Wide Format Substrates

[arossetti]

Anyone have experience trying to color manage wide format and all of its available substrates. I'm having an obvious issue trying to figure out how to deal with variations in white point for all the substrates we print onto. We have ~40 house stocks all with different levels of white and OBA. I know some RIPs allow you to specify the white point of each media and it then adjusts the destination profile accordingly but I don't think Caldera does. The idea of having a profile for each media, on each printer is not really feasible. Any ideas or resources I can research?

 

[kansasquaker]

Are you printing with UV or solvent inks? I can't speak to solvent, but I can tell you that UV inks react very differently to different substrates (and omg do the substrates vary in comparison to the differences in paper). The only correct solution, imo, is to make a separate profile for each substrate and print condition.

Like you, I agree that is impractical. So we profile the most common substrates we run and pick the closest type of substrate for short runs of odd ball materials. We also create one-off profiles for large runs or demanding artwork. The whole system is very ad-hoc compared to how we manage litho work. I don't like it, but I don't see a good alternative. How much thought can you put in to color managing two posters with a selling price of $75?

 

[arossetti]

This is for UV printing. The current setup has 6-8 profiles per printer with substrates grouped based on surface texture over white point. I'm thinking with UV that is less important since it cures quickly with little room for dot gain; anyone have thoughts on this? Should my groups preference white point or substrate texture? I'm currently chewing through Forgra PSD for wide format or what will one day hopefully become ISO 15311-3.

 

[pauly92]

I also use many different medias. I just create different profiles for each one of them. Yes there's minimal to no dot gain on UV printing, But different substrates respond differently to the inks.
Example is PVC and Aluminium composite sheets.
The aluminium needs more ink to reproduce the same colours as the PVC.
My aluminium sheets have a gloss and a matte side. Both sides come out different with the same profile.
It's mostly in the ink restrictions you can see the difference. So even if you do change your white point, the colours will probably be different.

 

[Tommyjt]

I have profiles for each substrate and print condition as well. I don't see a way around it. Now if we could just get consistent substrates we would be good to go!

 

[arossetti]

So this was another thought I had; let me know if it holds water. What if we used 1 linearization file for each printer to be shared with all the medias for that one printer? That way I could easily fix printer drifts globally but have individual transitions and profiles for each media. By its name a linearization shouldn't care what substrate it is on as long as the ramps show consistent progression right? I have 6 printers and like I said up to 40 substrates.

 

[Correct Color]

So this was another thought I had; let me know if it holds water. What if we used 1 linearization file for each printer to be shared with all the medias for that one printer? That way I could easily fix printer drifts globally but have individual transitions and profiles for each media. By its name a linearization shouldn't care what substrate it is on as long as the ramps show consistent progression right? I have 6 printers and like I said up to 40 substrates.

No, that does not hold water.

What an ICC profile is is a characterization of a device reproducing color in a certain state, and linearization is a key component of that state.

Fact is that if you're serious about color management in large format printing, the only option is to profile every media on every machine.

Period.

And it's more than feasible. It's done all the time. It is the accepted practice in this industry.



Mike Adams
Correct Color

 

[arossetti]

What an ICC profile is is a characterization of a device reproducing color in a certain state, and linearization is a key component of that state.

Correct Color

Obviously you have a lot of experience in this arena and I'm not here to debate what your saying just trying to get a deeper understanding. If the linearization brings the printer to a certain state, or known state for the ICC profile to sit on why would it matter if the linearization was done on the substrate that your are profiling versus a common substrate. It would still be a repeatable state every time you relinearize using that substrate would it not? Maybe I'm not looking at all the variables that the DFE uses off of the linearization process; I'm assuming that it is only making curves based on a density and not even concerning itself with LAB or white point. Maybe I'm thinking the ICC will compensate for a incorrect linearization too much. Coming from cut sheet digital this is a common practice to linearize the press to one substrate and share that lin with all medias.

 

[pauly92]

Obviously you have a lot of experience in this arena and I'm not here to debate what your saying just trying to get a deeper understanding. If the linearization brings the printer to a certain state, or known state for the ICC profile to sit on why would it matter if the linearization was done on the substrate that your are profiling versus a common substrate. It would still be a repeatable state every time you relinearize using that substrate would it not? Maybe I'm not looking at all the variables that the DFE uses off of the linearization process; I'm assuming that it is only making curves based on a density and not even concerning itself with LAB or white point. Maybe I'm thinking the ICC will compensate for a incorrect linearization too much. Coming from cut sheet digital this is a common practice to linearize the press to one substrate and share that lin with all medias.

From my understanding using onyx. Depending on your printer, could be a 3 or 4 step progress.
Ink restrictions: Your ink restrictions by printing out a patch set using minimum ink to maximum ink levels. They get restricted to once they produce maximum colour. (No substrate or media is the same, Some use more ink, some use less. Some will get less chroma and some will get more, all at different levels and curves)
Calibration: establishes a baseline on how uses the ink on the substrates / media. So when re calibrated from wear and tear over time, instead of making a whole new profile, it'll move the ink restrictions to reflect how the original calibration was.
Ink limits: looks for over inking when using single and multiple channels of ink.
ICC profile: creates the unique colour profile for the above settings.

So for example, Substrate A and substrate B could be both similar in white points, but how it holds the ink and the maximum colour gamut will be different.
I have 2 different brands of aluminium composite sheets. both hold ink differently.
I went from 1 type of coating to a new one for glass. The new coating uses a lot less ink but has the same gamut range. Actually it's slightly bigger.

From all the time stuffing around, you could create a profile for each substrate. Using the i1p I can understand it'll take some time. Or find someone with a Barbieri LFP. get a4 sizes sheets of each substrate and print the patch set on that and get it sent off to be read.
(you have to use Barbieri Gateway to make the patch set using an existing set, I.e your IT8.7/4 patch set or your eci2002 set. Or make your own from I1P software, What ever suits you. Just need to make them readable for an LFP.)

 

[Correct Color]

arossetti,

Obviously you have a lot of experience in this arena and I'm not here to debate what your saying just trying to get a deeper understanding. If the linearization brings the printer to a certain state, or known state for the ICC profile to sit on why would it matter if the linearization was done on the substrate that your are profiling versus a common substrate.

Because linearization measures dot gain, and dot gain is going to be at least somewhat different on every media.

It would still be a repeatable state every time you relinearize using that substrate would it not? Maybe I'm not looking at all the variables that the DFE uses off of the linearization process; I'm assuming that it is only making curves based on a density and not even concerning itself with LAB or white point. Maybe I'm thinking the ICC will compensate for a incorrect linearization too much.

ICC profiles don't compensate for anything. They characterize a certain device reproducing color in a certain state. Their function is to convert L*a*b* values -- pixel by pixel -- from the incoming color space to the destination color space. Those values in the destination color space are determined by all the variables created in the machine state -- linearization very important among them. By whatever amount any variable in the machine state is different from the machine state used to produce the patches measured to create any ICC profile, that profile is then that much invalid.

Coming from cut sheet digital this is a common practice to linearize the press to one substrate and share that lin with all medias.

You can try. You may even be happy with the result. But just note that that is not accepted industry practice. Further, my own personal rule of thumb is that in truly demanding critical color conditions, I never recommend relinearizing at all. The reason being that no two linearizations are ever exactly the same. And every ICC profile is built on the machine state of which a specific linearization is an integral part. My money says if it needs to be relinearized, it needs to be reprofiled as well.

What will happen if you go that route is that the farther you get from neutral, the closer your colors will match, but the closer you get to neutral, the more you'll have issues.

(Edited to add: I went back and reread this and noticed a couple people have said that UV printers don't have appreciable dot gain.

Not quite sure where that idea came from, but it is not so. I've made thousands of UV printer profiles over the years, and I can assure you that every single one of them has had its own unique dot gain curve -- I check dot gain curves twice in my profile verification process -- and not overall significantly any less of a curve than any other form of inkjet.

Keep in mind that the vast majority of dot gain is relatively instantaneous. The ink is a liquid and as soon as it hits the media, it disperses. By the time the light hits it, it's had plenty of time to gain.)


Mike