color matching

[SIGNTIME]

I am trying to print a matching color to 7125 burgundy. My process is... I take my sample color provided (in this case a piece of cut vinyl) and pick the closest match on a printed color chart on the wall. I then try to determine what adjustment to make to the color( more black, less yellow ect...) and print a bunch of blocks of different colors with different variations in increments of 5 or so( ill add 5% k 10% k 15% k ect..) then pick the closest and refine in smaller increments until I get a decent match, which is still usally not exact.

How do others complete this task. What tools do you use (charts, Spectrometers?). This process, depending on the color seems to eat up to much time and not produce perfect results.

 

[Kwiksigns]

I have never used it, but it seems like it would be helpful in your situation.

http://www.catzper.com/ISAPI/PC/CATZperISAPI.dll/$

 

[P Wagner]

Our process is to take a spot color measurement of the specimen using a spectrophotometer. This provides an accurate LAB-based definition of the target color (in your case it would be a sample of 3M Electrocut 7125 vinyl). From there, we can plot the color's position in relation to out printer's gamut on the media in question. This gamut is based on the custom ICC profile that is being used by the RIP, so it is highly accurate to our printer's ability to produce color on the specific media. These plots shows the location of Pantone 360 C in relation to the gamut of 3M IJ180C vinyl on an HP Latex 260, @ 8-pass, using 3D and 2D plot renderings.

Alternatively, we can generate a predictive Delta-E in the RIP. Either will give us a very accurate prediction of our ability to simulate the color(s) in question. In this case, the predictive dE was 1.88

By the way, we try to avoid using the term 'Color Matching', as it can reinforce unrealistic expectations. Terms like 'Color Simulation' are preferable.

 

[dypinc]

I am trying to print a matching color to 7125 burgundy. My process is... I take my sample color provided (in this case a piece of cut vinyl) and pick the closest match on a printed color chart on the wall. I then try to determine what adjustment to make to the color( more black, less yellow ect...) and print a bunch of blocks of different colors with different variations in increments of 5 or so( ill add 5% k 10% k 15% k ect..) then pick the closest and refine in smaller increments until I get a decent match, which is still usally not exact.

How do others complete this task. What tools do you use (charts, Spectrometers?). This process, depending on the color seems to eat up to much time and not produce perfect results.

A proper color managed workflow. The most important part is to create an ink limit/linearization/profile for maximum color gamut of your printer/media combination. If that doesn't get you there then read in the lab values from a pantone book spot color with your spectro and update those values in your RIP color library. If you can't get it close enough then your printer/media combination color gamut is probably not large enough to get it and you will just have to live with the best that you can do.

 

[SIGNTIME]

Ok and then if you don't have a spectrometer or a good grasp on profiling.

 

[petepaz]

if you have to do a lot of color matching the best/ easiest way to start off is get catzper by nazdar. leave all the guess work out and if you want to get involved get the spectrometer and learn how to use it.

 

[petepaz]

if you keep going they way you are doing it you waste a lot of time and material trying to get the right color

 

[P Wagner]

What's your preferred spectrophotometer?

Looking at finally getting one soon and not sure which way to go. Would likely be used in with Onyx Thrive and/or Roland Versaworks.

To do (almost) everything needed, while keeping budgetary issues in mind, I would recommend the XRite i1 Pro 2. If budget is not an issue, I would go with the Barbieri LFP. The XRite device is supported by both Versaworks and Onyx, among other RIPs. Versaworks supports the XRite, but I am not sure about the Barbieri device.

 

[Andy_warp]

A good profile is key, but most of us have done the trial and error method.

I've found most newer rips allow lightness, hue, and chroma variables.
I usually get a hit on the first try. Hard to get the flavor right through cmyk for me.

We have a managed color workflow, but still have to use this method for out of gamut spot colors.
(which is a third of them!!!)

 

[bob]

I am trying to print a matching color to 7125 burgundy. My process is... I take my sample color provided (in this case a piece of cut vinyl) and pick the closest match on a printed color chart on the wall. I then try to determine what adjustment to make to the color( more black, less yellow ect...) and print a bunch of blocks of different colors with different variations in increments of 5 or so( ill add 5% k 10% k 15% k ect..) then pick the closest and refine in smaller increments until I get a decent match, which is still usally not exact.

How do others complete this task. What tools do you use (charts, Spectrometers?). This process, depending on the color seems to eat up to much time and not produce perfect results.

There are no perfect results, just close enough.

Pay little heed to the profilistas in these waters. They're fond of polysyllabic technical terms but they don't produce any result that's in any way superior to simply matching a sample to a color chart. I've been doing this favor of work since their parents were making in their pants, I know from which I speak.

When you have some color or another for which you can't find a satisfactory match then pick a color on each side of your sample. Draw two squares. Aligned horizontally if you're anal but it really doesn't matter as long as they are separated by at least 10 or so of that square. Set one square to one of the colors you selected, set the other square to the other color. Select both squares and do a blend with 10 steps. What you want is each of your end squares and 10 blend squares in between. Print it. If one of those steps produces what you're looking for, then the color for that square is what you should use. If not try changing the color of one or both of the endpoint squares and do it again. Once usually works for most people.

 

[Asuma01]

You dont mention what RIP you use. Swatch Books in Onyx is a great tool for color matching if you have a spectrometer.

 

[derekw13029]

There are no perfect results, just close enough.

Pay little heed to the profilistas in these waters. They're fond of polysyllabic technical terms but they don't produce any result that's in any way superior to simply matching a sample to a color chart. I've been doing this favor of work since their parents were making in their pants, I know from which I speak.

When you have some color or another for which you can't find a satisfactory match then pick a color on each side of your sample. Draw two squares. Aligned horizontally if you're anal but it really doesn't matter as long as they are separated by at least 10 or so of that square. Set one square to one of the colors you selected, set the other square to the other color. Select both squares and do a blend with 10 steps. What you want is each of your end squares and 10 blend squares in between. Print it. If one of those steps produces what you're looking for, then the color for that square is what you should use. If not try changing the color of one or both of the endpoint squares and do it again. Once usually works for most people.

Yeah, I've never used spectro's so I can't say one way or the other.

But I work with old timers who definitely do things the old fashioned way.

If you are in the print business, you should have a good enough eye and an excellent working knowledge of color theory. There's really no reason why you can't print one sample test, see where it is, determine what colors need to move in what direction, and simply make it happen.

I do all my color matching (and yes, it IS color matching. There's no color simulation in fine art giclee printing. These people want an exact match, and we deliver) in Photoshop. Basically the only adjustments I need are levels/curves, hue/saturation, and selective color.

If I can't get it done with those options, then I've failed, not the equipment or my technique.

Yes, printing lots of proof is wasting time, energy, and money. But we charge our clients a fairly hefty proofing fee, because they want an EXACT match to their art. And the more you proof, the better you get. There are many times we will take a scan of a painting, print a raw proof, look it over and nail it on the first or second try.

There's no substitute for color theory knowledge.

 

[shoresigns]

Ok and then if you don't have a spectrometer or a good grasp on profiling.

Then you print a colour chart on each material that you use regularly, and match the colour visually like the rest of us.

 

[P Wagner]

Yeah, I've never used spectro's so I can't say one way or the other.

But I work with old timers who definitely do things the old fashioned way.

If you are in the print business, you should have a good enough eye and an excellent working knowledge of color theory. There's really no reason why you can't print one sample test, see where it is, determine what colors need to move in what direction, and simply make it happen.

I do all my color matching (and yes, it IS color matching. There's no color simulation in fine art giclee printing. These people want an exact match, and we deliver) in Photoshop. Basically the only adjustments I need are levels/curves, hue/saturation, and selective color.

If I can't get it done with those options, then I've failed, not the equipment or my technique.

Yes, printing lots of proof is wasting time, energy, and money. But we charge our clients a fairly hefty proofing fee, because they want an EXACT match to their art. And the more you proof, the better you get. There are many times we will take a scan of a painting, print a raw proof, look it over and nail it on the first or second try.

There's no substitute for color theory knowledge.

I agree that your business (art reproduction) is a very different animal than sign and display graphics. I would guess that you are using very different machines with a higher number of ink channels (Epson / Canon / HP). These machines often times have 8, 10, or 12 unique inks. You are also likely printing onto a fairly small population of coated photobase papers and coated art canvas, and the output is expected to be used exclusively indoors. The business segment that those machines are intended for is reproduction of fine art, and that market places a very high priority on gamut, color fidelity and accuracy. Running costs for traditional aqueous printers are relatively high (more expensive ink and media), which is not a major problem in this market segment, due to the relatively high prices that are charged for this output.

Those same traditional water-based Epson, Cannon, and HP printers are also ill suited to the market that this site caters to - the sign and display business. In this market, the ability to output to low-cost, outdoor durable, uncoated media is essential. This market has not historically placed it's highest priority on gamut, but rather on low running costs (ink and media) and outdoor durability.

The process you are describing is a classic 'closed loop' approach to color. You are modifying the colors in the file (using a tool like Photoshop or Lightroom) to make the printer produce the desired result on the media in question. That approach can work well when you are an experienced operator (which I'm guessing that you are), and the number of print-centric variables are low (few classes of print media), and often times dedicating a specific machine to one media.

As you say, there is no substitute for experience and practice when using a closed loop approach to managing color.

When you need to print your jobs jobs on very different printers (say latex and solvent and UV), or very diverse types of media (SAV, textiles, paper, backlit film etc), closed loop systems fall apart very quickly. For those willing to investigate a different approach, here is a link for an outstanding training program that is produced by IDEAlliance (I have no affiliation with it):

http://www.idealliance.org/videos/CMP2.0-Lesson1_Introduction_Color_Management/index.html

Happy printing (using any approach that suits you)!

 

[SIGNTIME]

if you have to do a lot of color matching the best/ easiest way to start off is get catzper by nazdar. leave all the guess work out and if you want to get involved get the spectrometer and learn how to use it.

I did try that out yesterday, only thing I had trouble with was determining the lab values for the color I was trying to reproduce. I ended up searching on google for the lab values of burgundy which got it close but still not right.

 

[dypinc]

I did try that out yesterday, only thing I had trouble with was determining the lab values for the color I was trying to reproduce. I ended up searching on google for the lab values of burgundy which got it close but still not right.

Sounds like your missing link is still an ink limit/linearization/profile of your printer/media combination.

 

[P Wagner]

I did try that out yesterday, only thing I had trouble with was determining the lab values for the color I was trying to reproduce. I ended up searching on google for the lab values of burgundy which got it close but still not right.

If all you require is the ability to capture and define the color of physical specimens, there are a new class of low-cost (< $200) color capture devices that have recently come on the scene. They are able to capture and quantify colors using several color models, including LAB:

http://palette.com/

https://nixsensor.com/

I have used the Cube device, and found it to be most useful for capturing colors out in the field. These units are not able to help with calibrating and profiling the devices in your workflow, their utility is limited to spot color capture.

 

[petepaz]

I did try that out yesterday, only thing I had trouble with was determining the lab values for the color I was trying to reproduce. I ended up searching on google for the lab values of burgundy which got it close but still not right.

catzper can be a pain with certain colors (red/burgandys and grays are the worst) that's not because of the program but because of the cymk printing.
one thing i have done is do my first round and pick the closest color on that sheet then run catzper again using the new close color and sometimes that will get you closer to where you want to be.
just remember you will never be exact and the customer has to be aware of that as well. the closest you may get could be like two or three shades lighter or darker. some colors are just like that and some colors you can get almost exact.

 

[derekw13029]

When you need to print your jobs jobs on very different printers (say latex and solvent and UV), or very diverse types of media (SAV, textiles, paper, backlit film etc), closed loop systems fall apart very quickly.

That's very true. When we get proofs that match, we make sure to note which printer they came from and strictly try to only use that printer to produce any future prints the artist may require.

I will certainly admit that fine art reproduction is a different animal than signs, banners, and graphics.

But I still use the same technique on those products as well. Don't get me wrong, I would LOVE to experiment with a spectro, but I guess at the moment I'll just keep doing what I'm doing.