Trouble color matching blues & purples - offset CMYK, G7 methodology, GRACoL 2013

[prepress labels]

Hello!

We're having trouble matching deep blues & purples on our Heidelberg Speedmaster CD-102. We've been doing our own press curves for about a year now (using Curve4 and an X-Rite ES-6000), and most of the time, it works out pretty well (until there's a change to fountain solution, ink, material, etc).

Our proofer is a newer Epson SC P7000, and we're running the GRACoL 2013 profile (we'd been running a custom profile for years--same issue).

We've historically had issues matching purple & blue in CMYK (not matching spot colors--this is CMYK both in proofing and at press), and we just had to play with the ink densities on press to deal with it. However, we're trying to get away from doing that, since we gang different items on practically every job.

We recently had to go in and tinker with some print-ready art files to achieve the dark blue & purple colors on our proofs (which also closely match the color on the Epson proofs supplied by our customer). We've also manually edited our press curve to add in (going on a couple weeks now--seems OK so far).

Being in prepress, I'm at a bit of a loss as to what else to do--this is an issue that's been with us since forever. We used to run a custom profile, and we'd have an outside company generate curves for us, and had the same issue. The curve would work until we ran into a press sheet where it didn't, and then press would report that "the curve went bad". We'd tweak it again periodically, then we decided to buy our own curve-generating software and do it all in-house. That seemed to calm things down for awhile, but then we ran into this blue/purple issue all over again. We did switch ink vendors recently, but we did a new curve after they were up & running.

Any suggestions? I know there's a number of potential contributing factors at press, but this seems to be pervasive and historic. Can anyone enlighten me on the effectiveness of G7, or GRACoL 2013?

Any input is appreciated--thanks!

 

[Stephen Marsh]

First question, is the press room “printing to the numbers” on a given press and stock – with the numbers being a repeatable SID value for solid C, M, Y, K… Which should translate to a repeatable L*a*b* colour value for solid C, M, Y, K and C+M, M+Y, Y+C… Forgetting curves for now, just solids.


Stephen Marsh

 

[DeltaE]

Hi there,

If I right understood, it is hard for you to match the offset printout with the proofing (from Epson which is simulated in GraCol 2013).

Here is my idea for you:

Your custom profile (your machine, curves, SID, Lab of CMYKRGB,...) now is not matched with Gracol 2013. It causes your difficulty when you want to adjust the offset printout as the proofing. You have two options:
- 1st: choose suitable ink, blanket, paper, plate,... and adjust the curves to print a finger print with IT8 or ECI2002, and repeat until your offset printout matches Gracol2013.

- 2nd: or you calibrate your Epson proofing (with Fiery EFI, GMG, Heidelberg ColorProof,...) to simulate what you print in normal condition of offset machine. Then you can predict the print result in offset before. Using your proofing and adjust the files until your proofing matches with clients' sample. Then you do not have to play with ink density anymore.

In my opinion, I will choose 2nd option because you never use same ink, consumables, papers.... forever!

Regards,
DeltaE

 

[D Ink Man]

Try switching rotation to M-C-Y-K. Worth a try; have seen it overcome this issue.

D Ink Man

 

[prepress labels]

First question, is the press room “printing to the numbers” on a given press and stock – with the numbers being a repeatable SID value for solid C, M, Y, K… Which should translate to a repeatable L*a*b* colour value for solid C, M, Y, K and C+M, M+Y, Y+C… Forgetting curves for now, just solids.


Stephen Marsh

Yes--we typically run one stock, and we're running to standard ink densities. In the past we'd play around with the ink keys a lot more to hit color, but we're trying to get away from that, since it's not always repeatable from job to job.

 

[prepress labels]

Hi there,

If I right understood, it is hard for you to match the offset printout with the proofing (from Epson which is simulated in GraCol 2013).

Here is my idea for you:

Your custom profile (your machine, curves, SID, Lab of CMYKRGB,...) now is not matched with Gracol 2013. It causes your difficulty when you want to adjust the offset printout as the proofing. You have two options:
- 1st: choose suitable ink, blanket, paper, plate,... and adjust the curves to print a finger print with IT8 or ECI2002, and repeat until your offset printout matches Gracol2013.

- 2nd: or you calibrate your Epson proofing (with Fiery EFI, GMG, Heidelberg ColorProof,...) to simulate what you print in normal condition of offset machine. Then you can predict the print result in offset before. Using your proofing and adjust the files until your proofing matches with clients' sample. Then you do not have to play with ink density anymore.

In my opinion, I will choose 2nd option because you never use same ink, consumables, papers.... forever!

Regards,
DeltaE

Correct--we're having trouble matching our printing (specifically blues & purples) to the GRACoL 2013 proof.

1st: we've actually already done this--printed & scanned the it8 chart. Still getting the issue on blues & purples.

2nd: Our Epson proofer is calibrated, we run nozzle checks with every proof, and we have the GRACoL 2013 profile plugged into our EFI XF proofing software. Our proofs already matching our clients' proofs--it's just the occasional blue & purple that don't match on press.

 

[prepress labels]

Try switching rotation to M-C-Y-K. Worth a try; have seen it overcome this issue.

D Ink Man

Thanks for the suggestion--does the ink rotation cause issues in other color combinations, in your experience?
BTW--we're running K-C-M-Y if that helps.

 

[curiosity]

What DeltaE said (2nd option)
Especially if what I understand to be true is that the file's cmyk values do NOT change from proof to press.
Can you confirm that?

 

[gordo]

Before you start randomly pulling levers and twiddling knobs -

what are your SIDs measured dry?
what are your solid ink Lab values measured dry
what are your RGB % trap values?
what are your RGB trap Lab values?
have you plotted/charted your dot gains? If so can you share?
are you using the same instrument to measure in prepress and pressroom?

BTW, the Gracol ink sequence is KCMY. Switching C and M to KMCY will shift the hue of blues and purples (as well as greens and oranges) but moving the K from first to last won’t since it is achromatic.
Color on press, especially older presses, can be affected by inline ink volume ussage in that area. That does not happen with proofers.

 

[prepress labels]

What DeltaE said (2nd option)
Especially if what I understand to be true is that the file's cmyk values do NOT change from proof to press.
Can you confirm that?

Yes--we're sending the same file to the proofer that we're sending to press. We apply the GRACoL 2013 profile at proofing, and we apply the press curve at plating.

 

[curiosity]

Do you have a profile of your press?
If so, open a blue/purple file color managed to GRACoL in Pshop.
Then assign your press profile. The color should shift.
Why?
Because you're color-managing files when proofing, but not when printing.
Curves can get you to neutral GB, but that software will not manage color per se.
For example, if you send 90c, 70m to proofer, it gets c-m'd to something different than 90/70. Let's say maybe to 95/70.
To get same results from press, your press profile might require a conversion to significantly different cyan and magenta values.

 

[D Ink Man]

Please twiddle the knobs as I suggest. You need to try.

D

 

[gordo]

Please twiddle the knobs as I suggest. You need to try.

D

I don’t agree. The first thing to do is establish your manufacturing’s current condition relative to your defined targets. That will help narrow down what you should try.
Otherwise you’re like a drowning person flapping their arms uselessly thinking that at least they’re trying something.

 

[DeltaE]

Please tell me the CMYK composition of your "Dark Blue and Purple"!

 

[Alois Senefelder]

prepress labels,



Instead of going round in circles and Up and Down - "Curves"! ............. post a good image of the Print Control Strip - inline with H/tone and Solid Image also.



Regards, Alois

 

[roger]

Prepress lables,

In what way do the blues/purples on the proofs differ from the offset printing? (lacking cyan, lacking magenta, weaker/stronger in proofs, or what)
What rip are you using to color manage the proofer?

 

[danremaley]

The purple colors are made from midtown (40-50-60%) areas of Mag/Cyan, they are the most difficult to control because of dot gain.
i would measure the mid tone gain of your “good” profile sheets & compare the gain of your current sheets. You will have to add a color bar that contains Solid&50% patches of each color. I would also add a grey patch too (50C-40M-40Y). Felix Brunner has it right saying 70% of color shifts are because of dot gain - only 10% caused by SID”.
So what do we measure?
Call with any questions 412.889.7643

 

[SMS]

If you are working with Gracol2013, I recommend contacting Idealliance or a G7 expert and have them walk you through this.
"Tinkering" with your proofing is a nono and same goes for playing with your ink densities on press.
Is you customer using the correct icc profile and are his or her proofs color correct for the destination/production paper (certified to a corresponding standard)?
You yourself should first of all be proofing to Gracol2013, corresponding to the paper you are about to print on, so you know your target colors, otherwise you might as well switch off the lights before you start printing. The icc profile used by your customer needs to be correspond to your output curves, adjusted for output according to Gracol 2013 - and then there is the press of course.

Do you know if your standard densities are correct - M1 LAB values once dry, according to the current ISO 12647 standard?

Best regards

Ingi

 

[roger]

Is this problem solved?
If not, we will need you to answer the questions that have been asked by everyone.

 

[Alois Senefelder]

Gentlemen,

Some information I hope of interest and value.


Regards, Alois.