Setting up reference printer for ORIS CGS

[MonsterBabyBLN]

hi everyone,
i would like to use my own media with my Z3200 and oris... therefore i tried to set up a reference printer numerous times.

i print a lin.. measure it.. set the max density a little less.. do another lin.. save it.. then print an it87.4 chart and build a beautiful reference printer.
a few moments later i start to set up a hotfolder.. using the just made reference printer .. use the same lin i just set up.. and print the chart (ECI2002?? i dont get offered anything else).. i measure it and boom.. avgt dE1.1 ok.. but yellow.. is max dE4.8!!! the same lin?? 5 mins later?

i do the first iteration and get avg 1.7 and max 6.7 ??? i dont understand how it is supposed to work?

 

[stu]

Why do you set the densities a little less? I used to set the maximum densities higher than would be achieved on press. If you limit the density on the yellow it may try and add other colours (usually cyan).

 

[MonsterBabyBLN]

i am talking about setting up a REF!! printer..

i thought you had to limit the densities a little, so you can always relinearalize to that set standard

 

[Johu]

i am talking about setting up a REF!! printer..

i thought you had to limit the densities a little, so you can always relinearalize to that set standard

Tell a bit more about your equipment:
CT version?
Which measurement device?
What sort of paper?

 

[MonsterBabyBLN]

what in the world has the paper and measurement device to do with the process of setting up a reference printer??

CT 1.2 web
DTP 70
semi gloss proofing paper

 

[Stephen Marsh]

MonsterBabyBLN, I can feel your frustration, you could post your question direct to the source:

ORIS Board

We have just become a reseller for CGS ORIS. From my limited experience, you should be able to get decent dE values with 3 iterations when setting up the new hot folder when based off a recent RFP... however due to my inexperience I hesitate to comment too much further.


Stephen Marsh

 

[kseldridge1]

Correct me if I wrong, I remember fighting with yellow in the beginning and i ended up realizing that the color of the paper (blue white epson) would never reproduce a 'litho' yellow. If you are setting up a hotfolder i guess you are working with the iterative color matching to a standard (ie. Gracol7). If you are then check the color of the paper and make sure it is within tolerance of the standard. Measured in Lab values, Gracol would be (a0 b-2) tolerance +- .5 I think.

 

[Johu]

what in the world has the paper and measurement device to do with the process of setting up a reference printer??

CT 1.2 web
DTP 70
semi gloss proofing paper

To answer in the same manner I would say crappy paper, crappy measurement device => crappy process.
But we have now established basics.

When you made the reference profile you probably noticed the dialog telling that it will make the gamut smaller than the one you measured. This way your target reference profile is smaller by a degree (there are 3 different levels, where minor is default). This means that even if you compare the measurement of which you created the rfp file you will see a difference.

What you can easily do, in order to check what is happening, open both your most recent measurement file and your rfp file in gamut viewer. Set you filter to primary colors and measurement points only. There you can see what are the differences in your rfp yellow compared to the measured yellow.

What sometimes happens is that the scanning measurement devices measures one point incorrectly which of course has an effect on your max Delta E (hence measurement device). Also another thing that can happen is that you paper gets too saturated which also can affect your measurements while the composition slightly changes during iteration (you've guessed it paper). And also the paper background color if it is not even.

I've done more than one hundred installations of these systems and seen all of the above. The only mystery part in your configuration is the HP. I have not once installed an HP printer. Only Epsons and Canons. Which brings another factor to our equation... Ink. Yellow ink to my experience tends to have the most drifting while drying up. But again I have no experince of HP ink. Still I must admit 6.7 is alot.

 

[MonsterBabyBLN]

well im glad you didnt say: crappy software, too .. haha ;-)

thanx for taking the time now.. and i totally understand what you are saying! i wasnt aware that YOU were actually the one who knows what im talking about ;-)

1. do you really think the DTP 70 is a crappy device? i thought from what i read all over the place it is one of the most reliable ones in remeasuring.. and i love the small patches and the choice of UV filter..

back to the topic.. yes. im aware that CT crops the gamut.. i chose of course minor.

i usually learned that when building profiles one should go with the cromatic values and not densities. this offer by onix (advanced?) doesnt allow me to change anything. and i dont get a good gamut.. i suppose because the contone drive is too limited then to use the RGB colors?..
so when choosing densities for profile building.. i just go with the highest value.. is that good? would you lower it a little for to be stable? or leave it high becuase everything is limited later on anyways?

 

[Johu]

well im glad you didnt say: crappy software, too .. haha ;-)

thanx for taking the time now.. and i totally understand what you are saying! i wasnt aware that YOU were actually the one who knows what im talking about ;-)

1. do you really think the DTP 70 is a crappy device? i thought from what i read all over the place it is one of the most reliable ones in remeasuring.. and i love the small patches and the choice of UV filter..

back to the topic.. yes. im aware that CT crops the gamut.. i chose of course minor.

i usually learned that when building profiles one should go with the cromatic values and not densities. this offer by onix (advanced?) doesnt allow me to change anything. and i dont get a good gamut.. i suppose because the contone drive is too limited then to use the RGB colors?..
so when choosing densities for profile building.. i just go with the highest value.. is that good? would you lower it a little for to be stable? or leave it high becuase everything is limited later on anyways?

Hi,
my point was not to say your measurement device is crappy (nor the paper) it was just an answer to your what has these got to do with reference calibration. I do have only one experience with DTP 70 and if I remember correctly it measures yellow a bit differently compared to old Gretag devices. But as long as your whole process is done with one device you should be okay. And about the different yellow if you are in US it should conform to your standards. I am only dealing with european standards so the use of UV filter is not that important to me, but I know it's nice if your paper/substrate is very glossy.

I still tend to use densities instead of lab values. But what I do is sometimes I measure the linearization chart in both density and lab values. Then I compare the lab values against the largest profile I would like to reproduce and I cut my densities based on this comparison. To be honest Color Tuner supports the use of both but I have never done full calibration with Lab values.

The problem is that I can not give any suggestion for suitable densities as they vary a lot between different printers and some what also between materials. For example Cyan in Epson can be totally different to Canon max cyan.

Basically your idea of using the slightly cut max density is pretty good. It mostly depends on whether you need to do a lot of max ink reduction. If so you might need to cut your max densities little more. But again with our seeing what is happening it is pretty hard to say something precise.

What I would do is measure linearization chart with Lab values, use measure chart function and save the measure file. Open the measure file in gamut viewer along with the profile you want ro reproduce. Choose measure points for both and set the filter to primaries. This way you can see your device primary colors compared to target profile primaries. I then see which printed point extends beyond target profile and mark the percentage. After this I cut my density values according to these percentages I have discovered in the previous process. This way I am cutting my gamut but I still have enough color to cover my target profile.

I hope this makes sense to you.

 

[MonsterBabyBLN]

thanks mr.. totally makes sense!

two things from my point.. CGS uses the contone driver with HP.. so i dont have to reduce ink limits at all. if i crop densities (even by lab) i get a lot smaller gamut.. specially in the dark blue. i have the feeling that this is then that the contone doesnt use the hp violett enough.. because i already cut the cyan density.

second. when building a reference printer profile.. i want my gamut as large as possible.. to be able to reproduce spotcolors as well.. dont i? so i dont like that clipping of densities just above isoCoated..

 

[Johu]

thanks mr.. totally makes sense!

two things from my point.. CGS uses the contone driver with HP.. so i dont have to reduce ink limits at all. if i crop densities (even by lab) i get a lot smaller gamut.. specially in the dark blue. i have the feeling that this is then that the contone doesnt use the hp violett enough.. because i already cut the cyan density.

second. when building a reference printer profile.. i want my gamut as large as possible.. to be able to reproduce spotcolors as well.. dont i? so i dont like that clipping of densities just above isoCoated..

I'll still insist about checking the primary curves of your printer. Usually cyan and magenta make sort of j-curve. So basically in density point of view the density actually gets smaller while you can visually definately see it gets darker. This j-curve some times can cause problems since you have saturated colors where the color in lab-space does not get directly closer to L-axis when you decrease color.

I might suggest that in your case you could try Lab linearisation instead of densities. Just to see whether that makes any difference. It automatically makes the cut based on the lab values. That is if you have the time to do experiments. I know that I could be much more of help if I could see what is happening, now it's a bit guess work on my part.

 

[MonsterBabyBLN]

oh i really apreciate the time you take!

i luv trial and error.. haha.

i already did the lab method.. and what really amazed me was.. it just does SOMETHING! i dont even see what this "advanced users" method does!.. thats rediculous... it just sets the limit.. boom!
and my profile was a LOT smaller in dark blue.. and didnt gain anything anywhere..

i am used to doing the limitations through cromatic values. as EFI and onyx does it this way. BUT these actually limit each channel.. also the blue and green and red ones.. then it makes sense!

but with CT.. i loose all the blue because CT cuts the cromix of cyan so early that the contone doesnt put any blue down.. is my feeling..

i dont really know what you mean with j curve ? the hooks in the croma ? but u mean.. i dont have probs with M and C...

back to my main question. i mean its an interesting discussion.. but the thing was that set me up..

why do i do a lin.. limit the densities a little bit . then print the it8.74 chart. meassure it.. build ref printer profile.
and 20 mins later.. use the same lin (correct ?) to set up my workfllow.. say use reference printer.. use the just made one.. and then i cant choose the chart for calibration but it always chooses the ECI2002.. ?? and i get this dE 5 off!!.. 20 mins after i just set up the ref printer it takes 4 iterations towards that ref profile.. and i cant get closer than max dE4 ?? thats my point