Pages: [1] 2   Go Down

Author Topic: How to build a 7 color profile for a RIP  (Read 14430 times)

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
How to build a 7 color profile for a RIP
« on: December 13, 2011, 07:21:13 pm »

hello ladies and gentlemen, i already asked this question in another forum but am still hoping to find help...

it seems like this is soooo rarely done.. ?? i dont know.

im lookin for someone who has actually succesfully built an ICC profile for a CMYKRGB device like my Z3200..

i love Qimage and the RGB driver of HP. no question.

the reason is on one of my papers the dark greens and blues get matte because to much ink is put down. so i want to use a RIP for that paper. i have a test chance of wasatch which is capable of limiting each ink seperately. which i did. the lin looks good. total ink is 275.

then i chose 7 channel profile in monaco, i was asked for the Lab values of my 3 RGB channels. i put in the numbers i read in the 100% patch on the lin. i asume this was right?.. then it decides i have to print a chart with 2154 patches. i do. thru the RIP with lin active. then i read the patches..
and now dont know which black generation to use.. tried the gradients of the RGB inks. profile looks OK, but some gradients (magenta to black and yellow to black) are totally bumpy and skintones are reddish..

there must be anybody who has actually done this before . ???

thanx for a professional help!
Logged

aaronchan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 619
Re: How to build a 7 color profile for a RIP
« Reply #1 on: December 13, 2011, 09:15:16 pm »

Sorry to use a question to answer a question.
But do you need to use a CMYK+3 to be your profile for your Z3200?
I have Z3200 and I have wasatch too, and I use CMYK profile only to drive my printer.

Aaron

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
Re: How to build a 7 color profile for a RIP
« Reply #2 on: December 14, 2011, 04:28:00 am »

hmm.. not sure if you read my maybe rather long question and if i understand your question.. smile

1. again: the reason is on one of my papers the dark greens and blues get matte because to much ink is put down.

2. i dont understand you. if you only use CMYK then why do you buy a printer that has 12 inks ?? if you use the contone driver of the HP to split into the 7 channels than again point 1) i cannot limit the blue and green inks more then cyan and yellow.
Logged

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4036
Re: How to build a 7 color profile for a RIP
« Reply #3 on: December 14, 2011, 05:38:35 am »

hmm.. not sure if you read my maybe rather long question and if i understand your question.. smile

1. again: the reason is on one of my papers the dark greens and blues get matte because to much ink is put down.

2. i dont understand you. if you only use CMYK then why do you buy a printer that has 12 inks ?? if you use the contone driver of the HP to split into the 7 channels than again point 1) i cannot limit the blue and green inks more then cyan and yellow.

1. You use the OEM Vivera ink? You could try to use another HP media preset or make another HP media preset with more ink limit to get what you want with the HP driver if too much ink is really the cause of your problem. Overall gloss enhancer applied should eliminate gloss difference. HP has some good documents on what ink is laid down and how to use third party papers, if not for the Z3200 at least for the Z3100.

2. Have you ever checked what the Wasatch SoftRip Z3200 profiles lay down in ink? The ones available on the site. I did for the Z3100. Put a microscope on the grey patches in a print, you will see all 6 color hues up to 60% K. Do the same on the HP driver output. If it is about driving an N-color machine Wasatch could learn a lot of HP's color management developers.

3. Creating N-Color profiles is an art, even for color gurus.

From an older message:

In contrast with that is the Wasatch SoftRip 6.2 that I can use with the Z printer models + the profiles as delivered by Wasatch for that paper and two other papers. It doesn't have a proper B&W driver that uses the 3 or 4 monochrome grey inks but defaults to black only by default then. Nor is there any chance to alter the grey inks partitioning like possible in the Quadtone RIP on Epsons. Worse is that its GCR in color mode is way too short for such a sophisticated inkset with 4 monochrome inks at hand. On a greyscale wedge printed on the same paper in color mode + the Wasatch OEM profile you will see not only CMmY in the greys from 60% downwards but even RGB ink dots. Their ink mixing algorithms and profiling is medieval compared to what is used in HP's driver. And it shows when I walk with two greyscale prints in my hand, one HP, the other Wasatch created, from one lamp to the other and outdoors. Being prone to metameric failure is one thing, it is also less consistent in color over time and it uses way more ink. When discussed with Wasatch they delivered some new profiles that were even worse. The rudiments of an RGB-device mode still available in that RIP, a leftover of the time that they couldn't make good N-Color profiles, was not recommended by them anymore. I have to agree, its hue angle/saturation choices for replacing CMY mixes with RGB inks seems broken.


met vriendelijke groeten, Ernst

330+ paper white spectral plots:

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm




Logged

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
Re: How to build a 7 color profile for a RIP
« Reply #4 on: December 14, 2011, 10:11:53 am »

ernst. thanks for your answer. its always great to see how much time you spend to help others!

1. yes. and im usually with you. i love qimage!.. its only for this one paper (lumijet nat pearl) and for my own ego to achieve the best. the glossenhancer doesnt work at all with that paper, not even with 20%. i tried less ink settings and even less. i know ALL the papers and docs that are out there (hp rules!) but even with the less settings im loosing some saturation compared to more ink. and as my problems are ONLY in the dark blue.. little in green. i figured i can do better! and i took it as a chalange. ;-) im realizing the air gets thin up here **

2. thats exactly my point. i am not sure if that is wasatch or the way i build the profile? i thought with the black generation I can decide which colors or K are put down for gray?

3. i wanna be a guru .. thats all what it is about *haha* so far i never print b&w.. so im not tooo concerned about the grays. although i of course want them neutral. but i would just like to be able to build a decent 7color profile.. with monaco or PM...

it must be possible
Logged

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4036
Re: How to build a 7 color profile for a RIP
« Reply #5 on: December 14, 2011, 10:29:06 am »

(lumijet nat pearl)

I thought Lumijet papers would no longer be manufactured/distributed by HM. Is it worth the effort? Shouldn't HM have an OMS + profile ready for the HP driver if it is still a going product?

I have removed all SpectrumViz Lumijet spectral plots but one for that reason. The White Velvet was still on stock but no longer produced and now stock is finished too. They can deliver the last if wished but you have to order a 1000kg or even more.


met vriendelijke groeten, Ernst

330+ paper white spectral plots:

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
Logged

LenR

  • Full Member
  • ***
  • Offline Offline
  • Posts: 197
Re: How to build a 7 color profile for a RIP
« Reply #6 on: December 14, 2011, 11:20:49 am »

For about 10 years I've been making 6 color profiles in PM for a Mimaki (2- six color sets of heads) with an OG set.  

I recently changed to Symphonic inks (LcLm).  Now I just go with 4C profiles.  If there's any loss in gamut it's really no big deal.

Wasatch allows you to (easily?) build a N-color inkset and as I recall (and I may be wrong) all LAB values will be saved automatically when it reads the linearization strip.

It's Separation Rules make nice 4color profiles using N-color inksets without the added cost of software upgrades.

How you build your profile (in PM. Monaco or whatever) is a black art.  Plan on spending A LOT of time and money on this learning curve.

Last year I got a 3200PS,  I don't print thru Wasatch, I use the HP Utility.  It's ridiculous how easy it is to linearize and profile media.

I love the option of using just the black inks for effortless B&W prints.  

It's so good I'm thinking about selling the Mimaki.
Logged

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
Re: How to build a 7 color profile for a RIP
« Reply #7 on: December 14, 2011, 12:06:30 pm »

smile..

it doestn help me at all.. but thanks for the input..

@ernst

well .. i dont have 1000kg.. but its worth the effort .. ::)

and as i said.. i am eager to learn !.. i just need a teacher ;-)

its not about IF HM has an oms or if something else is easier... its about building an ICC profile with CMYKRGB that works!
Logged

Scott Martin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1315
    • Onsight
Re: How to build a 7 color profile for a RIP
« Reply #8 on: December 14, 2011, 12:40:17 pm »

Not only is building this kind of profile a "black art" as LenR put it, the process of choosing a RIP for this purpose, determining optimal per ink channel limits, total ink limiting, linearization and optimizing separation parameters are also serious crafts that can take literally years to master. There are only a handful of people that are truly exceptional at this with a diverse range of experience.

It's true that the built-in RGB to 7+ color separation parameters built into many of todays drivers are quite excellent at this. Its hard to beat but it can be done especially if you're willing to prioritize certain gamut areas while sacrificing others.

All things considered, if you're not happy with the dark color saturation you're seeing I would start by doing two things:

1) Make sure your using Monaco Profiler or i1Profiler profiles with the saturation cranked up and print with the Perceptual intent. (this will show you more than the Z's profiles)
2) Compare different substrates and inksets (Epson9900,Canon8300) to see if one combination is better (there are differences, especially in the lower regions!)

In the end, I'm thinking it's easier and *more cost effective* to switch to another printer, inkset, media and profiling solution and continue using the driver than it would be do go down the "dark path" for trying to squeak more out of your 3200. And that's coming from someone who's made a living on that path for almost 20 years now.


Logged
Scott Martin
www.on-sight.com

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
Re: How to build a 7 color profile for a RIP
« Reply #9 on: December 14, 2011, 12:54:35 pm »

onsight.. thank you again as well for your input..

i thouhgt i might meet one of those handful of people who is willing to help me.. i realize.. it might be a desperate hope .. ::)

again please. i dont care about efficiancy or time... im not a person like "oh its hard? well lets take the easy way...."

so if you are experienced please lets continue your ideas..

i am pretty positive that my lin and ink restrictions are close to perfect after a lot of experience. its all the way linear. double checked. no bleeding or nothing.

its only about the profiling!!.. and not about saturation right now.. but about gradients and black generation!..

i have monaco or profile maker handy..
Logged

Scott Martin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1315
    • Onsight
Re: How to build a 7 color profile for a RIP
« Reply #10 on: December 14, 2011, 01:04:38 pm »

You want to look for chroma peaking when analyzing ink limits. Don't go for maximum density prior to bleeding. More ink is not necessarily better! At a certain sweet spot more ink causes problems and will sacrifice total ink limiting once all inks are combined. TIL of 275 suggests you might be putting down too much. Getting this just right is far more complicated than it looks... As a user of nearly all RIPs on the market, I'm not a fan of Wasatch on the Z series. I think you should consider another path.
Logged
Scott Martin
www.on-sight.com

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
Re: How to build a 7 color profile for a RIP
« Reply #11 on: December 14, 2011, 01:30:53 pm »

yes. i checked the 100% patches individually with spotmessurements compared to the 95% to make sure chroma is still changing.
how come you think 275 is too much for a pearl RAG paper?
and just because im "not talking" about this anymore doesnt mean, i havent spent days/weeks with this already ..  ;D  im just pretty sure i got it right and can move on.
of course.. i could be wrong ;-)
why dont you like wasatch with the Z?

but actually.. please.. if you are that experienced.. help me with the profiling within monaco f.e... THATs where im stuck..

when i start to build the profile, CMYK values are fixed. but for the other chanels i put in my Lab values from the 100% patches. right?
then print the chart thru the RIP. messure it..

then what about the seperations? right now i get very saturated profiles. grayaxis is sometimes good (but still has CMY in it. not only llK, lK and K) and gradients are harsh
Logged

Ellis Vener

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2151
    • http://www.ellisvener.com
Re: How to build a 7 color profile for a RIP
« Reply #12 on: December 14, 2011, 07:57:11 pm »

im lookin for someone who has actually successfully built an ICC profile for a CMYKRGB device like my Z3200

I print with a Canon iPF6300  which is also uses a CMYK+RGB ink set (R,G,B, Cyan, Photo Cyan, Magenta, Photo Magenta, Yellow, Black, Matte Black, Gray, and Photo Gray). I've been pretty  happy with the prints I've made on various papers using the custom profiles I've built using 1,800+ patch sets in Xrite's  i1 Profiler using an i1 Pro photospectrometer as measuring device. I print through Canon's IPF driver for Photoshop CS 5 as opposed to going through Photoshop and OS X's print driver. I send the files to the print driver/ printer as RGB files.

Are they technically "perfect"? I don't know, but the results are very satisfying.

My experience  does not address your specific workflow or tools of course but it does seem to be possible to build  good ICC compliant profiles for an RGBCYMK printer and a range of substrates.
Logged

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
Re: How to build a 7 color profile for a RIP
« Reply #13 on: December 15, 2011, 04:22:17 am »

thank you ellis.

but your workflow has nothing to do with my question as you are building RGB profiles and not CMYK plus
Logged

TylerB

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 446
    • my photography
Re: How to build a 7 color profile for a RIP
« Reply #14 on: December 15, 2011, 02:04:30 pm »

to your original questions... I have done a lot of CMYKOG, not 8 color, and my RIP, printer, and profiler are different than yours, so take this for what it's worth...
I don't know why a few of your ramps are lumpy, perhaps they are primarily on the gamut boundary, where smooth transitions are often not perfect. These kinds of tests sometimes reveal problems that will never be relevant to real world printing. Your skin tomes... sorry, not sure why that happened, other than the profile, and/or ink settings are not yet ideal. Often, we are finding irrational ink settings simply to satisfy quirky profiling software...
My only potentially helpful comment would be for K generation, I would find whatever max K generation setting your software provides. It will replace a lot of internal color inks with one, and your HP has several light blacks, so dottiness will not be a problem.
Tyler
Logged

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4036
Re: How to build a 7 color profile for a RIP
« Reply #15 on: December 16, 2011, 05:15:27 am »


My only potentially helpful comment would be for K generation, I would find whatever max K generation setting your software provides. It will replace a lot of internal color inks with one, and your HP has several light blacks, so dottiness will not be a problem.

Tyler

Which is the main problem with the Wasatch SoftRip if I am not mistaken. A maximum K generation should use all the grey inks and replace any CcMmY composite greys mixes not to mention any RGB ink "grey" mixes on the neutral axis. The few Z3100 profiles as delivered by Wasatch itself already show that they are not in that category. There are choices in the creation of N-color profiles where the extra hues, in this case OrangeRed-Mint Green-Violet, replace the cMmY mixes at the hue angles or at the extra condition hue angle + saturation. That kind of choices were available in the Wasatch SoftRip alternative RGB-device profiling but I do not think they are available in Monaco Profiler, at least not that transparent and adjustable for the user. Influence on the partitioning of the grey inks is not available in both Wasatch and Profiler in my (limited) experience. Then there is another obstacle; the Z3xxx inksets have no Cyan, just a Light Cyan. Part of where normally Cyan is used is done by Green and Violet. At least in the HP drivers, in a clever way, possibly adding something like a tone + hue angle condition on the shift from Light Cyan to Green. I wonder how that should be done in the Wasatch SoftRip and/or Profiler.  As I see it they are then forced to go strictly by hue angle replacement to integrate the Green and by that it will be difficult to avoid the RGB inks in the neutral spine. The profiles I have seen were worse though with little Grey inks used, more 7 channels than the 10-11 channels the HP driver uses and the RGB-device profile controls. One would expect an effort by Wasatch to deliver optimal profiles for a supported printer.

Monster wants to be educated and what I can tell does not bring him any closer to a solution if that is actually the goal instead of just education. True color experts may fill that in. Based on my experience with Z3xxx driver gloss paper printing I have no matte dark Greens and Blues but there can be some issues with the gloss; GE Economy mode bare paper white, which can be solved - bronzing which is less but not totally gone with much GE all over and suitable papers, more a B&W problem than with color (some customers like it :-). Either that Lumijet paper is simply not compatible for the Z3xxx inkset or you have a bad Green ink. I checked several prints, RC and Baryta and I do not see that matte dark Green and Blue you mentioned in your thread starter.


met vriendelijke groeten, Ernst

330+ paper white spectral plots:

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
Logged

MonsterBaby

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
Re: How to build a 7 color profile for a RIP
« Reply #16 on: December 16, 2011, 07:58:29 am »

thnks tyler. yes. i read a lot about max K. but realized it is not that easy and probably no one can actually say you have to use this and that setting, as numerous little test prints with different K settings bring totally different gradients. not only along the gray axis!

ernst.. great post! of course!.. even if it didnt help me precisly with my problem i appreciate your very found knowledge!... f.e. to know that HP creates cyan from green and violett. just those little information are what then makes things easier to understand. (when i had an epson next to my Z3100 i was frustrated so fast because there werent even information of what the paper presets do!)...

if i got everything right about hues and angles.. you cannot set anything in profile maker.. but in monaco you have a curve for every channel of RGV that i can set. for L start point of hue and curved to endpoint. plus i can use 2 slider 0-100 about hue and saturation range for the spot channel. what is your recomandation for those?

and now to the K and wasatch. THAT could bring the point!.. i thought max K means only use K thus llk, lk, k.. but.. is this a problem about wasatch or my profile?.. im not sure.. i was talking with other press experts that it is impossible to use a CMYK profile and get a RIP to only use K or neutrals in the gray axis. and the RIP will ALWAYS use CMYK for neutrals because of the way its built?..
Logged

TylerB

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 446
    • my photography
Re: How to build a 7 color profile for a RIP
« Reply #17 on: December 16, 2011, 12:48:26 pm »

Which is the main problem with the Wasatch SoftRip if I am not mistaken. A maximum K generation should use all the grey inks and replace any CcMmY composite greys mixes not to mention any RGB ink "grey" mixes on the neutral axis. The few Z3100 profiles as delivered by Wasatch itself already show that they are not in that category. There are choices in the creation of N-color profiles where the extra hues, in this case OrangeRed-Mint Green-Violet, replace the cMmY mixes at the hue angles or at the extra condition hue angle + saturation. That kind of choices were available in the Wasatch SoftRip alternative RGB-device profiling but I do not think they are available in Monaco Profiler, at least not that transparent and adjustable for the user. Influence on the partitioning of the grey inks is not available in both Wasatch and Profiler in my (limited) experience. Then there is another obstacle; the Z3xxx inksets have no Cyan, just a Light Cyan. Part of where normally Cyan is used is done by Green and Violet. At least in the HP drivers, in a clever way, possibly adding something like a tone + hue angle condition on the shift from Light Cyan to Green. I wonder how that should be done in the Wasatch SoftRip and/or Profiler.  As I see it they are then forced to go strictly by hue angle replacement to integrate the Green and by that it will be difficult to avoid the RGB inks in the neutral spine. The profiles I have seen were worse though with little Grey inks used, more 7 channels than the 10-11 channels the HP driver uses and the RGB-device profile controls. One would expect an effort by Wasatch to deliver optimal profiles for a supported printer....

Well this RIP functions in ways mysterious to me then,, and I should not muddy the waters with suggestions. My Monaco Profiler was not the Platinum version, so additional N color ink controls, I can  not see. i1Profiler, which I now use for 6 color, does provide controls over how the additional inks are used. My other profiler, Ergosoft GPS, give no user control of how additional colors are introduced, but frankly whatever it does by default works better than i1 with all it's control at this point. 9900 O and G are only introduced as the CMYK gamut runs out, and there is no O or G in the neutral build at all, or in interanl colors easily handled by CMYK. Not only does this keep total ink down, it provides higher resolution since green ink dots are more visible than yellow and light cyan combined to make the same color... the builds describe above seem overly complex to me...
Profilers would not offer controls over light black ink partitioning, they see K as a single color, so all that control is in the RIP, if available. I have no knowledge of Wasatch controls, Ergosoft does let me control the black inks, partitioning, dot size, percents, etc...

thnks tyler. yes. i read a lot about max K. but realized it is not that easy and probably no one can actually say you have to use this and that setting, as numerous little test prints with different K settings bring totally different gradients. not only along the gray axis!
...

and now to the K and wasatch. THAT could bring the point!.. i thought max K means only use K thus llk, lk, k.. but.. is this a problem about wasatch or my profile?.. im not sure.. i was talking with other press experts that it is impossible to use a CMYK profile and get a RIP to only use K or neutrals in the gray axis. and the RIP will ALWAYS use CMYK for neutrals because of the way its built?..


Well, hmm.. I never found a similar setting in Monaco (had to try to draw as straight a line as I could in the dumb chart), but there is a total K GCR setting in i1. They are wrong and right... there is generally some color ink in the neutral build regardless of setting simply because no K ink/paper combo ( including that channel's total build with the lighter Ks) is LAB neutral. So, color must be introduced to make nuetral. The amount, and the amount of K used there as well, depends on those GCR settings.
SOmehow I have the feeling none of this is moving you along though...
Tyler
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 21646
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: How to build a 7 color profile for a RIP
« Reply #18 on: December 16, 2011, 01:07:17 pm »

it seems like this is soooo rarely done.. ?? i dont know.

What are you hoping to achieve doing this, rather than using the RGB path of the native driver? And are you willing to invest in some 3rd party driver or RIP that allows control over this process?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Scott Martin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1315
    • Onsight
Re: How to build a 7 color profile for a RIP
« Reply #19 on: December 16, 2011, 01:17:23 pm »

the reason is on one of my papers the dark greens and blues get matte because to much ink is put down

Goto http://www.on-sight.com/2010/04/25/canon-x300-printer-review/ and scroll down the photo near the bottom called "gloss differential and bronzing comparison." Does that not show the matte like surface qualities you're seeing on your prints? If so, this is an inherent limitation to that inkset. I really think you need to compare other inkset and paper combinations or explore liquid varnishing.

Here's the direct link to my differential comparison image:
http://www.on-sight.com/images/3BrandPrinterComparisonGlossDiffBronzing_full.jpg
Logged
Scott Martin
www.on-sight.com
Pages: [1] 2   Go Up