universal human right to the effect that the stranger is going to interpret it correctly?

With the converse instance -- an untagged RGB file in was created in an
Adobe RGB working space that is opened under the assumption that it's untagged
ColorMatch RGB -- you'll have the result you disparaged in the last Makeready
article: an overly light file that's weak in saturation.

I had completely forgotten, until you brought it up, that my last two columns are about situations just like this: where a person assumes too much of his partner and fails to take elementary precautions against the assumption being wrong. I illustrated this with half a dozen catastrophes, one of them being exactly our case: the expert embeds a profile from a wide-gamut RGB before handing it on to an unknown printer; the printer ignores it; the job looks terrible; the expert calls the printer names and bemoans his own bad luck in having to deal with such incompetence.

For those who don't feel like having their jobs destroyed in this fashion, there are two alternatives. First, if you feel confident of your color skills, you can give CMYK. If that's undesirable, convert your RGB file to LAB before handing it off to the stranger and tell *him* to reconvert it to RGB. If he can't figure out what to do at that point you certainly don't want him hacking away at your RGB file, tagged or untagged.

It's all a matter of driving defensively, about having the proper suspicions about other people involved in the process. Those who do so may not think very much of their printer, but their jobs usually get printed correctly. To quote that column:

"The next time you hear someone with impeccable credentials wailing because a printer botched his job, or a client gave bad instructions, or color management got screwed up, ask yourself: does it seem like this person is consistently more unlucky than he should be? Because if he is, perhaps it's not a matter of luck at all."

Dan Margulis
________________________________________________________________________

From: Dan Margulis
Date: Sat, 5 Jul 2003 14:22:45 EDT
Subject: Re: [colortheory] Who's at fault?

Dennis writes,

It's amazing to see how this thread dances with an obvious point but just
doesn't seem to ever get close enough to it. In this line of work communication
is essential. Without it disaster is inevitable. Both client and vendor need
to do everything necessary to be certain that the expectations are set and met
properly.

Yes, it's amazing that so much heat was generated, considering that the basic premises are uncontested by anybody, to wit:

1) Client and printer definitely have to be a lot more motivated to communicate with one another than this client and this printer were.

2) It's a good idea for service providers to understand what embedded profiles are and how they may be of use.

3) Nevertheless, a very large number certainly, and IMHO a healthy majority, neither know nor care about the topic.

4) If a file prepared for Adobe RGB is misinterpreted as being something else, the likely result is a ruined job.

Given those premises, one's course of action in the real world seems obvious. Getting off on tangents as to how certain operations *do* pay attention to profiles, or trying to make this into a referendum on color management generally rather than on this tiny part of it, serve no purpose. In fact, they are counterproductive in that they imply to people like Alain that there's no need to prepare for the likelihood that the printer won't honor the profile.

In my current column, which speaks to situations like the one Alain engineered himself into, the headline reads as follows: "Publishing today is very much a partnership game--but there's a limit to how far you should trust your partner. Cater to the possibility that the client, or the printer, may not be an expert, and you'll be a winner in the long run."

Dan Margulis
________________________________________________________________________

Date: Mon, 07 Jul 2003 15:15:47 -0700
   From: "J J Smith"
Subject: Re: Re: Re: Who's at fault?

 Martin Orpen wrote:

The problem is that they need to get work to pay for the sales reps Mercedes
and are drawn like flies to a turd to the source of some major image
problems: PHOTOGRAPHERS  and DESIGNERS.

Having spent the past 25+ years as both a photographer and a graphic designer, I'm certain I've run into a couple of attitudes like those above - but, thankfully, not many.

Keeping the lines of communication open between design, prepress and print is a function I've personally put at the top of the list since I first began in this business. Bringing in both downstream functions at the outset of a project to discuss how everybody can make each others life easier, while keeping the client (each of our clients) happy in the process has totally eliminated the part of the process where everyone points the finger of blame at one another (generally found at the end of a project).

And guess what - everybody always gets paid to boot.

It's called the team approach. It requires the mutual respect of all members to work. There's no place on our team for the kind of attitude expressed above. The people we work with seem to appreciate the fact that we ask for their assistance in getting things done that we may be unfamiliar with - and they'll bend their own rules to help when needed. As a result, most of the people we wind up working with actually look forward to getting our files and clients wind up getting everybody's best efforts.

And, besides, it's a lot more fun that way.
--
Jeff Smith

Smith/Walker Design and Photography

P. O. Box 58630
Seattle, WA  98138
ph: 206-575-3233
fx: 206-575-3960

________________________________________________________________________

Date: Tue, 08 Jul 2003 14:38:09 +0100
   From: Martin Orpen
Subject: Re: Who's at fault?

on Mon, 07 Jul 2003 15:15:47 -0700, "J J Smith" wrote:

Having spent the past 25+ years as both a photographer and a graphic
designer, I'm certain I've run into a couple of attitudes like those
above - but, thankfully, not many.

Jeff

Saying one thing while meaning another is a something that people have used to good effect in English for about 5 Centuries - I guess that's why the plays of Shakespeare were so different from much of the dull stuff that preceded them.

However, you are the second person to highlight one particular sentence from my post and appear to attribute it directly to me as if it were my own view. Fortunately your post did not lead on to a bunch of idiotic insults like the one that Dennis Dunbar posted on Friday.

I chose not to respond to that post as the thread looked like it was coming to a welcome end and verbally beating up somebody who has both limited comprehension skills and less knowledge & ability in this area than one of my trainees would have been embarrassing for the *APA Digital Committee Chair*. (Little wonder he didn't bother to sign that particular message with his official title!)

But, for the record, I'd like to point out that this sentence does NOT reflect *my* attitude. It was an illustration of how stuff can go bad when people take on work for which they do not have the requisite skills to complete.

***BEGIN SARCASM ALERT***

I was forgetting that there are some English speakers whose forebears spent the whole of The Enlightenment with their noses firmly wedged in the pages of one particular book. For whom all sentences must be plainly stated and truthful and for whom the such modern methods of expression are incomprehensible.

***END SARCASM ALERT***

Perhaps explicit warnings like those above will help?

--
Martin Orpen
Idea Digital Imaging Ltd -- The Image Specialists
http://www.idea-digital.com
 ________________________________________________________________________
 
Date: Tue, 08 Jul 2003 09:21:45 -0700
   From: "J J Smith"
Subject: Re: Re: Who's at fault?

Martin Orpen wrote:

...But, for the record, I'd like to point out that this sentence does NOT
reflect *my* attitude. It was an illustration of how stuff can go bad when
people take on work for which they do not have the requisite skills to
complete.

I took the opportunity to respond to the remark because it's far from the first time I've heard it. There are many people in the graphic arts that share the view that their life would be infinitely better if there were no designers or photographers to have to deal with.

In all honesty, there are a few designers I've met over the years that I'd be reluctant to work with, too. Generally these are the people who seldom wind up getting the results they want because they lack the skills to communicate effectively with the people that must produce their projects, and the talent to provide them with the materials they need to realize their expectations.

But in the current world where we all share in the benefits and problems of new technologies that are creating new ways of getting our work done and blurring the traditional lines between functions in the process, doesn't it make more sense to work in an atmosphere of mutual respect and trust. It does to me.

As a photographer, the shift to digital imaging took away my security blanket of an approved transparency to hand off for prepress. CTP technology has made the search for a good yet inexpensive method of proofing our files prior to release for prepress critical. Without a good 'validation' proof (not a contract proof necessarily) we're flyin' blind - and creating potential problems for the prepress people and printer that means remakes and additional charges to the client.

The best way I've found to cope with the problems presented by the new technologies is to share information - openly. Communication allows everyone to progress and grow. Photoshop is a great example. There must be three ways to do everything in Photoshop. Friends in prepress are constantly surprised at how creatives use the program, and creatives could often use a few of the methods prepress people use everyday. For me, it makes every day I come to work an opportunity to learn how to be a better designer, a better photographer, and a better resource for my clients. The people that work for me have the same attitude - how can we get this done; what resources can we use to solve this problem; who in the production chain is likely to have seen this before?

After all, in the eyes of the client, we're all part of the same problem - or part of the same solution. It's our choice.

--
Jeff Smith

Smith/Walker Design and Photography

P. O. Box 58630
Seattle, WA  98138
ph: 206-575-3233
fx: 206-575-3960
________________________________________________________________________

Date: Thu, 10 Jul 2003 10:27:07 -0400
   From: Dolores Kaufman
Subject: Re: Re: Who's at fault?

Jeff Smith wrote:

After all, in the eyes of the client, we're all part of the same problem
- or part of the same solution. It's our choice.

Well said, Jeff. I think that's the reason most of us are here. Dan's the Man. Makeready put me on the right track and there are others on this list who make valuable contributions from their own extensive experience, as well, adding to the nourishing stew from which we can all partake. I just wish some would refrain from tossing their large egos into the pot, as well. Makes the stew taste bad.

Dolores
DGK Creative Imaging

And with that, there was a two-week ceasefire. But hostilities recommenced in earnest with an even longer thread on handoffs to strangers. See "Deliver in LAB" in this section. Meanwhile, the question of whether CMYK profiles should be honored  in the absence of an instruction was answered by Photoshop CS, which by default ignores CMYK profiles.--DM

Adobe Photoshop training classes are taught in the US by Sterling Ledet & Associates, Inc.