Profiles

[5624GM]

New to Color Management and our files man uses a default profile(US Web Coated(SWOP)v2) This seems to be working fine. Can someone breifly explaine Profiles and why would you make your own?

 

[JoshB]

So you've got data coming in (hopefully) in CMYK, and there's a patch of color that says 30%-Cyan. OK, great! Now what does that mean exactly? 30% of the cyan ink you can lay down on uncoated sheet on your old offset press at 150lpi with an elliptical dot? 30% of the cyan toner that your Xerox can lay down on coated stock at 200lpi? 30% cyan is very nearly meaningless.

For your purposes, think of a profile as describing a device. The SWOP profile that you're using is for a theoretical coated web offset press, and if your press behaves kind of like that theoretical machine you'll probably be close enough for most customers without doing much beyond keeping your press calibrated.

But what if you're not using a standard web offset press? What if you're sending a small reprint to your Xerox? It should look about the same, right? Thankfully, you've got a different profile for your Xerox, that describes how that machine behaves. The controller for the Xerox sees that your incoming file was setup for SWOP, and it translates the color values using the two profiles so that your output should match.

So, when would you want to use a different profile on the file? Well, maybe you're not usually printing on a standard web offset press on coated paper. In that case either:

A -- You have no further color management, and your colors are probably fairly far off.

B -- You have a color managed rip that takes the incoming file and translates the colors for your device. In which case you've probably limited the colors more than necessary for your output device.

That Xerox, for instance, can almost certainly print a much greater range of colors than a SWOP device, so if you're prepping all your files for SWOP, then you've unnecessarily restricted your available colors.
-----------------------------------

Why would you ever make your own profile? If you're not doing lowest common denominator printing and your customers are at all picky, then you probably need to be able to exactly match a proof to your final product. In that case you want a profile for your specific press, on your paper and inks, with your RIP settings. Get good profiles for your proofer and for your press, and you should be able to confidently hand out a proof and say "Yes. It will look like this."

 

[David Milisock]

roll your own profiles

roll your own profiles

Let me ask you this, what output device are you using?

 

[5624GM]

We are an offsett and wide format printer using nexus,Lotem platesetter,Epson 9900 proofer.We are G7,I was just questioning (US Web Coated SWOP v2) And Thanks JoshB

 

[David Milisock]

reply

reply

I've been reading this forum for a day or so now and it appears that many do not understand the process of setting up the plate setter and press verses the use of ICC profiles.

Some call it footprintng the press I call it linearization/dot gain/ink limiting but what ever you call it the process in many cases is propriatary for the plate setter, ink set, blanket and ambient conditions at the time of the process.

The plate setter manufacturer has a process to set the RIPPed file to the plate in a manner to compensate for dot gain, this goes in conjunction with the ink film requirement of your ink manufacturer. They know what density their inks need to run. Onces this process is done the press has had an established total ink coverage for the paper used at the time, a designer can then use any CMYK ICC profile that has a TIC that is equal to or less than the TIC for the press.

Inother words sen CMYK only and use the ICC profile that gives you the required TIC and the transition through the curve you like. For sheetfed quality work I use a Kodak TIC 360 for web work I use a Kodak TIC 300, there's a GRACOL or two I like also.

The only ICC profile you really need to make are for your proofing devices

 

[David Milisock]

The only ICC profile you really need to make are those for your proofing device

 

[5624GM]

David, thanks alot. Please forgive me for not totally understading, I just won't to make sure I know this. I work on the proofing,plating side and we are happy with hitting proofs. I am weak on the front end and won't to make sure our customers are getting all the color they want. When a file first comes in they are using a SWOP profile on images that are not assigned one. I know this controls a color gamut. So do we need to look at everyone using another one.

 

[David Milisock]

Missing profiles

Missing profiles

If your clients are not tagging CMYK images then they are making you guess, if you're from the Westen Hemi-sphere using the US Web coated CMYK is a good guess as this is the default install for Adobe and CorelDRAW.

Depending on where proofing is in your process it may be irrelevant, if you're proofing the RIPPED file the system most likely uses postscript color management then an ICC profile for the proofer.

If you're editing these images you have to guess when you open them and then you can tag them after you edit them solving the guessing issue.

The real issue is only Adobe Photoshop or CorelDRAW/ Photo-PAINT will display them in a manner that you can have confidence in the image.

Soft proofing untagged images can be interesting. Do you have calibrated displaya and a controlled work environment.

 

[Correct Color]

Here's what I'd consider a much more concise answer:

Every device that reproduces color reproduces it to some degree differently. A profile is a description of how a particular device reproduces color in a certain calibrated state.

And the two things to keep in mind about any profile are that whatever the limits of a particular profile, they are the limits you apply to the device you're using it to drive every time you use it; and, the closer the profile you use matches how the device you're driving it with actually prints, the more you'll get predictable results.

You said you were large-format and offset.

At present, those two arenas are traveling in somewhat divergent paths regarding color in general. In offset, there's a tendency to define process ink values to standards for chroma and density. Then if a press can be linearized--by whatever means, such as G7--theoretically there's no need for a specific profile for any press. One size fits all...

Yay! Printing as a commodity.

In large format, the straitjacket is still at least in the other room. It's pretty much imperative if you want to get full capability out of a large-format workflow that you profile each printer on each media in your specific environment.

And if you want to learn how to do it correctly, it'd be very beneficial to hire someone to come in and set this all up for you.


Mike Adams
Correct Color

 

[David Milisock]

Here's what I'd consider a much more concise answer:

Every device that reproduces color reproduces it to some degree differently. A profile is a description of how a particular device reproduces color in a certain calibrated state.

And the two things to keep in mind about any profile are that whatever the limits of a particular profile, they are the limits you apply to the device you're using it to drive every time you use it; and, the closer the profile you use matches how the device you're driving it with actually prints, the more you'll get predictable results.

You said you were large-format and offset.

At present, those two arenas are traveling in somewhat divergent paths regarding color in general. In offset, there's a tendency to define process ink values to standards for chroma and density. Then if a press can be linearized--by whatever means, such as G7--theoretically there's no need for a specific profile for any press. One size fits all...

Yay! Printing as a commodity.

In large format, the straitjacket is still at least in the other room. It's pretty much imperative if you want to get full capability out of a large-format workflow that you profile each printer on each media in your specific environment.

And if you want to learn how to do it correctly, it'd be very beneficial to hire someone to come in and set this all up for you.


Mike Adams
Correct Color

I would agree with large format devices, however 90% + of all sign companies are using applications and devices with canned profiles. Not for me but their market is ok with the end result.

It seems that the bigest problem this poster has is the untagged CMYK images and how to handle them. His people have a reasonable approach and it seems to be working.

I also agree that if a press can be linearized, however each and every press needs to be linearized and I would suggest two processes one for the house coated and one for the house uncoated stock. I wouldn't call it printing as a commodity I wouls call it practical linearization. At the cost of a days production making a linearization curve for each and every media is IMO BS.

Now to this gentlemans issue once the presses are set up they can actually use any CMYK ICC profile that has a TIC equal to or (and I stress slightly less then the linearization curve of the press. There must be a hundred or so commercial CMYK profiles out there, pick one that fits the TIC parameter and that has a gray balacne curve you like and have at it. I print stuff all over the world using Kodak ICC profiles, I'm not big on the ECI CMYK profile they look too saturated for me, the gracol are not bad but in reality just a rehash of the old Kodak work.

I will say this the Kodak and gracol are more difficult to actually print but IMO give a better end result due the the increased use of CMY for gray values near and right under ther mid tones.

 

[Correct Color]

It seems that the bigest problem this poster has is the untagged CMYK images and how to handle them.

Oh?

New to Color Management and our files man uses a default profile(US Web Coated(SWOP)v2) This seems to be working fine. Can someone breifly explaine Profiles and why would you make your own?

Maybe it's just me, but I'd say he wanted a brief explanation of profiles and a reason for making custom profiles.

 

[David Milisock]

Oh?



Maybe it's just me, but I'd say he wanted a brief explanation of profiles and a reason for making custom profiles.

His third post pointed to the fact of non-tagged CMYK images as being the big issue. Judging from his shop layout he says he's hapy with the proof matching the press so no profiles needed there. Unless he's doing posters and art prints he dosen't need anything.

 

[gordo]

I also agree that if a press can be linearized, however each and every press needs to be linearized and I would suggest two processes one for the house coated and one for the house uncoated stock. I wouldn't call it printing as a commodity I wouls call it practical linearization. At the cost of a days production making a linearization curve for each and every media is IMO BS.

Now to this gentlemans issue once the presses are set up they can actually use any CMYK ICC profile that has a TIC equal to or (and I stress slightly less then the linearization curve of the press.

Could you explain/elaborate on what you mean by "linearization" in your response?

best, gordo

 

[David Milisock]

linearization

linearization

Well the process can be different with different hardware/software combinations but it's similar. In the end you produce what Heidelberg and Rampage call a profile, other processs call them something else (sets) comes to mind but the point to this part of the conversation is if they are called profiles thay ARE NOT ICC profiles. They are a set of commands that produce a curve. The best writing I've seen on the subject was Heidelbergs white paper on the META work flow which as most presses are is postscript color managed. It completely explaned the use of color space arrays/color rendering dictionaries and mentioned that the RIP had the capability to use ICC profiles if the RIP server contained all the source and destination profiles in a specific folder. However Heidelbergs detailed instructions for this product are for postscript color management and they recommended it be set up that way.

Ok the process starts with choosing an ink set, paper, blankets, packing materials and making sure the press is in good working order, I've gone on site and spent three days with a press repairman before anything could start.

On the plate setter you run a test pattern, it differs with different manufacturers but the concept is to get the plate setter running screens of the propper percentages, 5%, to 100%
some system have a 3% they start with. Once the setter is linearized (producing the propper screen percentages) the plates are taken to the press and printed to the propper ink density based on the ink manufactureres specifications. The test pattern is read and adjustments are made inthe plate setters command software to adjust for dot changes. A set of plates are made with these changes and the test run again and read.

This creates an ink limits, linearization dot gain curve for maximum gamut from the press with the specific media and ink set. The only critical aspect of that for a designer it the TIC (total ink coverage) for that media. As postscript color management will be used simply convert your RGB to a CMYKprofile with a TIC qual to or slightl less then the presses TIC and send the file.

It's similar to what I do on my Onyx production house RIP when I create a new media and like Onyx the manufacturer of the specific devices controls all of it with their propriatary hardware/software.

It's time concuming especially with the software issues from different manufacturers but not that tough. The tough part is working with a company that has not information on their equipment.

 

[gordo]

Ok the process starts with choosing an ink set, paper, blankets, packing materials and making sure the press is in good working order, I've gone on site and spent three days with a press repairman before anything could start.

On the plate setter you run a test pattern, it differs with different manufacturers but the concept is to get the plate setter running screens of the propper percentages, 5%, to 100%
some system have a 3% they start with. Once the setter is linearized (producing the propper screen percentages) the plates are taken to the press and printed to the propper ink density based on the ink manufactureres specifications. The test pattern is read and adjustments are made inthe plate setters command software to adjust for dot changes. A set of plates are made with these changes and the test run again and read.

This creates an ink limits, linearization dot gain curve for maximum gamut from the press with the specific media and ink set.

OK, I think I understand what you mean.

In a CtP workflow, the process of linearizing the plate in order to run the plate on press to get the data to build a dot gain compensation curve is redundant. It is not needed. All that's actually needed is a "target" curve - i.e. the tone reproduction that you want and a "current" curve - i.e. how the calibrated (may not be linear) plate prints. With a target curve and a current curve you can build a press curve to apply to plate in order to achieve the tone reproduction you want.

In your last sentence you write "linearization dot gain curve". That confuses me since a linearization curve is used to linearize output (either plate or presswork). A dot gain curve is not necessarily a linear curve (actually it's unlikely to be linear). So, a "linearization dot gain curve" seems to be a contradiction in terms.

best, gordon p

 

[David Milisock]

linearization dot gain

linearization dot gain

Yes it's a two step process, the plate setter needs lineraized, then the test is printed and press sheets read for press dot gain or dot reduction. The plate setter is then adjusted so the (for example) if the setter placed a 50% dot and the press printed 52% the curve reflects that so that the plate curve is moved, so that when the ink density is hit the new dot produces 50%. In reality it processes 4 functions, ink limit, dot gain/reduction, plate setter setup linearization media lineaization.

Once this is done you run your quality sample sheet and adjust the density to achieve the graybalance.

There is no other way to do this as the setter manufacture requires a ( set or a command profile) to print to so plate can be made. The real issue is I see so many with just one profile command. How can that work for coated and uncoated? It can't but people are what they are!

After this is done any CMYK profile with the previously mentioned TIC limitation just can drop in as postscript color managed.

 

[gordo]

Yes it's a two step process, the plate setter needs lineraized, then the test is printed and press sheets read for press dot gain or dot reduction. The plate setter is then adjusted so the (for example) if the setter placed a 50% dot and the press printed 52% the curve reflects that so that the plate curve is moved, so that when the ink density is hit the new dot produces 50%.

I've asked this in another thread. Is your idea to have a 50% tone in the original file print as a 50% tone on the press sheet.

thx, gordo

 

[David Milisock]

I've asked this in another thread. Is your idea to have a 50% tone in the original file print as a 50% tone on the press sheet.

thx, gordo

The concept is to linearize the start media (the plate) print that plate, read the print adjust the plate output to correct for the press.

This through the plate setter RIP software creates the LAB equivalent of the Presses Color Gamut (I.E. the color rendering dictionaries) for the connection of the assumed CMYK of TIC 400 to the destination press TIC of what ever.

It's basic postscript color management and worked long before ICC profiles were in the mainstream.

NEVER, NEVER, NEVER for RGB to CMYK conversions