Discussion: View Thread

Multiples of the same model printer

  • 1.  Multiples of the same model printer

    Posted 04-04-2017 02:34 PM
    Hello,

    I wanted to get others opinions on profiling multiple of the same model printer. I have groups of the same printers and was wondering what others do for profiling. Do you give them all their own curves and icc profiles? or is it better to have one icc profile for all of them and linearize each printer individually? They are setup to run the same media on each group of printers.

    Thanks,
    Nick

    ------------------------------
    Nicholas Reed-Kutz
    Color Technologist
    Sign-Zone, Inc.
    Anoka MN
    (612) 224-4766
    ------------------------------


  • 2.  RE: Multiples of the same model printer

    Posted 04-05-2017 09:25 AM
    Hi Nicholas,

    Ideally, you will want to calibrate each device individually to keep each device consistent to itself. To answer the  "how many profiles" question: It depends on your Color Tolerance (Expectation factor for color match) that you are willing to accept compared to how close all printers actually print to one another. If your Expectation factor for a given color match is 3 (delta E (00) at the 95%) then we can compare all 25 printers to one another, and if they are all within three, you can get buy with one curve or one profile. And if not, you can assess how close they are, and group them based on your Expectations. In fact, you can compare all 25 printers (even on 25 different substrates) and we can organize them into as many groups as your Expectation factors requires, for instance, recently I was comparing 25 different print conditions, (different printers and substrates), and if I asked for a 3 Expectation factor it would group them in 13 groups, if I asked for a 5 it grouped them in 6 groups. And it shows me which device is most like the others in each group, so that would be the one to profile or curve!! 

    I did this with multiple linked spreadsheets for years, but now it is automated and on line, you can try it for free. Works with any color target, and any measurement file. Thanks, 

    ------------------------------
    David Hunter
    Principal
    Pilot Marketing Group
    Forest Lake MN
    651-717-0590
    ------------------------------



  • 3.  RE: Multiples of the same model printer

    Posted 04-05-2017 09:25 AM
    A common procedure is to use a common linearization for printers of identical model and media and to profile or optimize each one individually, depending on what options the software provides. That optimization may be a relinearization or something else, like an abbreviated reprofiling--again, it depends on the software. It's even possible with some RIPs to optimize to a reference profile, which can be a press profile or another printer. 

    ------------------------------
    Mike Strickler
    MSP Graphic Services
    ------------------------------



  • 4.  RE: Multiples of the same model printer

    Posted 04-05-2017 09:25 AM
    Nick, 

    I assume your goal is to make them all print as close as possible, or match each other. If that is the case, using a single .icc profile and linearizing each may be best. To go a little deeper, and assuming inkjet technology, you may want to mimic the saturation point of the target device to all other devices of the same model when doing your linearization. This helps the presses mimic the target device during linearization rather than letting each device vary its saturation point. This will help align all presses to your target device. Once you have done that, using the same .icc profile should give you very similar results across all devices of the same model. 

    Please contact me if we can assist further. 

    Allen

    ------------------------------
    AllenFilson
    Sr. Manager Professional Services
    Canon USA
    Boca Raton, FL
    ------------------------------



  • 5.  RE: Multiples of the same model printer

    Posted 04-10-2017 09:41 AM
    Yes they are all inkjet printers. we do base our tolerance on the 3 Delta E. We are using the caldera rip software and it allows us to use any icc and curves. Our main color space aim is Gracol. Would a good practice be to find out which printer profile is closer to Gracol in each group of printers and set them all to that icc and then just linearize them individually with a P2P to keep them in line?

    ------------------------------
    Nicholas Reed-Kutz
    Color Technologist
    Sign-Zone, Inc.
    Anoka MN
    (612) 224-4766
    ------------------------------



  • 6.  RE: Multiples of the same model printer

    Posted 04-11-2017 09:33 AM
    Nick,

    You've got the right idea. Now it's time to experiment and review the outcomes. Nothing like a little action to move you towards the perfect answer for your individual application. Please keep us posted as I for one would love to know how this works out and what works best for you.


    ------------------------------
    James Raffel
    Color Management Consultant & CEO
    G7 Expert
    262-820-1131
    ------------------------------



  • 7.  RE: Multiples of the same model printer

    Posted 04-12-2017 11:22 AM
    Nick,

    My experience says that if you are trying to achieve a close visual match between printing devices several things have to happen:
    1. Printer gamuts have to be pretty close between devices. This of course has a lot to do with substrate texture and ink texture (rough textured media and/or UV inks both exhibit more light scattering properties due to their roughness vs. solvent on glossy substrates).
    2. Evaluating more than just the worst ∆E value is necessary. You need to know how all the patches in a control strip compare in ∆E (the more, the better within practical daily use consideration), not just the worst or the average. The more patches under 1 ∆E, the more likely the printing is visually close, as you are comparing all patches in the strip to themselves and ranking them on visual closeness.
    3. You can't compare to an industry reference, like GRACoL, when visually comparing devices. You have to compare one device as the reference to the other, because that's what you're looking at in the viewing area. You can't see GRACoL, as there is no perfect GRACoL proof, but you certainly can see the difference between printer A and printer B, so make printer A the reference when comparing those two devices. Hopefully with grouping tests you can compare multiple devices to one device.
    Tight calibration of the device and the ability to truly recalibrate back to the same known state the device was in when profiled is key. I don't see what is being called recalibration working too well in the field. Some RIPs are better than others, but the bottom line is for true recalibration to work successfully it has to be a two-part process. First, you have to achieve the same solid ink value that was in the original calibration, and second, you have to then create the same curve along the values between 0% and 100%. Most RIPs do the latter, but few actually do the former during the recalibration process. At any rate, if you can't fully recalibrate the printer, the original profile is going to eventually be far enough off the mark, that it becomes useless.

    Another point I've proven out many times when calibrating two exact devices is that very often they do not print the same right out of the box. There is no way to successfully use a single profile for multiple devices that aren't even close and achieve a tight visual match. My advice is to target the same source reference space (GRACoL as an example) and calibrate and profile them as carefully as possible to achieve a tight match (you can also compare profile output data between calibrated devices to see just how different they are once calibrated). Now that being said, RIPs that have iterative optimization have a much better chance of achieving a tight calibration between multiple devices than RIPs that can only rely on ink limits, linearization, and icc profiles.

    You certainly can and should run comparison tests between all your devices, ideally on a single substate all devices can print on, to identify which devices are the closest to one another and group them accordingly. The point here is to get to know each and every device (it's gamut, how consistently it prints, etc.). Maybe you get lucky and find several devices that actually are close enough to calibrate using a single profile. Only this experiment will tell. I have installed many pairs of Epson aqueous printers and have never found two that calibrate the same or profile the same.

    My two cents. Hope this helps.

    ------------------------------
    Bruce Bayne
    President
    Alder Technology
    Portland OR
    503-226-7598
    ------------------------------



  • 8.  RE: Multiples of the same model printer

    Posted 04-10-2017 09:51 AM
    Hello Nick,

    Some of this depends upon the RIP technology you are using to drive these printers. In a perfect world I'd want to do the following.

    1. Set the primary ink restrictions of each printer to match. While time consuming this is a fairly easy to implement once you have completed all these steps for 1 printer.

    2. Create a printer calibration based on a known good print condition. I'm a pretty big fan of G7. I find this favorable over built in RIP linearization when matching between devices is the end goal.

    3. Build an ICC profile based upon #1 and #2 above. 

    4. If the output of that ICC profile meets your expectations, then perform steps #1 and #2 above on the second printer, and apply the ICC profile built in step #3 above.

    5. As David Hunter suggested, use software to evaluate the results achieved in this small test group. 

    6. Depending how #5 goes, decided if you can continue the process on each printer. As you go, the printers may place themselves in buckets of go/no go. 

    7. The no go printers may need maintenance or just may need to be grouped in such a way (again, as David suggested) so that multiple profiles can be utilized. 

    Off the top of my head, that's how I'd attack this.


    ------------------------------
    James Raffel
    Color Management Consultant & CEO
    G7 Expert
    262-820-1131
    ------------------------------



  • 9.  RE: Multiples of the same model printer

    Posted 04-11-2017 11:49 AM
    Nick
    What version of Caldera are you using and do you have the Easy Media plugin?


    ------------------------------
    Bruce Compton
    Print R&D Director
    Zon Retail Environments
    Long Lake MN
    (952) 473-5449
    ------------------------------