Dan Margulis Applied Color Theory

Uncoated Stock, Coated Profile

   Date: Mon, 15 Dec 2003 07:17:26 -0800
   From: Doug Walker
Subject: printing uncoated stock printed with coated specs

I have a client who has switched papers from a matte coated to a matte uncoated stock but who still is converting images with U.S. Web Coated (SWOP) v2.

What is basically happening if one converts to a Coated profile when actually using an Uncoated stock?  Does it mean ink flow out and thus less sharpenss?  Muddiness?  Heavier blacks?

TIA

Doug Walker, FP
“Specializing in Coprorate People on Location in a Clean, Bold Classic Style!”
website: http://www.walkerphoto.com
Member, PPW, ASMP, APA SF
___________________________________________________________________________

   Date: Mon, 15 Dec 2003 14:22:34 -0500
   From: Terry Wyse
Subject: Re: printing uncoated stock printed with coated specs

The biggest difference in the separation itself should be the Total Ink Limit/Total Area Coverage. IOW, if it’s separated for WebCoated, the ink limit (280-300%) will be too high for an uncoated stock (should be 240-260%). This won’t cause “heavier blacks” as the maximum density on uncoated will be much less but it WILL likely cause plugging/blocking of important shadow detail.

Solid ink density will drop of course but “it is what it is” so you wouldn’t necessarily compensate for this in the separation. If the solid ink density is kept at the proper level, dot gain does not increase significantly in my experience.

A quick and dirty fix would be to perform a profile-to-profile conversion from USWebCoated to USWebUNCoated to correct the total ink limit and dot gain. Dot gain would be easy enough to fix using curves but a total ink limit correction would be difficult at best. Profiles would be by the easiest and most accurate way to do this. RelCol rendering with BPC on would be your best bet.

Cheers,
Terry
—
__________________________________
WyseConsul
Color Management Consulting
v 704.843.0858
__________________________________
___________________________________________________________________________

   Date: Tue, 16 Dec 2003 07:58:55 EST
   From: Dan Margulis
Subject: Re: printing uncoated stock printed with coated specs

Doug writes,

What is basically happening if one converts to a Coated profile when
actually using an Uncoated stock?  Does it mean ink flow out and thus
less sharpenss?  Muddiness?  Heavier blacks?

Muddiness is probably the best description because the dot gain is being underestimated, therefore the separation is being made too dark, even though it will preview correctly on the screen.

To get an idea of what is occurring, open a typical CMYK file and set your Color Settings>CMYK to Custom CMYK, dot gain 20%. Assume for the sake of argument that what you see on the screen now is what you would like to see in print.

Now change the dot gain to 25%. The new preview will give you a reasonable idea of what it *will* look like in print if you were assuming coated and in fact it prints uncoated.

Dan Margulis
___________________________________________________________________________

   Date: Wed, 17 Dec 2003 12:19:57 -0500
   From: “Remaley, Dan”
Subject: RE: printing uncoated stock printed with coated specs

The uncoated sheet will have more gain than the coated sheet. Typically, the pressman will run the density ‘down’ to compensate for the higher gain, resulting in a ‘flatter’ reproduction. Removing about 4% in the midtones would help.

Dan Remaley
Process Control Mgr.
Graphic Arts Technical Foundation
412.741.6860x450
___________________________________________________________________________

   Date: Wed, 17 Dec 2003 13:28:51 -0500 (EST)
   From: Ray Swanson
Subject: RE: printing uncoated stock printed with coated specs

Hi Dan:

Does “gain” actually mean “spread”... that is, a higher gain means a wider spread of ink, ergo, the tighter the paper, the less gain/spread and therefore the higher potential resolution? Is this elementary or is there more to it than this?

Ray
___________________________________________________________________________

   Date: Thu, 18 Dec 2003 14:50:48 +0100
   From: Axel Ritz
Subject: Re: printing uncoated stock printed with coated specs

Hi Dan:

On uncoated paper it’s the dot-gain only, if you exaggerate the thickness of the printed ink film layer.

If you use an adopted ink film thickness, you will have the same dotgain on both papers.

On uncoated paper it’s a brow whereas on art paper you get a bright orange.

The difference between brown and orange is the amount of black. Consequently you have to reduce the amount of black in pre-press-work in all pictures printed on uncoated paper compared to prints on art paper. (I’ve been talking on offset, flexo inks doesn’t have a draw back like offset, they stay on the paper surface an have more brightness on uncoated paper.)

Axel
___________________________________________________________________________

   Date: Fri, 19 Dec 2003 07:23:06 EST
   From: Dan Margulis
Subject: Re: printing uncoated stock printed with coated specs

Axel writes,

On uncoated paper it’s the dot-gain only, if you exaggerate the thickness of the printed ink film layer. If you use an adopted ink film thickness, you will have the same dotgain on both papers.

I am not familiar with the phrase “adopted ink film thickness”. I believe it probably corresponds to the English “reduce ink density” or to the phrase used by Dan Remaley “run the density down”.

If that is the meaning, no, the dot gains will still be different. The inherent characteristics of the sheet cannot be changed. This tactic merely substitutes one problem for another.

If a pressman is truly faced with attempting to match a coated result using the same digital file but now printing on uncoated, then he will probably decide that the least evil approach is to do what you suggest, reduce the amount of ink hitting the paper. Now the uncoated image will not look as dark and muddy as it would otherwise; however it will appear flat because the shadow will unavoidably be lightened also.

The correct approach is a different separation that leaves the shadow as dark as it was (consistent with the ink limit) but lightens the midtone. Without such a separation, no amount of inking shenanigans will be able to match the desired look.

On uncoated paper it’s a brown whereas on art
paper you get a bright orange. The difference between brown
and orange is the amount of black.

No. In a bright orange there will be little to no cyan or black. 0c50m90y is a bright orange even if printed on newsprint.

Even in a considerably duller orange, little black is found even when extreme GCR is being used. The culprit, if there is one, would be the cyan, which would be five or ten times heavier than the black in most orange areas and much more subject to the effect of dot gain.

Consequently you have to reduce the amount of black in pre-press-work in
all pictures printed on uncoated paper compared to prints on art paper.

One has to reduce the amounts of *CMY* in such circumstances. It’s customary, but not required, to reduce the black in the lighter half of the image as well. However, in the shadow region of the image the black actually needs to be *increased* in comparison to a coated sep, as the CMY components of the shadows will have to be reduced to stay within the lower total ink limited required for uncoated printing. The increased black is the only way to retain a dark shadow while honoring the ink limit.

Dan Margulis
___________________________________________________________________________

   Date: Fri, 19 Dec 2003 12:14:58 -0800
   From: Doug Walker
Subject: Re: printing uncoated stock printed with coated specs

On Friday, December 19, 2003, at 04:23  AM, Dan Margulis wrote:

The correct approach is a different separation that leaves the shadow as dark
as it was (consistent with the ink limit) but lightens the midtone.

Would using a profile such as the SWOP TR001 CHROMiX ltGCR280.icc instead of USWebCoatedSWOPv2 be a step in the right direction in lightening the midtones?  This profile would aready have the light instead of medium and 280 instead of 300.

Or is it not that simple.

Doug Walker, FP
___________________________________________________________________________

   Date: Fri, 19 Dec 2003 21:24:57 -0500
   From: Jim Rich
Subject: Re: printing uncoated stock printed with coated specs

On 12/19/03 3:14 PM, “Doug Walker” wrote:

Would using a profile such as the SWOP TR001 CHROMiX ltGCR280.icc
instead of USWebCoatedSWOPv2 be a step in the right direction in
lightening the midtones?  This profile would aready have the light
instead of medium and 280 instead of 300.
 
Doug,

It almost sounds like you are searching for a magic bullet  to solve a number of color reproduction issues. I think it is not that simple.

The main purpose  of specifying total area coverage in percent dots is to help the printing process and get the optimum amount of ink on paper with a certain kind printing press. There might be other benefits to TAC such as a change in the black printer that might affect midtone or even affect midtone placement.

Theoretically, if you have a good color reproduction process that can print say 280%, 300% and 320% and you make color separations for each of those processes with the same intent to have them look the same, the end results should be that the reproductions  look almost the same.
 
IMHO you might want to consider getting a custom profile that is made for your process instead of trying to use another type of profile and hope that it works. If you can find one then you lucked out but the question is what does that cost you in time and wasted prints etc.

Just my .02.

Jim Rich
___________________________________________________________________________

   Date: Sat, 20 Dec 2003 07:47:01 -0000
   From:Stephen Marsh
Subject: Re: printing uncoated stock printed with coated specs

Doug Walker wrote:

Would using a profile such as the SWOP TR001 CHROMiX ltGCR280.icc
instead of USWebCoatedSWOPv2 be a step in the right direction in
lightening the midtones?  This profile would aready have the light
instead of medium and 280 instead of 300.  Or is it not that simple.

Hi Doug, simply put - no.

Both the Adobe v2 and the Chromix profiles describe TR001 - and both produce near identical values in the midtones as they are both describing the same gain.

The Chromix range of profiles offer a more flexible choice of GCR settings or UCR and various TAC which is not SWOP.

Your initial post indicated that the preview of the v2 profile is great, but the seps it creates are not ideal for press, which would prefer 280% TAC and light K gen - which is not what the v2 profile offers.
 
Stephen Marsh.
___________________________________________________________________________

   Date: Sat, 20 Dec 2003 07:55:56 EST
   From: Dan Margulis
Subject: Re: printing uncoated stock printed with coated specs

Doug writes,

Would using a profile such as the SWOP TR001 CHROMiX ltGCR280.icc
instead of USWebCoatedSWOPv2 be a step in the right direction in
lightening the midtones?  This profile would aready have the light
instead of medium and 280 instead of 300.

I have no idea as I have never seen the first profile. Fortunately, you can find out for yourself. Make yourself two copies of a typical CMYK picture. With the first, Image: Mode>Assign Profile>SWOPv2. With the second, Image: Mode>Assign Profile>Chromix.

Now compare the two side by side. If the Chromix version looks darker, then it’s a step in the right direction (that is, it would generate a lighter sep from the same RGB data, which is what you want).

However, rather than just guess at what profile is right, or get into the cycle of having other people generate profiles for you, I’d just learn how to make the adjustments in Custom CMYK and forget about these problems.

Dan Margulis
___________________________________________________________________________

   Date: Sun, 21 Dec 2003 02:04:13 -0000
   From:Stephen Marsh
Subject: Re: printing uncoated stock printed with coated specs

I wrote:
 
Both the Adobe v2 and the Chromix profiles describe TR001 - and both
produce near identical values in the midtones as they are both
describing the same gain.

When using RelCol, the Adobe SWOP and the Chromix TR001 profiles deliver near identical results - which is what I based the above comment on. I originally mentioned that this Chromix profile was not as tame as the Adobe one, as the Chromix one has a more ‘creative’ perceptual intent than the Adobe profile that describes TR001. The Chromix profile does deliver an approx. 4% lighter midtone when using the perceptual intent.

When I originally mentioned that there was a Chromix profile which did have 280TAC and Lt GCR which was based on TR001 data, the intent was that this profile did offer black ink and paper white simulation in soft-proofing - which the legacy custom CMYK controls do not offer, as Doug had good previews with these options in the Adobe SWOP profile.

To refresh the lists memory on this particular thread:

Doug has great previews when he uses the Adobe SWOP v2 profile for soft-proofing. The press would ideally like 20% lesser TAC and lighter K generation than this profile delivers.

I see three options:

i) Separate with the Adobe profile and let things ride/and or hand edit the TAC - but the GCR is more complex to change via channel mixer

ii) Separate with legacy custom CMYK settings for total control of TAC and K gen, then soft-proof using the Adobe SWOP v2 profile (preserve) colour numbers on when soft-proofing).

As the custom CMYK is not the same as the v2 profile, this will not be a faithful reproduction of the original and will either require edits to the custom CMYK inks and or post separation edits to get things back into shape.

iii) Use the Chromix profile which does offer lower TAC and lighter K generation. One can then choose to soft-proof with the known Adobe profile or see how the Chromix profile soft-proofs (preserve) colour numbers on when soft-proofing).

Doug’s question about having a lighter dot gain in his separations is new, and perhaps picked up from another thread? The press guys would like lower TAC and lighter K generation, this does not mean they would like more gain built into the seps.

So Doug, are your seps now too dark?

Stephen Marsh.
___________________________________________________________________________

   Date: Mon, 22 Dec 2003 07:56:00 -0500
   From: Dan Remaley
Subject: RE: Re: printing uncoated stock printed with coated specs

Hi Stephen, I have a question about Photoshop. The settings for SWOP coated are equal dot gains for Y-M-C-K - True? If you look at the GRACoL chart, yellow is always 2% less and black 2% more with cyan and magenta of equal value. In fact this IS "gray balance" at press. Any comments, please.

Dan

Dan Remaley
Process Control Mgr.
Graphic Arts Technical Foundation
412.741.6860x450
___________________________________________________________________________

   Date: Mon, 22 Dec 2003 10:35:37 -0500
   From: Dan Remaley
Subject: RE: printing uncoated stock printed with coated specs

Hi Ray, yes there are 2 components to 'dot gain' (or Tone Value Increase - TVI) optical and mechanical. Most people never realized the 'optical' component. If you look at a traditional Matchprint, its dots are much smaller than a printed sheet (little or no mechanical gain). On press the ink spreads and is absorbed by the paper (mechanical and some optical gain). Then how can they 'match'? The proofs optical gain is equal to the press mechanical gain and  a similar 'weight' is achieved. Even today with digital proofs these numbers are reflected at <swop.org> density, dot gain, print contrast are referenced to 'visually' match a SWOP press sheet. The coated sheets have less gain because the smooth surface holds the ink away from spreading.

For anyone who's interested - send me your e-mail address and I'll return some PDF's on process control and how to use color bars and proof comparators or visit <www.gain.net>. GATF offers a new consulting service called "System GATF", a group of 2 or 3 consultants work together in solving your color matching problems. Any questions please call. . .  

Dan Remaley
Process Control Mgr.
Graphic Arts Technical Foundation
412.741.6860x450
___________________________________________________________________________

   Date: Mon, 22 Dec 2003 22:16:32 -0000
   From: Stephen Marsh
Subject: Re: printing uncoated stock printed with coated specs

Dan Remaley wrote:

Hi Stephen, I have a question about Photoshop. The settings for SWOP coated
are equal dot gains for Y-M-C-K - True?

Hi Dan, if discussing the legacy custom CMYK engine, which is now profile based - then yes in standard mode this is the case (in [custom] curves the cyan runs at 24%c 20%myk - which Dan Margulis has commented on before).

The Adobe U.S. Web Coated (SWOP) v2 profile is based on TR001 data, as is the Chromix TR001 range of profiles - which are the other main options apart from a custom profile, if I understand Doug's situation. These profiles do indicate different gains in each plate (I measured this a while back and posted it to the list) - but I do not have the data at hand (same gain, same colorimetric description, same measurement data, different separation options, different vendors, different results).

I think the time has been and gone for Adobe to revamp the legacy custom CMYK engine - now that ICC profiles are here. That being said, users can make their own adjustments to the defaults if they wish. The current 'de-facto standard' Adobe v2 profile for TR001 does have different gain for each plate which is a step in the right direction

 I am not sure if it meets the magic number aim-points dictated by SWOP, but from poor memory it does sound close.

Stephen Marsh.
___________________________________________________________________________

   Date: Mon, 22 Dec 2003 16:18:23 -0800
   From: "J Walton"
Subject: RE: Re: printing uncoated stock printed with coated specs

Hmm...

That doesn't sound gray balanced to me, but I'm not all that familiar with the GRACol chart.  Shouldn't yellow be more than 2% low?

J
___________________________________________________________________________

   Date: Wed, 24 Dec 2003 01:36:45 EST
   From: Dan Margulis
Subject: Re: Re: printing uncoated stock printed with coated specs

Stephen writes,

Hi Dan [Remaley], if discussing the legacy custom CMYK engine, which is now
profile based - then yes in standard mode this is the case (in
[custom] curves the cyan runs at 24%c 20%myk - which Dan Margulis has
commented on before).

In Custom CMYK, even if you go in and edit the setting yourself, if you enter "Standard" for Dot Gain and pick a number X at random for your percentage, you will get MYK=X and C=X+4. This is totally balled up and dates from the early 1990s.  Thomas Knoll says that it was a kludge for something but he forgets what. Presumably, his seps were coming out a little cold and he hacked it this way to try to fix it, rather than reaching the correct conclusion that his black was too heavy.

Those who use Custom CMYK should correct this. After choosing what seems to be the appropriate dot gain percentage, enter the "Curves" option and reduce the cyan setting so that it equals that of the magenta. The black definitely needs to be set higher and I would recommend four points although some authorities say two. And the yellow, in principle, should be set two points lower than cyan and magenta, although in real life it doesn't make much of a difference if this step is omitted. The black increase, however, is quite important for accurate separations.

I think the time has been and gone for Adobe to revamp the legacy
custom CMYK engine - now that ICC profiles are here. That being said,
users can make their own adjustments to the defaults if they wish.

Custom CMYK can make excellent profiles but it's a real beast if you haven't had a lot of experience in doing this kind of thing by hand. Because its so heavily kludged, just measuring ink values with some artificial instrument and plugging them in won't work. Instead, you have to enter new ink values in LAB or xyY based on your own visual perceptions of what's wrong with the existing profile.

Me, I've profiled so many different devices using so many screwed-up interfaces that I find no difficulty doing this, but it's hell for the average user--or the average color management consultant, for that matter. Plus, there really aren't enough options to do a perfect job with it, which is ridiculous in this day and age.

But, until Adobe wakes up and does what Stephen suggests, we're stuck with it. If one thing is clear after nearly six years of experience with Photoshop and ICC color management, it's that high-end users won't migrate to something that doesn't allow them to edit profiles within Photoshop. That Adobe hasn't added a full-featured profile editor during this time is very regrettable.

Dan Margulis
___________________________________________________________________________

   Date: Wed, 24 Dec 2003 11:41:40 +0100
   From: claudio.corvino
Subject: Re: Re: printing uncoated stock printed with coated specs

Those who use Custom CMYK should correct this. After choosing what seems to
be the appropriate dot gain percentage, enter the "Curves" option and reduce
the cyan setting so that it equals that of the magenta. The black definitely
needs to be set higher and I would recommend four points although some
authorities say two. And the yellow, in principle, should be set two
points lower than cyan and magenta, although in real life it doesn't make much of a difference if this step is omitted.

Is it true for Euroscale coated v2 also? So the idea that cyan must be set higher due to "inks impurity" is a tale?

The black increase, however, is quite important for
accurate separations.

Can you please go deep into it? I undestand that light black is good (coated)

Thank you!

--
Claudio Corvino
Florence, Italy
___________________________________________________________________________

   Date: Wed, 24 Dec 2003 21:19:09 -0600
   From: "Howard Smith"
Subject: Re: Re: printing uncoated stock printed with coated specs

  Dan Margulis writes:

In Custom CMYK, even if you go in and edit the setting yourself, if you enter
"Standard" for Dot Gain and pick a number X at random for your percentage,
you will get MYK=X and C=X+4. This is totally balled up and dates from the early 1990's.  Thomas Knoll says that it was a kludge for something but he forgets what. Presumably, his seps were coming out a little cold and he hacked it this way to try to fix it, rather than reaching the correct conclusion that his black was too heavy.

 Dan,

      My understanding is that Thomas Knoll's awareness of the relative weakness of cyan compared to magenta and yellow caused him to conclude that cyan required a higher dot gain value to compensate for that weakness.  I'm curious to know if this account is an urban legend.  Wasn't a higher dot gain value used for cyan before Thomas Knoll's work?   As for your point about reducing cyan dot gain and increasing black dot gain, that makes sense.  The small increase in black substitutes very nicely for Cyan in addition to giving slightly greater contrast to the image.  But while you recommended raising the percentage of black, you commented at Knoll's black was too heavy.  Is this a contradiction, or did I miss something?

  Howard Smith
___________________________________________________________________________

   Date: Fri, 26 Dec 2003 19:33:29 EST
   From: Dan Margulis
Subject: Re: Re: printing uncoated stock printed with coated specs

Claudio writes,

Is it true for Euroscale coated v2 also?

Euroscale coated v2 is a non-editable profile. You have to take it or leave it the way it is. Based on what I have seen of it, I'd forget it and try to work with the older Eurostandard settings, which have a ridiculously low default dot setting but at least can be edited.

And yes, the error of setting the cyan dot gain four points higher exists in the Eurostandard profiles as well.

So the idea that cyan must be set higher due to "inks impurity" is a tale?

Yes. That is compensated for elsewhere.

Can you please go deep into it? I undestand that light black is good
(coated) Thank you!

There are several threads on this general topic (GCR) at

http: //www.ledet.com/margulis/ACT-postings/SeparationIssues/Separation.htm

Unlike RGB, where a given color can only be described in one way, in CMYK almost all colors can be described in several different ways, by adjusting the black up or down and making a counteradjustment in the CMY. So, a black channel can be light or heavy and it theoretically makes no difference as long as the countermove in CMY was done correctly.

The problem with heavier blacks is that black is such a powerful ink that if there is any error in separation or in printing the job can be ruined. With lighter blacks the problem is much less severe, which is why historically the industry has preferred a "skeleton", or light, black channel.

The Custom CMYK profiles we have been discussing give a) a relatively heavy black; b) one that underestimates its dot gain, therefore producing a heavier black yet. The combination of the two is deadly. Those wishing to use Custom CMYK are best off changing both.

Dan Margulis
___________________________________________________________________________

   Date: Mon, 29 Dec 2003 09:24:29 EST
   From: Dan Margulis
Subject: Re: printing uncoated stock printed with coated specs

Howard writes,

Dan,
My understanding is that Thomal Knoll's awareness of the relative
weakness of cyan compared to magenta and yellow caused him to conclude that cyan
required a higher dot gain value to compensate for that weakness.  I'm curious
to know if this account is an urban legend.

As indicated in my initial post, Thomas says he does not remember what his reason was. As this involves work he did more than ten years ago, I think we can accept that at face value.

Wasn't a higher dot gain value used for cyan before Thomas Knoll's work?

No. The relative dot gains of the various inks has been understood for a long time.

As for your point about reducing cyan dot gain and increasing black dot
gain, that makes sense.  The small increase in black substitutes very nicely for
Cyan in addition to giving slightly greater contrast to the image.  But while
you recommended raising the percentage of black, you commented at Knoll's
black was too heavy.  Is this a contradiction, or did I miss something?

I did not recommend "raising the percentage of black." I recommended increasing the assumed dot gain to conform to the known reality that black has greater dot gain than CMY.

Since the Photoshop defaults underestimated the black dot gain, the separations using those assumptions were too heavy in the black.

Dan Margulis

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