Mailing Lists: Apple Mailing Lists
Image of Mac OS face in stamp
Re: U.S. Web Coated (SWOP)v2 vs. SWOP2006_Coated5v2
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: U.S. Web Coated (SWOP)v2 vs. SWOP2006_Coated5v2

Hi Todd,
FYI, I'm a Certified G7 Expert (big deal...not) and know a few things about these new SWOP profiles.

Comments inserted....

On Jan 4, 2008, at 7:11 PM, Todd Shirley wrote:

This is more a "topic for discussion" than an actual problem, but it does effect my work and it IS driving me a little crazy, so I'd really like to hear from the colorsync list!

As most of you know, the default CMYK profile in the color settings of photoshop is U.S. Web Coated (SWOP)v2.icc. The profile on my machine is from 2000 and this has been the default as far back as I can remember. This is based on web printing on a #5 coated stock. As most of you also know, there is a much newer profile for web printing on a #5 stock called SWOP2006_Coated5v2.icc which is freely available at Would anybody like to expound on the differences between these 2 profiles? I'd love to hear it!

I know that the new profile is based on averaged (massaged) data from a number of presses calibrated using G7 methodology and that the old one is based on presses set-up to the old SWOP TVI/density numbers, but that's about all I know. Where did Adobe get the old SWOP profile? Did anybody ever use it outside of photoshop? Does anybody know the "story" of that particular profile? On other lists there have been long debates (arguments) about how it turns certain blues purple and its many other shortcomings, but it must have had some validity to become the photoshop default... right? There are a number of differences I can spot easily in Colorthink, not the least of which is that the new profile is about 5x larger (2.6mb vs 550kb) and that it's gamut volume is about 4% larger, but there is some interesting/troubling behavior in photoshop that I'm hoping someone can shed some light on.

You have it mostly right. The new SWOP v11 specification and the resulting data sets and profiles were created using the colorimetric data from ISO 12647-2 (paper types 3 & 5) for solid primaries and secondaries with the "G7 Method" used to create the tone response or "press curves". The G7 method uses what's called the NPDC (Neutral Print Density Curve) instead of TVI/dot gain to calibrate the tone curve of the printing press. The NPDC focuses primarily on 1) gray balance characteristics and 2) tone response as defined by either visual density or L* but NOT TVI/dot gain. Many of us consider this a superior method to dot gain alone since dot gain CANNOT guarantee a specific gray balance nor can it guarantee a consistent visual appearance across different types of printing processes. The G7 method attempts to define a methodology for attaining a common/shared visual appearance (neutrality and tone response) no matter the printing process used and it mostly delivers on that promise.

The "old" or current SWOP profile in Photoshop was derived from the TR-001 characterization data which was based on solid ink density and TVI specifications.

In a perfect world, the new "G7" SWOP and the old SWOP will appear very similar visually but the differ mainly in the methodology used to "get there". There are some slight differences (about 1-2% difference in midtone gray balance) but they are very similar. Also, the new SWOP2006 profile for #3 paper was a long time in coming. The profile/ data set for #3 paper is closer to the reality of what's being printed on today for most web publication printing (#3 is a slightly whiter (less yellow) and brighter stock than the old #5).

Let's say (hypothetically) that I have an untagged CMYK image that I have no idea where it came from or what profile (it any) was used to make it, and I have no way of finding out, and yet I'd like to convert it to RGB to do some color work and then further convert it to a known good CMYK profile (such as SWOP2006_Coated3v2.icc) to proof and release. In the past when faced with this I generally had pretty good luck assigning the photoshop default (U.S. Web Coated (SWOP)v2) and then converting to my RGB working space. Today I was testing this workflow by assigning different CMYK profiles to my mystery CMYK image and looking at the resulting Lab values to see what was "going on", and that's when I really started to wonder about the differences between U.S. Web Coated (SWOP)v2 and SWOP2006_Coated5v2.

Through most of the image there is not too much difference between the two, with the new one being about 2-3 points higher on the L axis. But when I get into the deep shadow, where image detail disappears into black, the old profile causes the image to plug up at a much lighter point and more "sharply". Here are a couple CMYK values and the resultant LABs so you can see what I mean. (Of course you can easily test this yourself).

C, M, Y, K     new SWOP Lab     old SWOP Lab
46,51,59,82      17,4,5                 14,4,7
58,57,63,90      10,3,3                   5,2,4
68,63,62,94	  6,2,1                   1,1,2

You're basically seeing the results of the new profile having slightly less "dot gain" in the midtone and much improved shadow contrast. Depending on the rendering intent you're using, you're also seeing basic difference in the profiling applications themselves. From what I understand, the Photoshop profiles are "hand-built" whatever that means while ALL the new G7 family of profiles (GRACoL and the two new SWOP profiles) are built using Monaco Profiler v4.8. You'll also note that the new profiles having a slightly less "blue" or cyan gray balance and a bit more aggressive GCR and K generation. These are all good things in my book. At the same time, anybody is free to download the GRACoL/SWOP characterization data and build profiles with whatever level of GCR you want (for obvious reasons, one should stick with the recommended total ink limit settings (GRACoL=320%, SWOP=300%) depending on the printing process).

FYI, this image has a max ink limit of about 290%. The first value is somewhat typical of the bulk of the image, and the other 2 show how assigning the new profile leaves the deep shadows more open, with detail visible right up to the ink limit, while assigning the old value plugs up the detail at about 280%. This might seem minor, but it can have a huge effect on dark images, and I am just very curious why it does this. Did old SWOP presses plug up much easier than they do now? Is this just another example of how the old profile is "broken"? In ColorThink I can see that the A2B LUTs are about 13x bigger in the new profile - is there just not enough data in the LUTs of the old profile?

Uh...yes? :-)

A bigger LUT/more profile nodes could be one reason. The other reason could well be rendering intent. From what I know, the Photoshop SWOP profile really only supports the colorimetric rendering intents and not perceptual whereas the new Monaco profiles should have a very good perceptual rendering which would improve shadow detail rendering somewhat.

Taking a quick peek at the profiles using ColorThink Pro, the new SWOP profile has the expected 17 grid points for the A2B while the old SWOP profile has only 9 grid points.

Hope this helps,
Terry Wyse

_____________________________ WyseConsul Color Management Consulting G7 Certified Expert email@hidden 704.843.0858

_______________________________________________ Do not post admin requests to the list. They will be ignored. Colorsync-users mailing list (email@hidden) Help/Unsubscribe/Update your Subscription: This email sent to email@hidden
 >RE: Colorsync-users Digest, Vol 5, Issue 4 (From: "Mark Rice" <email@hidden>)
 >U.S. Web Coated (SWOP)v2 vs. SWOP2006_Coated5v2 (From: Todd Shirley <email@hidden>)

Visit the Apple Store online or at retail locations.

Contact Apple | Terms of Use | Privacy Policy

Copyright © 2011 Apple Inc. All rights reserved.