Forgive my ignorance - is it this simple ... yet this complicated?

[ScotJ]

I'm coming from a digital large format background, and we moved into a QM DI about 6 months ago.

I self taught the basics of desired dot gains, required densities, and the perils of reaching these targets.

We are in the process of installing a 6 colour akiyama - and I'm curious if the colour configuration is that "simple." I know the targe densities vary from press to press, but beyond reaching desired dot gain curves and appropriate densities - is there much more to it?

We plan to bring in an experience colour management consultant to help with this, but I'd like to have a knowlege of the objoectives we're trying to reach none the less.

Any info would be appreciated!

 

[glenncolorguy]

Simple? or Complicated?

Simple? or Complicated?

The concepts involved in process control for offset are reasonably simple.....application is not.

Density and dot gain may rough you into the territory...if you have accurate proofs to begin with.....but there is much more to the process.

Density needs to be based on L*a*b* values, and will vary with different ink types. MY, CM, CY overprints are as important as primaries and must be taken into consideration. Graybalance-the most important component-is to some degree predicted by density+dot gain but not entirely, and this must also be under control. In addition, determining press conditions is far more difficult than often thought because press variability is a major factor. Was your test calibration run really representative of typical press output (Almost certainly not).

But far beyond all technical considerations are the human factors. How will you implement process control in the pressroom? How will you achieve needed cooperation between prep and press? How will you collect data to assure continuing compliance with objectives? What procedures do you have in place to adapr to changing conditions on press? What will your team do when running to the numbers seems not to work?

Feel free to contact me directly on any of these issues.

 

[gordo]

Density needs to be based on L*a*b* values, and will vary with different ink types.

Perhaps worthy of a separate discussion.
I can't quite agree with that statement.
Density is an indirect measure of ink film thickness.
L*a*b* values are a measure of color.
These are two separate but related factors that need to be dealt with.
The function of the press is to lay down a film of ink. It does not know or care what color that film of ink is.
So, IMHO, density needs to be correct in order to avoid issues like ink tailing/slinging, and over emulsification, etc. At the correct ink density (ink film thickness) the color (L*a*b* values) should be within the target for the process (e.g. ISO 12647-2)
Put another way, you need to hit the correct density at which point your L*a*b* values should fall into place. If not you may have to select a different primary ink hue set/series.

best, gordo

 

[David Milisock]

yes it is that simple

yes it is that simple

A new press will require that a media linierization, ink limit dot gain curve be created this will create plates that have a preset curve for your paper. Many companies only use one some use two, one for coated paper and one for uncoated paper.

These curves will generate a TIC total ink coverage for each media, simply use ICC profiles for your applications that are equal to or less than the TIC for the media press.

Once this is done DO NOT change inks, use the same inks used for the creation of the curve.

 

[Bloodsaler]

As I know,DI press is hard to calibrate by human,it`s almost autne client of my boyfriend have brought this press,but the sample often reddish,even the heidelberg engineer can`t solve this problem~

 

[David Milisock]

Heidelberg have people and then they have people, Kodak has the same issue.

 

[gordo]

A new press will require that a media linierization, ink limit dot gain curve be created this will create plates that have a preset curve for your paper. Many companies only use one some use two, one for coated paper and one for uncoated paper.

These curves will generate a TIC total ink coverage for each media, simply use ICC profiles for your applications that are equal to or less than the TIC for the media press.

Once this is done DO NOT change inks, use the same inks used for the creation of the curve.

I find this very confusing.

What do you mean by "media linierization"?

What do you mean by an "ink limit dot gain curve"?

TIC is a function of how an RGB to CMYK conversion is made. Not curves (plate or press).

thx, gordon p

 

[fiatlux]

Heidelberg have people and then they have people, Kodak has the same issue.

Those other people you refer to...are they men without souls....zombies?

FL

 

[David Milisock]

Those other people you refer to...are they men without souls....zombies?

FL

I only know this I was buying printing from a company and for some unknown reason their color would just go bonkers, New Heidelberg install with service contract so all the configuration was part of the contract. Well afte about 3month they ask if I'd set in and observe the process. I suggested after my observations that theyrequest another technician, the second guy didn't dowell eithr but the third guy got it done.

Kodak never could get the Nextpress working so I finall calibrated it for them. I did several of these units they are unique.

 

[David Milisock]

I find this very confusing.

What do you mean by "media linierization"?

What do you mean by an "ink limit dot gain curve"?

TIC is a function of how an RGB to CMYK conversion is made. Not curves (plate or press).

thx, gordon p

People some times have a hard time understand that PS and ICC color management are much the same thing. Just like an ink jet media a press produces a different curve for each paper. So the plate is linerazed - printed to a media (paper) result is non linearized and read by densitometer/spectrophotometer- curve adjustments are made to plate setter - new curve printed - result read again until paper produces liner result with test.

During this process an ink limit is defined for each channel, it may be 86% DMAX and 2% MIN. 86 x 4 is 344 TIC When talking PS color management curves is a better term then profile this result will produce the color rendering dictionaries for the postscript interpretation process.

TIC is simply a term total ink limit, be it RGB to CMYK or CMYK to CMYK which is what PS color management was designed to work with ONLY!!

Is this stuff really a curve?

No, there is a LAB color model equivalent of the CMYK TIC 400 color model, PS CM systems assume this as the source space and for press purposes CMYK can be considered to be an absolute color space for the source.

The destination space is the space you created by the plate lineraization, ink limit process the TIC and curves also have through the software process a LAB color madel equivalent.

The plate making process simply connects the assumed color space with the descination color space via LAB.

Assume this, any ICC compliant application Adobe/Corel whatever, with a working CMYK of US web uncoated, anyone worth their salt knows this is a TIC 260 profile.

The designer uses a green C100 Y100 M5 K0, imports a CMYK image converted to the propper CMYK profile and produces a PDF for output, (if done correctly it will an uncolor managed PDF). As we read the green color all through the process it will read C100 Y100 M5 K0 and we all know the paper is not going to hold that ink however spot reading of the image show that its CMYK numbers are being passed along also.

Nothing in the ICC controlled process before the plate settter allows for the conversion to the plate destination color space for the non-images objects in the file.

The file is opened in lets say Super Trap and then trapped, then passed along to Signastation and imposed, still at all these points the CMYK numbers read just as they did in the native application.

From Signastation it is RIPPed VIA PS to the destination plate, asssumed source space to created destination space.

The display of the native application always tries to simulate what C100 Y100 M5 K0 looks like but no real change ever took place in the application to the green object, the image was already changed so the display (if calibrated) may be fine.

So the imaged passed through to the plate unchanged as the CMYK numbers were already in small destination space, the green objects was changed because most likely the paper was only going to hold less then the 100 for the C and Y channel. However the change was expected by the designer because their display had a simulation of a small destination color space.

I've never seen any plate for any CMYK job created in any other manner.

 

[gordo]

So the plate is linerazed - printed to a media (paper) result is non linearized and read by densitometer/spectrophotometer- curve adjustments are made to plate setter - new curve printed - result read again until paper produces liner result with test.

So you start the process with a linear plate (50% in the file is 50% on the plate). And your target is to end up with 50% in the file resulting in 50% on the press sheet?

thx, gordo

 

[David Milisock]

Yes as close as possible across the entire curve from 1% to 100%, the different applications have different levels of control. I would like to say you get what you pay for but not so, some less expensive units do as well or better than systems costing $30,000 more.

Of course this is where the ink limit comes in and the generation of the CRD's from the process. PS colormanagement when working with CMYK to CMYK only works as well if not better than any device link profile I've ever seen.

I've never seen it work any other way with any plate setter.

Now the cool part of thisis after all this work company A follows the rules and gets great repeatable color fom their units. Company B has a press room manager that decides to try a new ink set or blanket company the next month.

 

[meddington]

David, forgive me, but after reading and re-reading your posts, I can't make heads or tails of your actual methodology. I'm coming away from this with the perception that you're "linearizing" the press sheet. (50% from the file measures 50% on the printed piece), then compensating for this with CRDs, is that right? And after all this, what does the 50% from the file actually read on the press sheet? I don't understand why you would want to "linearize" the printed sheet.

 

[buckeye]

Put another way, you need to hit the correct density at which point your L*a*b* values should fall into place.

Gordo,

I thought it was explained to me by our G7 Expert that the desired L*a*b* values on the printed sheet gave you your standard ink densities. I very well could have misunderstood him or maybe I'm misunderstanding what you're saying.

Erik

 

[David Milisock]

David, forgive me, but after reading and re-reading your posts, I can't make heads or tails of your actual methodology. I'm coming away from this with the perception that you're "linearizing" the press sheet. (50% from the file measures 50% on the printed piece), then compensating for this with CRDs, is that right? And after all this, what does the 50% from the file actually read on the press sheet? I don't understand why you would want to "linearize" the printed sheet.

Actually all I do is follow the Heidelberg META directions in their manual. Linearize plate material so the plate produces the test patten correctly, run proper test patter en on press to proper ink densities, read test sheet enter values in the RIP/Plate setter controls, The software creates a set in which you print to for making a plate. It's time consuming but real simple. This creates ink limits, linearizes and controls dot gain/reduction. The method is very similar for Rampage and I was just at a shop this morning with a Harlequin setup and they were doing their system and it seems very similar. After the process any postscript file with CMYK color data that fits within the limits passes through unchanged.

Now the Rampage system I worked on would not allow the use of ICC profiles. This process creates the CRD's for the destination color space for postscript.

Now you'll get some technicians that will tweak curves especially with systems like META that have them to tweak but if you have patience and do it right on a well maintained press IMO no tweaking ever helps.

I prefer to leave the choice of a CMYK color profile to the designer within limitations of course. For quality sheet fed work I like a Light UCR GCR CMYK 360 proofing by Kodak. For standard commercial work I like a similar TIC 300 by Kodak. I have used a gracol TIC 300 once and liked it too. I use uncoated and uncoated web variations of the Kodak for uncoated work and news print.

 

[gordo]

Gordo,

I thought it was explained to me by our G7 Expert that the desired L*a*b* values on the printed sheet gave you your standard ink densities. I very well could have misunderstood him or maybe I'm misunderstanding what you're saying.

Erik

No, the desired L*a*b* values on the printed sheet do not give you your standard ink densities.

CIEL*a*b* values define a color.

Ink density is an indirect measurement of ink film thickness - neither of which tell you what the color is.

best, gordo

 

[meddington]

There's latitude with ink film thickness (density) as there is with CIELab values. You might find that you window for operation with a given inkset is quite wide, so adjusting densities up or down to hit desired CIELab values is often quite doable. Part of optimizing your process might include printing a destiny sweep, where you print from low to high ink film thickness, measuring the wet ink densities, then later measure the dry CIELab values to determine the ideal target at press, keeping in mind that lowest delta E isn't always the best target, and that secondary colors might take precedent. I'd also note that delta E is not ideal for process control tolerances, as it can be a wide window. Give a pressman a density values to shoot for with tolerances based on what the process is capable of holding (six sigma approach).

My 2 cents...

 

[TerryWyse]

Now the Rampage system I worked on would not allow the use of ICC profiles. This process creates the CRD's for the destination color space for postscript.

Not true David....RAMpage has options for applying "in-RIP" ICC color management as well as "post-RIP" color management via RAMproof Direct+ICC or INKdrop+ICC. RAMproof Direct (RPD) is typically used with proofing devices while INKdrop is typically used just prior to imaging plates for the press. With RPD and INKdrop, you can use a standard ICC transform (source/destination) or ICC device link profiles. Furthermore, if you have the ICC option installed, I believe many, if not all, of the various RAMproof options will support insertion of ICC profiles.

I prefer to leave the choice of a CMYK color profile to the designer within limitations of course. For quality sheet fed work I like a Light UCR GCR CMYK 360 proofing by Kodak. For standard commercial work I like a similar TIC 300 by Kodak. I have used a gracol TIC 300 once and liked it too. I use uncoated and uncoated web variations of the Kodak for uncoated work and news print.

Wouldn't it be better to use standard profiles (GRACoL_Coated1, SWOP_Coated3, ISO variations...) designated to be used by their respective standards organizations instead of custom profiles conjured up by Kodak? Personally I wouldn't be trusting custom profiles unless I knew exactly what data set was used to build that profile.

Just sayin'

Terry

 

[David Milisock]

Not true David....RAMpage has options for applying "in-RIP" ICC color management as well as "post-RIP" color management via RAMproof Direct+ICC or INKdrop+ICC. RAMproof Direct (RPD) is typically used with proofing devices while INKdrop is typically used just prior to imaging plates for the press. With RPD and INKdrop, you can use a standard ICC transform (source/destination) or ICC device link profiles. Furthermore, if you have the ICC option installed, I believe many, if not all, of the various RAMproof options will support insertion of ICC profiles.



Wouldn't it be better to use standard profiles (GRACoL_Coated1, SWOP_Coated3, ISO variations...) designated to be used by their respective standards organizations instead of custom profiles conjured up by Kodak? Personally I wouldn't be trusting custom profiles unless I knew exactly what data set was used to build that profile.

Just sayin'

Terry

Not sure how old the version of Rampage I was on, seemed low end but I just follow the books.

>Wouldn't it be better to use standard profiles (GRACoL_Coated1, SWOP_Coated3, ISO variations...) designated to be used by their respective standards organizations instead of custom profiles conjured up by Kodak?

Why? If I open the profile up in ProfileMaker Pro and like how it transends the gray balance and it's conversion why not use it? Most presses run by smart operators are Postscript color managed anyway so all I need do is select a low enough TIC and I'm good to go. I've done nearly 11,000 jobs with these Kodak profiles I hope I don't have to give the money back.

I've taken Press ready PDF files created with Quark 7 MAC jobs opened them in PC Acrobat with Pitstop, converted fonts to outlines, taken those files to CorelDRAW X5 for the user and had the color elements in those file match from Corel Pblished ODF files other jobs printed 2,000 miles away and at a second printer 400 miles away so I'm fairly confident about my color theory.

>Personally I wouldn't be trusting custom profiles unless I knew exactly what data set was used to build that profile.

Isn't that what you're asking people to do when you build a profile for their press? Quite frankly do you think the clients understand the data sets?

 

[TerryWyse]

Not sure how old the version of Rampage I was on, seemed low end but I just follow the books.

RAMpage ICC has been around a while, at least since v9....but it *is* an option a customer has to purchase...so maybe your customer didn't have the ICC option. Just sayin' that making a blanket statement that implies RAMpage doesn't support ICC color managment is dangerous unless you have all the facts.

Most presses run by smart operators are Postscript color managed anyway so all I need do is select a low enough TIC and I'm good to go. I've done nearly 11,000 jobs with these Kodak profiles I hope I don't have to give the money back.

I'm getting the nagging feeling that YOU may be talking about digital presses while I'M assuming we're talking about web and sheetfed offset litho presses...is that maybe the confusion here? In any case, my experience with color-managed workflows differs significantly from your experience.

I've taken Press ready PDF files created with Quark 7 MAC jobs opened them in PC Acrobat with Pitstop, converted fonts to outlines, taken those files to CorelDRAW X5 for the user and had the color elements in those file match from Corel Pblished ODF files other jobs printed 2,000 miles away and at a second printer 400 miles away so I'm fairly confident about my color theory.

Speaking of "Acrobat", that sounds like a lot of hoops to jump through just to get a job into a workflow....and you use CorelDRAW as part of your normal workflow recommendations? Interesting.

>Personally I wouldn't be trusting custom profiles unless I knew exactly what data set was used to build that profile.

Isn't that what you're asking people to do when you build a profile for their press? Quite frankly do you think the clients understand the data sets?

I think we're talking two different things again. What I'm saying is that as far as a SOURCE profile (or a "target" profile that I want to match if you prefer), I wouldn't generally recommend the use of a non-standard or custom profile. Of course the press (destination) profile would be custom....but I'm not necessarily advocating profiling an offset press unless the customer is looking to "color-manage" the press data for ink savings or they have a specific (non-standard) printing condition such as Hexachrome or wide-gamut CMYK that needs to be custom-profiled for use in a proofer or whatever.....otherwise, even for proofing, I'm generally recommending *against* custom press profiles and sticking with industry-standard profiles...and then *calibrating* the press to meet that specification. That's one of the tenets of the G7 method is 1) use of a *standard* specification/profile for proofing and 2) calibrate but don't necessarily color-manage the press to hit that specification. IOW, have the press match the (standard) proof rather than the other way 'round.

Terry