Go ahead make my day, C40 M48 Y48 K8 go ahead and print it with EcoSolvent inks on an Arlon DPF 4500LX media and laminate it...
Well, see, here's the problem right off the bat. Not only would I never try to do that, but part of the service I offer my clients -- and since they get lifetime free tech support, if it doesn't sink in right away, they always have me to fall back on -- is that that's a fool's errand.
See, when you say C40 M48 Y48 K8 or any other CMYK value, you have to define a specific CMYK color space in which those values represent a specific color.
And understand, that's all they do. They represent a specific color in a particular CMYK color space.
And I'm sorry, that's not theoretical, that is the real world.
So let’s just do a little research here, shall we?
Now, you rattled off Gracol and ECI and something to do with ISO, but just to keep it easy for this test I’m just going to assume those numbers in the almost universal US CMYK default, SWOP.
And I get them as representing L55 a9 b8 in Lab.
Now again it’s extremely important that this is not in any way “theoretical.” This is absolutely real world, and whether you want to understand it or not, or whether you want to accept it or not, this is how digital printing works.
Every pixel in every digital image is nothing more than a box full of numbers. And the numbers relate to some color space, whether the file is tagged or untagged. And that color value in that color space relates to some specific value in a “profile connection space” – which is the real, actual color value of that pixel. (This is why, incidentally, it does not matter whether you send a RIP RGB or CMYK files. In that case, the terms RGB and CMYK represent pixel definitions, not printing processes.) And again, that’s not theory. That’s simply the way it is.
Okay, so, let’s just take a look at what CMYK values L55 a9 b8 represents in some profiles I’ve made on other inkjet printers.
Here’s a Mimaki JV33. Inkset is Mimaki EcoSol CMYKcm. RIP is Onyx. Stock is cast vinyl.
C35; M59; Y59; K10.
A Seiko M64S. Inkset is CMYKcmk. RIP is Onyx. Stock is cast vinyl.
C24; M46; Y30; K27
A Seiko H274S. Inkset is CMYKcmkllk RIP is Onyx. Stock is cast vinyl.
C27; M49; Y37; K29.
So what does this show?
Several things, actually. And the first is just how important black and light-light black are to controlling hue shift.
Because it is imperative to understand that hue-shift is caused by creating the grey base neutral colors out of CM&Y out of printing condition necessity, as opposed to creating them out of black. The earlier and the more aggressively you can bring in black, the more you can control hue-shift. That is the reason for light black and light-light black. The problem is that a lot of people who write profiles for light and light black inksets don't understand that. And if they're not properly taken advantage of in a profile, they might as well not be there.
I made all of these profiles, all of them are for vehicle wrappers, and one intent of all of them is to control hue-shift in neutrals to the largest extent possible.
And understand the problem: Can I bring in the black on any or all of these machines earlier and more aggressively?
Absolutely.
And, if I do, the hue shift will get correspondingly better – at the expense of graininess. So it’s a judgment call. And many times it depends on what the client wants the final piece to do, and where he wants it to go. But, having been doing this for ten years, for clients all over the world – the real one, not a theoretical one – I can tell you it works for them.
So, yes, it’s true, the machine you’re using puts you at a disadvantage right from the start with this particular issue. But don’t get upset with me for pointing that out. And note that it doesn’t change the fact that if you have competitors with the other machines shown above, if they’re profiled correctly, they have a pretty good leg up on you in this area.
Ohh… And note that since every one of these profiles was made to be used with laminated media, they were all made with the profile patches laminated before reading. That way they are in their true final color state laminated, and their profiles direct the printers accordingly. Meaning that they are completely color accurate outside, laminated.
So, what would I do in your little scenario outlined above?
What I’d do is take a swatch of the actual color the client wants to hit, and I’d read it into the RIP with a spectrophotometer. Then I’d give that color a name, and that Lab value and name would be that color. And it’s that name I’d use in all my design applications.
Then, as long as my printers are properly profiled in my environment, I can be absolutely confident that every time I send that color to print, I am getting as close a match as it is physically possible for me to get.
And it works. I’ve set up workflows just like this from coast to coast. And since every client gets
The Correct Color Guarantee, and lifetime free tech support, if you want to deny that it works, fine, you can.
Just as tomorrow morning you can deny the sunrise.
Just to add:
What I've seen happen with manufacturer created profiles is in more recent cases is that they are made outside the U.S., so ISO - ECI, Japanese - ISO profiles are used when generating test prints for profiling…
Untrue.
An ICC profile of any device is a characterization of that device reproducing color in a particular state. So, the way the profile is made is that the device is brought into that state, and then it is sent a series of raw color values, it reproduces them, then those reproduced colors are read with some sort of color reading device, and those readings are used to create the characterization – the ICC profile.
Let's stick with one RIP, Rasterlynk 6.6.
Oh, let’s not…
See, what any RIP does is it converts pixel information into dot information based on information in profiles.
In profiles.
See, that’s what a lot of people – you evidently included – seem to have trouble wrapping their minds around. The RIP creates an RTL file, which is the printing dots. The RIP has no idea how the printer actually is going to print those dots.
None. All it does is make dots per the profile.
So the more the profile captures every bit of your printer’s capability, and the more it’s designed to do exactly what you want to do on a particular media, and the more it matches how your printer in your environment actually prints, the better the printing dots you’re going to get.
So in this business, profiles are everything.
And yet unlike every other RIP out there, Rasterlink does not come with the tools needed to create custom media profiles. Not even the in-RIP machine-state parts.
To me, that rules it out as a serious tool for anyone in any way serious about color.
…we lose $1200 in production creating a profile.
Creating profiles is an art, and a science. But on machines such as your machine – assuming it was run with a real RIP – I typically make about five profiles a day.
The medias change sometimes 3 times a year without any acknowledgement from the manufacturer, we use 15 different medias at least.
Funny, every profile I make I test four separate ways before I install it and run a test print. And one of those tests is for white point, primaries and black point. So I am pretty familiar with the white points of all the major media, and I can tell you from my experience, well, hell, in all the time I’ve been doing this, I’ve never seen them move.
I have seen some issues with laminates wandering around a little. But even that is minimal enough that virtually every client I have would rather laminate with what they currently have on the floor for profiling than to run unlamented profiles for lammed applications.
Ohhhh – Kay.
15 profiles.
That’d be a three day project. That’s $1250 a day plus expenses. Let’s say expenses run another $1250.00.
You’re looking at 5 grand.
Plus you’d need a real RIP to do this.
Cheapest you could get to do all you need to do would be Onyx RIP Center – which I can sell you. It doesn’t have Media Manager, but you wouldn’t need it because I’d be writing your profiles.
$1695.00
So, you’re into this now for $6695.00
Now of course you’re going to rant and rave and make a snide comment about a stripper’s butt, or a transexual…
But just suppose you’re wrong.
Just suppose it works.
Just suppose all your color issues vanished, just like that?
Would it be worth it?
You do the math.
Please try and explain to my clients what is and what isn't a reasonable lighting condition, I need the humor.
So…
You really think your clients are all
that unique?
You really think that there in Pennsylvania you have some sort of monopoly on clients who are discriminating about color?
Let me tell you something: They aren’t. You don’t.
Fact is I can’t even begin to tell you how many times I’ve heard words to that very same effect, and yet the fact is in my lifetime I’ve dealt with many, many world-class and very discriminating graphic designers, and I know how to talk to them, and in the end what they’re all after.
And in the end what they’re all after is to make sure that whomever they’re working with understands the process they’re working with, understands the physics involved and is getting for them every bit of what is physically possible out of that process.
Now, fact is that in your current environment, you’re not. A competitor right down the street with a real RIP, an H2 Seiko, or an M64S Seiko, or a CMYKOcmk Epson Sure Color properly profiled who understands how to create a spot color can blow you into the weeds on hue-shifting neutrals.
And yes, it is true that competitor will still have some hue shift on that color, but it will be in the “sweet spot” as I like to call it, and since that competitor has done everything he can do to eliminate the problem, and if he can explain it step by step to the client, then the client will be satisfied.
Mike Adams
Correct Color
