Photoshop training: Converting 300% Ink Limit to 245%

hello colortheory, I need to print a dozen head & shoulders portraits on newsprint.. current files are cmyk/300ink limit/light GCR (no profiles currently attached) that printed well on coated stock..what is the best method for reusing these files for output onto newsprint.. I was planning to Mode>RGB then back to custom cmyk.. your suggestions would be appreciated. thanks for your time, jim bean

___________________________________________________________________________

Try a test of a couple of common shots.

i) (assign/assume correct profile) Old CMYK > New CMYK, Profile to Profile

ii) (assign/assume correct profile) Old CMYK > RGB > New CMYK

Option (i) will probably be the simplest and technically 'cleaner'.

As you are going good 4C > poor 4C, option (ii) may not be needed, but if I was repurposing a poor quality 4C sep into a better sep (opposite to your task) then a RGB side-trip to improve shadow density may be in order.

Stephen Marsh.

___________________________________________________________________________

You could try converting from your current CMYK profile to US Web Uncoated v2. Of course if you can get a profile from your newspaper then that is best. There is a standard in the newspaper industry called SNAP, but as far as I know there are no commercially available profiles like there are for SWOP.

Good luck,

Eric Bullock

PixelPusherz Imaging

www.pixelpusherz.net

___________________________________________________________________________

Here's what I would do:

1. Download IFRA's ISONewspaper30v4 profile from http: //www.ifra.com/ website/ifra.nsf/html/ISO_downloads_USA and install it in an appropriate location.

2. Create a Custom CMYK profile based on the settings for the existing separation and save it in the same location.

3. Open the images, and assign the Custom CMYK profile you created to them.

4. Convert the images to ISONewspaper30v4 using the Perceptual rendering intent.

5. Save the images in whatever format works best for you.

There is no reason at all to convert to RGB as an intermediary, and doing so will in fact degrade the images more than necessary due to the extra trip through Lab color. I also recommend saving the newsprint CMYK separations without an embedded color profile, just in case something further down the line doesn't agree with the ISONewspaper30v4 profile and decides to re-re-separate the images.

In my work as the photo editor of a community college newspaper, I've found that the ISONewspaper30v4 profile produces vastly superior output than any other CMYK profile for newsprint, especially the brain-dead options provided by Custom CMYK or whatever your press tells you to use.

Be warned: the ISONewspaper30v4 profile is *no good for soft proofing*. However, if you adjust your images in a color-managed environment to look exactly how you want them to on-screen *before* you convert to it using its Perceptual rendering intent, you'll likely be surprised by how good (for newsprint, at least) the printed results are.

—-

Erik Swanson

___________________________________________________________________________

hello eric and stephen, thanks for the help.. I could not convince myself to utilize the profile to profile approach, and as stated, Better to Worse printing conditions allowed me to be more comfortable with the cmyk>rgb>cmyk.. right or wrong we will see the results within a few days.. and if I made huge errors I will share the results... (additional option may have been to submit the 'ink heavy' files and hope that the newspaper's rip 'might' pull that ink down..but I trusted my files better than their rip)

everyone have a great weekend, jim bean

___________________________________________________________________________

I don't know if this is just being superstitious or not, but I never go CMYK - RGB - CMYK; if I can't (or don't want to) do a profile-to-profile conversion (which is normally the way I go, especially for ink limit and too-rich-black adjustments) I always use CMYK - Lab - CMYK.

BRIAN PYLANT

Manager, Electronic Prepress

:::::::::::::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::::

Disc Makers

7905 North Route 130, Pennsauken, NJ 08110

Toll free: 1-800-468-9353 ext. 5539

___________________________________________________________________________

I just flat ignored the obvious and likely better method.. I also believe cmyk>lab>cmyk would be superior than going through rgb... thanks, jim bean

___________________________________________________________________________

I've done it both ways. The method I use depends on what sort of tonal or color tweaks the file might need in the interim space. If your tweaks are more effective in RGB, then, by all means, use that.

___________________________________________________________________________

True. I was answering in regard to merely making the two mode conversions as opposed to doing any further editing during that process.

BRIAN PYLANT

Manager, Electronic Prepress

:::::::::::::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::::

Disc Makers

7905 North Route 130, Pennsauken, NJ 08110

Toll free: 1-800-468-9353 ext. 5539

___________________________________________________________________________

Jim,

I doubt it. When you use Convert to Profile to make a CMYK to CMYK conversion, you're risking changes in your final image due to changes in the Black channel because changes in the Black channel affect color in the image. However, conversions from CMYK to either RGB or to LAB seem to work well. If you compare the Black channels produced by the CMYK to RGB to CMYK conversion and by the CMYK to LAB to CMYK conversion, you won't see any difference between them. There may be some advantage to going the LAB route instead of the RGB route, but I've never seen anything to support this approach other than strong opinions.

Let us know how your newspaper copy turned out, would you?

Howard Smith

___________________________________________________________________________

Whoops! My earlier post was not all that clear. Instead of saying "If you compare the before-and-after Black channels when using the latter method, you won't see any difference" I should have said "If you compare the Black channels produced by the CMYK to RGB to CMYK conversion and by the CMYK to LAB to CMYK conversion, you won't see any difference between them." And I might well have added that you will see a difference between these two Black channels and one resulting from a CMYK to CMYK conversion. If you'll look closely, you should also see a difference in the image resulting from the CMYK to CMYK conversion vs. the other two images.

So much for posting without thinking.

Howard Smith

___________________________________________________________________________

On Jun 17, 2005, at 5:54 AM, colortheory@yahoogroups.com wrote:

Huh? When you do a Profile-to-Profile conversion the intermediate space (or Profile Connection Space) is LAB, so doing it as a two-step process absolutely does not make one bit of difference. The black channel gets re-generated either way, unless you use a device link profile, which would preserve the black channel.

Eric Bullock

PixelPusherz Imaging

http://www.pixelpusherz.net

___________________________________________________________________________

How did you arrive at this conclusion? I just converted a file using the two methods described and they are not the same. {US Sheetfed Coated v2 > US Web Coated (SWOP) v2 versus US Sheetfed Coated v2 > Lab > US Web Coated (SWOP) v2}

Jonathan Clymer

___________________________________________________________________________

Jonathan, I agree with Eric in general terms. My understanding is that the image itself is not transformed into LAB, even if that is the PCS used in the profile.

By going into LAB mode in Photoshop on regular bit data, one is loosing a chunk of L data and introducing other errors into the file. This may not be visible onscreen or in print, but it is happening. Which is why a working RGB space is good as an intermediate space, if one is going to do things this way instead of CMYK > CMYK. There are going to be differences in how some CMYK colours are handled if converted to RGB or LAB before going to the new CMYK which may not be there if simply going CMYK > CMYK via profile to profile.

Off hand, I only see the intermediate stop/step as necessary in two cases:

i) When thre is no access to the RGB, and one must go from poor CMYK to good CMYK and one must restore shadow density etc.

ii) If you wish to see if there is any difference when repurposing a CMYK to CMYK via convert to profile and you don't know or trust if things will be the same, so as insurance you try both ways just to see. It may not be needed, but what if the 100th time you do things without checking you get a different result due to not making the correct sacrifice to the ICC gods? I don't know enough about profile technology to guarantee that the process of transforming colour and

tonal values is always going to be the same, independent of the source

and destination profile, if one uses CMYK > CMYK or CMYK >

LAB/RGB > CMYK.

I do not see any visual difference between a Uncoated CMYK > SWOP CMYK vs Uncoated > LAB (no edits) > SWOP CMYK. If you are doing difference blend tests - then due to going to LAB you will have a more major slight difference, and even if not going to LAB using dither in the conversion may give different results when difference blending.

Stephen Marsh.

___________________________________________________________________________

Stephen, in many ways the ICC technology is superior to the old colour separation tables and system that Photoshop offered in previous years. But in other ways, despite being technically better - it can lead to inferior results in the real world compared to the old system due to users not having good profiles for the end device and being forced to use one that is less flexible than the old custom CMYK system.

An ICC profile is built for one specific condition. And for photographic content and holding shadow detail, the total ink and the neutral CMYK build for the total ink is not going to be the same as say for a large black panel in a page layout. So for an image, having less than 90% K can be good...

Speaking of the SWOP v2 profile, if memory serves correct a 0RGB value will convert to 90K and not 85K.

But this is the whole issue for more advanced users or for those who would like to make the separation suit the image, rather than forcing every image into one separation.

With only one profile to choose from, one is forced into making post separation edits to the CMYK data rather than using a separation method that delivers the required GCR or K limit or total ink limit etc. Then one must ensure that total ink limits are not blown and that the shadow is neutral and that detail is not lost and that tonal transitions do not go funky.

Another method, could be to make a dupe of the original image and convert that to custom Light GCR or UCR CMYK with a 100K limit, and then apply this K in darken mode to the K channel of the SWOP v2 CMYK conversion (perhaps pulling a lightening curve to the midtones of the custom CMYK first).

ICC technology is good but not great. It does not know what the image content is like. One can't easily shoot for the same colorimetric response but alter the black generation amount, K limit, TIL, TVI etc. Some vendors supply a range of CMYK profiles for one output intent, with different GCR, Total Ink Limits and other values. But this is clunky and the user has to choose the right profile for the right image, just as in custom CMYK one must dial in the correct values for the image at hand.

I personally think that Adobe should offer a Newsprint ICC profile, and also make different GCR, K limit and Total Ink limit versions of it's v2 profiles. It is great that Adobe have given it's users the ability to make use of ICC technology inside Photoshop, but it is a shame that ICC technology at this point is somewhat lacking in what all print users would require of it.

Stephen Marsh.

___________________________________________________________________________

Because press operators are murderous and its not good to piss them off by messing up their presses. You should only be using CMYK if you are preparing something for printing. You should use that profile if you are using coated paper on a press that has individual sheets of paper being fed into it. There are other profiles for other situations like uncoated web press....like a newspaper.

There are several reasons to avoid K=100. Inks stick to paper better than to other inks...so you want some of the paper showing through so that the inks can find it. And what with the pressures of the printing process the inks spread a bit so 85 becomes 90-95 anyway.

Now...that is not to say that you cannot do it. These profiles are sort of generic. If you know your press and what it can handle then go ahead and tweak it...but it has to be reasoned out.

Then again I haven't run a press in 20 years....maybe things have changed.

___________________________________________________________________________

Hello Stephen - It isn't uncommon for a shop to have their Spectrum (or MatchPrint, or Approval or...) dialed in to their printing conditions, so I'm not surprised you are getting the results you are getting. But this begs the question. Why aren't you doing your own proofing? It has become relatively affordable to do so, offering a lot more control to the user. Case in point, an Epson 4000 or 4800 with a ColorBurst RIP will allow you to produce a SWOP certified proof for your web jobs, and also allow you to plug in any device simulation you want for your jobs that run sheetfed. There is also a TR004/Gracol profile in there, but if your printer gives you one of their profiles (for their press or proofer) that can be used in its stead.

Its worth looking into, instead of being at the mercy of a prepress shop you aren't happy with. Here is a link to CSE's Website, the developers of the ColorBurst RIP. For around $2500 you've got a top notch, easy to use contract proofing system.

www.colorburstrip.com

Regards,

Eric Bullock

Mac Business Solutions

www.mbsdirect.com

___________________________________________________________________________

1) I don't think your problem has much to do with total ink limit. As long as you have 90%+ K in the shadows, you won't see much difference on a proof and probably zero difference on a press sheet even if your total ink ranges from 280% to 340%. I've done several press tests where the actual Dmax of a press was reached at around 100%K/280% total ink versus ANY amount of total ink where the K was <90%. 95%K @ 300-320% will look fine.

2) I believe your problem is tone reproduction/dot gain or lack thereof. If you've separated using USSheetfedCoated, that profile assumes around 24% dot gain, hence it will compensate for that gain by generally REDUCING the amount of ink anywhere on a gray scale and in colors relative to a profile that assumes, say, 20% dot gain. My guess is that this shop is linear CtP which means they are probably MUCH lower than 24% gain, probably closer to 12-16% gain and they've quite possibly linearized their Spectrum proofs to simulate this dot gain. The result is that your tone reproduction (quarter/mid/three-quarter tone total ink) is too low for their proofing system, hence the "flat" or weak look. Again, total ink coverage has really nothing to do with this.

The simple answer, if the Creo Spectrum proof is going to be your reference proof, is to have a custom profile made of the Spectrum. When that profile is created, it can be built with any total ink amount and K limit you want, that way you can use a limit of 320-340% and, say, a K limit of 97% and then you won't have to artificially raise the K limit post-conversion.

The other thing to understand about a proofing system like the Spectrum is that they will have perfect "wet ink" trapping characteristics. The result is that a Spectrum proof will show an increase in density right up to a 400% total ink limit whereas a press is likely to stop increasing in density well before you reach 400% total ink. It's not always reasonable to assume the press is going to show the same degree of separation in deep shadows compared to most digital halftone or analog proof systems. It's worth noting that GRACoL (the proposed "standard" for commercial sheetfed printing) has suggested a total ink limit of 320% for sheetfed printing, well below the 340-350% ink that you're pushing for.

Regards,

Terry

_____________________________

WyseConsul

Color Management Consulting

704.843.0858

http://www.colormanagementgroup.com

___________________________________________________________________________

My guess is that your correct! Not only are they proofing to their press but it sounds like a linear plate as well (no contrast). The final 2-weeks of the National Tour on Color Reproduction begins Tuesday in Atlanta - Thursday in Lancaster, PA - then New York & Boston. Check it out at <www.gain.net> under process control.

Dan

___________________________________________________________________________

Just out of curiosity, who have you been using in San Francisco? I'm familiar with a few of the shops downtown; I wonder if they are the same places.

I do think it's a bit odd that you increased the black in curves and are complaining about the blacks being a bit dark. That, of course, does not explain the light 1/4 tones.

I'd have them proof some scales, and then read them on a densitometer. What is happening in their proofing system? There should be some dot gain, so a 50% won't read 50%, but it should be a smooth curve.

BTW, embedding CMYK profiles is generally frowned upon.

J

___________________________________________________________________________

The place I WAS using was Cenveo. Formerly Weller Press, then Anderson Litho. Now the giant corporation Cenveo. I got suckered in with being a fan of Anderson Litho. But I guess those days are gone.

Thanks to those of you who posted your helpful comments!!

The 2nd proof (of the exact same files) looks GREAT! No changes.

The 2nd proof was made by a different vendor.

It is what I suspected all along.... vendor 1 was calibated to their specific web press conditions and could not mimic a change in press conditions on their proofer.

Which prepresses do you like in San Francisco?

Whay are you anti embedding CMYK profiles?

They are easy enough to remove if the press does not want to use the info (profile.) But are great if a conversion needs to take place downt the road. It also tells the prepress (or Press) how the file is setup.

Stephen

___________________________________________________________________________

On 6/28/05, Stephen Austin Welch wrote:

Hmm... I didn't know Anderson was bought out. That is a big company.

I wouldn't expect them to - unless they knew for sure what the press conditions were. It's kind of tough to go to a printing company for a proof that represents what another printing company would produce.

I'd probably recommend PDI or XYZ, although I don't know how much proofing-only they really do. The best thing for you would be to get a press profile from your printer, and an Epson/Colorburst and just do it yourself. You'll get closer with that setup than by taking a chance with whatever setup a prepress place happens to have.

In general, that question has been answered quite a few times on the list. Specifically for your situation, I'll answer it below.

That assumes the printing company knows what to do with it.

But not so great if a conversion is not wanted. Too many times an embedded profile prompts an unexpected conversion by an inexperienced operator. So much for all the work you did seperating the files...

Here's the interesting part - the profile in your case doesn't really describe how the file was setup. You said you embedded Adobe Sheetfed coated v2, but then made additional changes after conversion. If the changes were minor, then great, but I don't think you'd get 350 DMAX with a minor tweak. I don't know what that profile has as a maximum density, but I can't imagine it's that high.

If the profile doesn't truly represent the file, then the only way it could benefit you is by accident. I just don't like taking those kinds of chances.

J

___________________________________________________________________________

On Jun 28, 2005, at 1:01 PM, Stephen Austin Welch wrote:

This topic has been covered many, many, many.....times before on this list.

I don't look at this issue as ALWAYS do that, NEVER do this sort of thing. It, well, DEPENDS.

Being a color management idealist (sort of), I like the idea of embedding profiles and passing this bit of information on to the prepress dept. and/or printer. The color management realist in me, however, tells me to never send out a CMYK image with an embedded profile UNLESS I've had a discussion with the vendor first and I'm convinced they're going to do "the right thing" and not muck things up. The color pessimist in me tells me to leave well enough alone and never send out a CMYK image with embedded profile, period (RGB is completely different case and I think most would agree that embedding a profile in an RGB image is almost always a GOOD THING).

Bottom line, for your own internal purposes and re-puposing needs, always embed the profile. For any image/job that's going to an outside vendor, especially if they are unknown, then DON'T EMBED is the safe thing ASSUMING you've separated the image to the proper press/print conditions. If you separate to something totally foreign to the vendor you're sending the job to, just don't expect good results. It'll be up to you to first find out what they recommend (USWeb? USSheetfed? whatever?) and separate to those conditions.

Terry

_____________________________

WyseConsul

Color Management Consulting

704.843.0858

http://www.colormanagementgroup.com

http://www.wyseconsul.com (coming soon)