Creating curves using Curve2

[QualityPrint]

We are building curves based on the G7 methodology using paper that is not G7 compliant. How can we tell the software the lab value of the paper so that it can account for it in the curve. The reason I am asking is that we are seeing more ink put into the quarter tones and lower in the curve and the proof is not showing this color in those areas. We are thinking that the curve2 software isn't taking the paper value into account when building the curves.

Under the gray balance option tab it asks for an a* and a b* but doesn't ask for an L. Is this where we input this data? Is the L value not used?

thanks in advance

 

[mglouis]

The NPDC targets are non-negotiable, meaning you cannot override them in Curve 2. The paper is always zeroed out, regardless of the L* value.

1/4 tones should match closely (+/- 0.02 Den) in terms of NPDC between press and proof, even if the whitepoints do not match.

I would run a P2P through the proofer and make sure it is G7 compliant too.

Matt Louis

 

[airyk]

FYI There are some tutorial videos on youtube for the Curve2 software use

 

[TerryWyse]

With G7 and Curve2, gray balance is always calculated *relative* to paper tint. 100% of the paper tint is removed from the calculation at 0% ink, 50% of paper tint is removed at the midtone, etc.

In Curve2's "Gray Options" you have the choice of calculating gray balance to whatever paper a*/b* you want but it's generally best to click the "Paper White" button so it uses the current paper tint from the P2P measurements....but if you want to force the gray balance calculated to a different paper a*/b* value, you can do that....although it will impart a "cast" to the gray balance. The idea is that gray balance generally looks visually "correct" when it's relative to paper (blue-ish paper should have blue-ish gray balance, yellow-ish paper should have yellow-ish gray balance etc.) as opposed to calculating it as an absolute value.

Paper L* doesn't really come into play since the NPDC is designed to be a relative density scale (density of tint above paper white density) which, again, tends to be more visually "correct".

If you're proofing using, say, the GRACoL ICC profile or the GRACoL dataset and your press stock is quite different from the standard GRACoL-defined paper tint of L*95 a*0 b*-2, you'll naturally see some difference between the proof and the press sheet. An option here is to use the newer "GRACoL Relative" approach and adjust the white point/paper tint in your proof to match your press sheet. The simplest way to do this would be to edit the standard GRACoL profile with the actual L*a*b* of your paper and rename this as your new "GRACoL Relative" profile.

Terry

 

[TerryWyse]

If you're proofing using, say, the GRACoL ICC profile or the GRACoL dataset and your press stock is quite different from the standard GRACoL-defined paper tint of L*95 a*0 b*-2, you'll naturally see some difference between the proof and the press sheet. An option here is to use the newer "GRACoL Relative" approach and adjust the white point/paper tint in your proof to match your press sheet. The simplest way to do this would be to edit the standard GRACoL profile with the actual L*a*b* of your paper and rename this as your new "GRACoL Relative" profile.

Just to be clear on this last point, this assumes you're using "absolute colorimetric" rendering when proofing...if you use relative colorimetric or, god forbid, perceptual rendering, no amount of editing the GRACoL white point will have any effect. If you're using relative colorimetric rendering currently, that could be a reason you're not able to match your proofs now.

Just wanted to make this clear since the guy I quoted didn't exactly spell it out in detail.

Terry

 

[QualityPrint]

What I think I am seeing is the addition of dots in lower percentage areas to compensate for the stock difference. Our press sheet specs out to l 92.5 a 1 b -4. So we have a bluish/red paper that is darker than the spec. I'm thinking that the Curve2 software is reading lower percentage areas as blue/red thus cutting back the blue/red in the curve. Our proofs always look warmer than the press sheet. I believe that if we start with a G7 specified sheet we will avoid this addition of dots and get a more accurate curve that will work on various stocks. Am I totally wrong?

 

[meddington]

Not sure the subtle change in recommended plate curves between your press paper and a "G7 specified" paper would be enough to bridge the visual difference between your proof (emulating the Gracol2006coated1 white point) and your press (using a non-standard stock). I personally think your best bet is the advice Terry gave you...adjusting your proofing profile to emulate your press stock.

 

[TerryWyse]

Our press sheet specs out to l 92.5 a 1 b -4. So we have a bluish/red paper that is darker than the spec.

Anytime I see that strong of a -b* coupled with a +a* value I automatically think "optical brighteners". The low L* of the paper tells me this is either a commercial web offset stock or a very poor commercial sheetfed stock. Kind of puzzling values to be honest...you generally don't see an L* that low with such a strong blue tint or optical brighteners. Just for grins, you should try another spectro just to make sure it agrees with these values....best option would be to use the spectro that will later be used with your proofing system to create the proof/output profile.

I think the GRACoL colorimetry is going to get mucked up as well.....for example, if you try to hit the green/yellow/red absolute colorimetry with a paper that blue, you're liable to overshoot the yellow ink density by a good bit.....what I mean is, by the time you hit the yellow b* value (chroma) for GRACoL, you might be laying down way too much ink...it'd be interesting to see what the yellow gray balance curve looks like....I'll bet it's a bit whacked and out of balance compared to the cyan and magenta curves.

I have some ideas on how to handle this....contact me off-list if you like.

Terry

 

[tmiller_iluvprinting]

We experienced the same issue when we used the GRACol data set to set up our proofer. The press sheets were heavier in the highlight/quartertone areas than our proof created with the GRACol data set. We switched and used our press data to create our proofing profile and the issue went away. We print using 200LS and it is my understanding that the GRACol data was created from press sheets printed with 175LS. I am wondering if this was the root cause of the issue? FYI-our house sheet is GRACol compliant. We also had issues with the red overprints matching, and we still have issues getting our red overprint to agree with the GRACol Lab values.
Regards,
Todd

 

[QualityPrint]

Thanks for your reply Todd. Can you tell me what type of paper you are using? I haven't been able to find a sheet with a b* lower than a -3.8. We are switching to an eye 1 instead of a dtp70 for a few reasons and hope that this along with using the right paper will help us out.
Jeremy

 

[tmiller_iluvprinting]

Our house sheet is Productolith Gloss(just qualifies!). It is a New Page product. There are other sheets out there as well, Unisource gloss would be another.
Regards,
Todd

 

[TerryWyse]

Thanks for your reply Todd. Can you tell me what type of paper you are using? I haven't been able to find a sheet with a b* lower than a -3.8. We are switching to an eye 1 instead of a dtp70 for a few reasons and hope that this along with using the right paper will help us out.
Jeremy

I doubt whether changing to a different spectro is going to make a difference.

Back to your original post, if you're using the Curve2 software, you can either accept the measured paper white L*a*b* or input a custom value....you could play with that and see what affect it has on the G7 curve values and make a judgement and what value to use.

Since you have a DTP70 and you can measure the P2P targets with either UV included or excluded, measure the P2P targets with both filters and use that in the Curve2 software.....you can easily import both sets of measurement data and switch them on/off and look at the curve values you get....you seem to have an idea of what you want already so just use the data set/filter setting that comes closest to that.

Terry

 

[TerryWyse]

Under the gray balance option tab it asks for an a* and a b* but doesn't ask for an L. Is this where we input this data? Is the L value not used?

The L* value of the paper is irrelavent since G7 by definition is paper-relative (paper excluded).

If your typical press stock has an L* value that's more than about delta 1-2 L* from GRACoL and you're using GRACoL for proofing, then I believe you'll have to use the "GRACoL Relative" approach and adjust your proofing system accordingly.

Terry