Dan Margulis Applied Color Theory

SWOP Gray Balance

   Date: Tue, 8 Feb 2005 18:32:05 -0800
   From: Steven Barton
Subject: SWOP gray balance

The values needed to make a neutral gray in a SWOP print condition at the 50% point, according to the SWOP web site and GATF are 50C, 39M, 39Y, which has an 11-point spread between the cyan and the magenta.

Yet, the industry standard for conversion to CMYK for the SWOP print condition is the Photoshop profile USWebCoatedSWOP-v2, which produces a much different gray balance (44C, 36M, 37Y, 2K)--only an 8-point spread. The Photoshop profile is built from the SWOP colorimetric data.

Does anyone know why there is such a significant discrepancy?

Steven Barton
Imaging Sciences
Portland, Oregon
___________________________________________________________________________

   Date: Wed, 9 Feb 2005 08:45:48 -0500
   From: Dan Remaley
Subject: RE: SWOP gray balance

I believe the Photoshop settings are wrong. The 50C-39M-39Y patches are to be placed on the proof and press color bar. The  SWOP numbers at press are Y-1.00-18% gain  M-140.-20% gain C-1.30-20% gain  K-1.60-22% gain (this works for sheetfed as well) AND all the seps. are made SWOP specs.

Here's Photoshop's numbers for gain Y-20  M-20 K-20 C-24 Know anyone that prints like that - I don't!

Photoshop has a 4% Cyan bias to the neutral in the midtone.  example- if you take 50C-39M-39Y and apply maximum GCR - what would you expect - (Answer - 50% Black).  If done in Photoshop you get 46% Black 4% Cyan - HUH?   Oh well.
Dan

Dan Remaley PIA/GATF
Process Control Manager
412.2591814
___________________________________________________________________________

   Date: Wed, 9 Feb 2005 11:27:30 -0500
   From: Terry Wyse
Subject: Re: SWOP gray balance

For USWebCoated(SWOP)v2, the gray balance is 50/41/41 or 50/42/42, depending on whether you ignore K or not (it's best to reference a 50% cyan if you want to compare apples2apples).  It's important to understand that the relative "spread" between cyan and m/y for a neutral 3-color gray varies depending where on the scale you take your measurement. At 5%, it's only 1-2%, at 25% it's around 6-8%. Generally at 60% cyan is where the spread "peaks" at around 9-12% and stays there until you reach max density/total ink limit, sometimes actually narrowing a bit depending on the neutrality of the black.

You have to be VERY careful about the K ink. It's almost never perfectly neutral and can have quite an impact on the 3-color gray balance in an ICC profile, especially as you reach the total ink limit. For example, the K coloration or hue with SWOPv2 is a bit on the magenta side. For USSheetfedCoatedv2, it's just a bit on the blue side. In either case, this would cause the profile to alter the 3-color gray to counter the hue of the K ink resulting in a separation that's not by "the numbers" but is actually more neutral. On the other hand, at least in the case of a blue/cold K ink, a neutral gray that's a bit on the blue side is usually less noticeable than a gray that is tending towards magenta. For my money, USSheetfedCoated produces the closest thing to "classic" neutral gray balance numbers that I was used to using back in the days of drum scanning. The only drawback of that profile is that it assumes a bit too much dot gain (22-24%) and thus produces a separation that would print too clean/sharp on something targeted closer to SWOP and especially linear CtP. For my own stuff, I've actually edited that profile to give me something much closer to SWOP/GRACoL dot gain/TVI while still giving me the classic gray balance numbers.

Regards,
Terry

_____________________________
WyseConsul
Color Management Consulting
704.843.0858
http://www.colormanagementgroup.com
___________________________________________________________________________

   Date: Wed, 9 Feb 2005 11:15:31 EST
   From: Dan Margulis
Subject: Re: SWOP gray balance

Steven Barton writes,

The values needed to make a neutral gray in a SWOP print condition at
the 50% point, according to the SWOP web site and GATF are 50C, 39M,
39Y, which has an 11-point spread between the cyan and the magenta.

Correct.

Yet, the industry standard for conversion to CMYK for the SWOP print
condition is the Photoshop profile USWebCoatedSWOP-v2, which produces a
much different gray balance (44C, 36M, 37Y, 2K)--only an 8-point
spread.

That profile is certainly not an industry standard--it can be a useful starting point, but it has many holes, of which this is one. There was a lengthy thread here a couple of months ago about the fact that this profile displays light neutralish areas as too blue in comparison to how they would be printed almost anywhere. The gray balance issue is the other side of that coin: it wants to make the midtones warmer than they actually are, therefore it considers actually neutral colors to be too cold.

The Photoshop profile is built from the SWOP colorimetric data.

Whether built from "colorimetric data" or banana peels, the quarter- to midtones separate too warm/display too cold.

Does anyone know why there is such a significant discrepancy?

Accurate CMYK profiles are hard to make. The mere fact that machines have measured certain values by no means guarantees an effective result, but it cons some neo-calibrationists into thinking things must be right when they are obviously wrong.  This profile displays too blue for all print conditions; also it produces a darker separation than the Photoshop "sheetfed" profile, which is preposterous. Yet there are those who would tell you that because a machine measured them that way, it must be right, and all the printers are wrong.

Dan Margulis
___________________________________________________________________________

   Date: Wed, 9 Feb 2005 11:52:32 -0500
   From: Terry Wyse
Subject: Re: SWOP gray balance

Yo DanR,

Those numbers are what you get with the semi-retarded "Classic CMYK" engine in Photoshop. If you build a "SWOP" profile that way versus what you get with USWebCoated(SWOP)v2, you'll get two different profiles. If you want a better "SWOP" profile in terms of gray balance, use an edited version of USSheetfedCoatedv2 that corrects the too-high dot gain/TVI assumed by that profile. Now you'll get the numbers that Mr. Remaley and us ex-drum scanner guys know and love!

Take Care,
Terry

_____________________________
WyseConsul
Color Management Consulting
704.843.0858
http://www.colormanagementgroup.com
___________________________________________________________________________
 
   Date: Wed, 9 Feb 2005 11:37:12 -0700
   From: Ron Kelly
Subject: Re: SWOP gray balance

How does one edit a profile such as this? Is it done through photoshop, or do you need third party software? If so, such as?

Thanks,
Ron Kelly
___________________________________________________________________________

  Date: Wed, 9 Feb 2005 14:37:08 -0500
   From: Terry Wyse
Subject: Re: SWOP gray balance

On Feb 9, 2005, at 11:15 AM, Dan Margulis wrote:

Accurate CMYK profiles are hard to make. The mere fact that machines have
measured certain values by no means guarantees an effective result, but it cons
some neo-calibrationists into thinking things must be right when they are
obviously wrong.  This profile displays too blue for all print conditions; also it
produces a darker separation than the Photoshop "sheetfed" profile, which is
preposterous. Yet there are those who would tell you that because a machine
measured them that way, they must be right, and all the printers are wrong.

Right or wrong, SOME presses will produce neutral gray using non-standard values. Is the press wrong or are the numbers wrong? Neither in my view. A press printing at a given condition can produce any combo of 3-color gray values and the profile will accurately reflect that. The fact is, the profile is showing you the gray recipe for that particular press assuming it's printing the same as it did when the profile was made. And there's nothing at all "wrong" about that as long as the printer understands the implications. The implications are 1) this press profile will work well for any internal separations that they make, 2) this profile will work well if any of their customers provide separations using this profile, 3) any outside tagged separations converted to their print conditions using this profile will print well and 4) any separations done NOT using this profile will work less so! So right/wrong is all a matter of perspective.

At the same time, in my press profiling work I advocate adjusting the press, either mechanically or using plate curves, to a standard dot gain/TVI/solid density spec using a target that's appropriate for their type of printing. This usually means either SWOP or GRACoL numbers. Once this has been accomplished, you profile the press and, 9-time-out-of-10, the gray balance will be either spot-on or within 1-2% of the target. Is this "better"? Only in the sense that it allows them to take in all manner of outside separations and print/proof them with a minimum of fuss. *I* think this is a good thing because it LESSENS their requirement to get neck-deep into color management but I'm not so arrogant to say that this is the only "right" way and that printing to some internal non-standard is "wrong", it's just different. While I respect the "historical" value of how things have been done (I've been there too), I also don't think it's a inherently a bad idea to explore alternative ways of calibrating or optimizing a press to achieve the highest quality separations possible (look at Bill Atkins work over in Asia as an example).

As soon as I can convince somebody to give me a few skids of paper, a 40" press and about 3 weeks of time, I'm ready to do some exploring! Dan Remaly, how about loaning me "your" pressroom for some really radical ways of approaching press optimization? Oh, and I also need to get PAID for all this experimentation! :-)

This could be very enlightening even if points back to the fact that we've been doing it the "right" way all along!

Cheers,
Terry
_____________________________
WyseConsul
Color Management Consulting
704.843.0858
http://www.colormanagementgroup.com
___________________________________________________________________________

   Date: Wed, 9 Feb 2005 14:37:18 -0500
   From: Terry Wyse
Subject: Re: SWOP gray balance

On Feb 9, 2005, at 1:37 PM, Ron Kelly wrote:

How does one edit a profile such as this? Is it done through photoshop,
or do you need third party software? If so, such as?

3rd party. I use the Gretagmacbeth Profile Editor module and also the profile editor that comes with MonacoPROFILER. In keeping with the current topic, I use this to "correct" certain press profiles that I've done to give them typical gray balance behavior. The only catch is, the press has to be "modified" (plate curves usually) to reflect the edits in the profile.

Terry

_____________________________
WyseConsul
Color Management Consulting
704.843.0858
http://www.colormanagementgroup.com
___________________________________________________________________________

   Date: Wed, 9 Feb 2005 20:54:12 -0500
   From: "Jerry P'Simer"
Subject: Re: SWOP gray balance

On Feb 9, 2005, at 11:27 AM, Terry Wyse wrote:

You have to be VERY careful about the K ink. It's almost never
perfectly neutral and can have quite an impact on the 3-color gray
balance in an ICC profile, especially as you reach the total ink limit.
For example, the K coloration or hue with SWOPv2 is a bit on the
magenta side.

This is interesting because we were just informed by Kodak that they are changing the pigments of their donors for the Kodak Approval. Most notably the black, which will be changed first, is even more magenta than it has been. In my opinion it was too magenta to start with and to make it more so is disturbing to me. Does anyone know why this is? I asked the Kodak tech guy that was upgrading our Approval software and our Kodak sales rep but could not get an answer.

Jerry P'Simer

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