Using LAB values to create a color chart - what can go wrong?

[Marie Meyer]

I've been asked to bid on the creation a marketing piece for a manufacturer of fine art inks. They want it to include a color chart, as accurate as possible with 4 color offset litho. Their actual specifications call for it to be "matched to Pantone colors". But I know they want it printed in 4 color and will not pay for spot colors. So I'll assume they mean Bridge colors. However, I'm inclined to skip that and go the direct route: My plan is:
1) make ink swatches
2) measure with spectrophotometer 3x; average results
3) check LAB data against GRACoL profile in PatchTool or ColorThinkPro to make sure everything is in gamut
4) turn on color management!
5) design piece, using the LAB data to color the swatches.
6) make jolly sure it goes to a printer using the GRACoL profile!

What haven't I thought of - what can go wrong?

 

[gordo]

IMHO - this is a waste of time - too many uncontrolled and uncontrollable variables.
You can create a color atlas like one of these: http://the-print-guide.blogspot.ca/2009/07/color-atlas-helping-designers-to.html however it would only be valid as printed by individual printshops and only as a rough guide if used for several print shops. The fact that a printer is using a GRACoL profile does not mean that they print to GRACoL. In many cases they might have been set up to do so at one time but have drifted away.

IMHO they should drop the idea of such a color chart in their marketing material.

 

[Marie Meyer]

The color chart is a must for them. Is there a way to verify that a printing company is "printing to GRACoL" correctly before sending them the job?

 

[gordo]

The color chart is a must for them. Is there a way to verify that a printing company is "printing to GRACoL" correctly before sending them the job?

If you're color critical you would have to have the printshop do a press test (on a substrate that meets the specification) using at least some version of the P2P25 target, measure the result and run that data through the IDEAlink Curve tool (or equivalent). At the least you would have them print and measure the IDEAlliance ISO 12647-7 3-Row Digital Control Wedge 2013 which may be sufficient (although it's primarily for proofers).

That should show you if the printshop is within GRACoL 7 parameters.

You've stated that "Their actual specifications call for it to be "matched to Pantone colors". But I know they want it printed in 4 color and will not pay for spot colors." That, to me, indicates an issue with setting expectations and a potential liability for you. IMHO, you should confirm with them that they understand that printing 4/C process will only simulate maybe 30%-40% of the Pantone color library and that percentage may not be included in the spot colors that are selected to be simulated in this project.

Since this is for inks for fine art application it might be better to imply gamut (which is what I assume they're trying to show with the Pantone simulations).** Something like this gamut visualization:

That shows the GRACoL gamut inside a larger translucent gamut (e.g. your fine art inks). The dots represent individual Pantone spot colors.

**This is a hot button for me >

(Warning - rant follows please don't take it personally) When I used to do graphic design, clients would often tell me what they want in the design (e.g. "The color chart is a must for them") and, being a bit cocky (or professional), I would tell them that their job was to articulate what their marketing/communication problem was - not to tell me what to do. Just like when you go to a doctor, a car mechanic, or some other PROFESSIONAL. You don't tell a car mechanic what to do or how to do it - that's their job. You tell them what your problem is (the car makes a chirping sound when I turn left but not when I turn right). It's their/your job to probe with questions to get a more in depth understanding of what your/the client's problem is so that they/you have a better sense of how to solve it.
This video shows humorously when people are not performing their proper roles: https://www.youtube.com/watch?v=BKorP55Aqvg

 

[rich apollo]

If you find that the colors they want to show are out of gamut, does the color chart still have value, or relevance? And, since this is about "fine art inks" I think it's safe to assume that they will be very demanding. I foresee great wailing and gnashing of teeth.

 

[Stephen Marsh]

Start with a bright orange, similar to Pantone 021C, and a “dark but bright” blue such as Pantone Reflex Blue or Pantone 072C etc. What are the Lab values of the orange and blue fine art inks?

From this worse case scenario, you should be able to illustrate the issue with notoriously out of gamut colours for common press CMYK conditions.


Stephen Marsh

 

[David Milisock]

>what can go wrong You've ask the wrong question. The correct question is what wouldn't go wrong?

 

[Marie Meyer]

Rich, Stephen, I've made swatches and it appears that all but two are in gamut (GRACoL). They've agreed that they are OK in principle with flagging those two with a note of some kind. Instead of allowing the software to clip them to make the closest in-gamut simulation, I'm going to make my own determination of what in-gamut color is the closest match for hue and value, and then attach a note saying that IRL the chroma is X% higher.

David, any specifics you have in mind?

 

[Stephen Marsh]

Marie,

Consider that there are many combinations of CMYK values that in theory (say proofing) equal the same Lab value – however on press there will be natural variation.

Speaking of variation, is your client willing to accept that there could be a +/- difference from the “ideal” colour (proof) to the press run? Say +/- 3-5 de76?


Stephen Marsh

 

[David Milisock]

All of them, paper changes during the run. An example would be the time we spent making swatches and profiles for an inkjet, all was well until three days later the new lot of media was different, significantly different. In my view balancing color management with production realities are not what many color management professionals seem to say. With that said good solid practices and running to something like GRACOL or swop or some industrial standard is very profitable. Making swatches maybe not so much.

 

[gordo]

And keep in mind the potential impact of OBAs on the color - especially any pastels. Try and use a paper that has as low an OBA content as possible.

 

[cosmo]

Good luck, and you are the printers printing the job? Theres a reason there are spot colours, theres a book to match to.

Lab is just another way of defining color, at the end of the day it still has to be mapped back to CMYK, so it heavily relies on the machines you are running.

 

[Werby]

A job like this where color patches have to be extremely accurate require partnering with a printer and doing some testing back and forth, ideally including a press test on the actual stock. As has been mentioned, even a good printer could vary 3 dE from your approved proof and still be within spec for gracol. There is a tolerance built into any spec and I've seen plenty of proofs and/or printed pieces that fall within gracol specifications but look significantly different from one another. If these swatches are really supposed to match the actual specialty inks, there will have to be some very tight coordination with the printer.

 

[David Milisock]

Good luck, and you are the printers printing the job? Theres a reason there are spot colours, theres a book to match to.

Lab is just another way of defining color, at the end of the day it still has to be mapped back to CMYK, so it heavily relies on the machines you are running.

Sometimes printer is spelled REALIST!

 

[meddington]

Whether or not this is successful is up to setting realistic expectations. The onus is on the printer/separator to convey the inherent limitations of the color gamut and the certainty of process variation in production. The level of variation may be acceptable. Other industries rely on different methodologies to ensure accuracy…like how paint chip cards are made:
https://www.youtube.com/watch?v=roMJblx9FBM