Creating an ICC profile from a fingerprint?

[a_keyes]

Hello PrintPlanet,

Noob here, requesting some help from all you experts out there. I have recently been tasked with a difficult assignment. I have been sent a physical press fingerprint from a printer we're doing prepress for and I need to build an ICC profile from that fingerprint so that we can output color accurate proofs for the client. The fingerprint was generated with GMG OpenColor we believe. We use EFI Fiery RIP here in house. We had a color consultant out to help us with a few things and I asked him about this. He says that we need reference information (preferably CGATS) to compare the chart against. Without that, there's nothing we can do. The printer refuses to supply us with anything additional.
My question to you is, can this be done? How can I get this to work? We will most likely need to invest in some software, which we're willing to do. We're always trying to grow our capabilities.

 

[benstarr]

The color consultant is correct, you need more information from the printer to do what you're trying to do. If the chart is a standard chart in a standard layout (IT8/7.4 or ECI2002) you may be able to get what you need off it without additional information from the printer.

 

[gordo]

The color consultant is correct [re: "He says that we need reference information (preferably CGATS) to compare the chart against."] [snip]

Wouldn't that only be true if the printer was targeting an industry standard/specification? The fact that prepress can't use an industry standard proofer profile suggests that the printer isn't targeting an industry standard - otherwise why profile the press?

To the OP - Make sure that you validate the mechanical integrity of the press sheet you were supplied with before you attempt to measure it. If the press sheet shows any mechanical issues (like slur, over-emulsification, etc.) then don't bother measuring it because it's unlikely to be repeatable.

 

[a_keyes]

gordo, you are correct about the printer not using industry standards. When I requested the reference data, I requested CGATS data and the response from the printer was that CGATS is an industry standard and "we don't do that here".

 

[benstarr]

Wouldn't that only be true if the printer was targeting an industry standard/specification? The fact that prepress can't use an industry standard proofer profile suggests that the printer isn't targeting an industry standard - otherwise why profile the press?

To the OP - Make sure that you validate the mechanical integrity of the press sheet you were supplied with before you attempt to measure it. If the press sheet shows any mechanical issues (like slur, over-emulsification, etc.) then don't bother measuring it because it's unlikely to be repeatable.

Gordo I believe the issues is that the OP does not have the reference data for the chart that is printed on the finger print, and without that there is no reasonable way to read the chart and create a profile. And of course you are correct on validating the integrity of the press sheet (preferably sheets so an average can be made).

a_keyes If the printer will not provide you with the reference data will they provide you with a PDF of the layout or at minimum the chart? If you can get that there are tools (Color Think, Babel Color) that will let you extract the reference from an image. However getting the reference data is only one portion of the equation. Is the chart you got from the printer formatted for your measuring device? Do you have software that can make the profile?

 

[a_keyes]

Bennstarr, thanks for your help. I have obtained a copy of the chart which was used on the fingerprint. I have downloaded PatchTool from Babel and I'm playing around with that now. I don't believe our i1 measuring device is compatible with the chart. Our color consultant mentioned that this a newer chart and i1 device is getting up there in age. We're looking into multiple profile creation solutions. Do you have any recommendations?

 

[benstarr]

i1 Profiler and BasICColor Print are the two applications I'd recommend for profile creation. Depending on what version of EFI/Fiery you have you may have the option there also.

If this is your first time creating profiles, I'd highly recommend bringing in someone to help you get it done it sounds like you're probably going to need to invest in some software and hardware. I'd also invest the money in having someone come in and train you on how to do it.

 

[mastegman]

Seems an odd way to go about this. Sounds like the printer has only read half the manual. As previously stated you need the reference data to generate a profile from a printer target but giving you just one is pure folly. As any pressman will know the printing press is something of a 'beast' and suffers from variability during any run. As a result you need to read several targets and average the data to get an accurate profile of the characteristics of this particular printing condition. That printing condition depends on a host of determinants principally, the combination of press, substrate, ink and operator. The reason that standardised printing conditions were developed was to make all this easier and ensure that different printers in different locations could achieve an acceptable and consistent result without 'chasing their tail' and driving up the cost of the process. Standardised reference printing conditions reflect a range of achievable printing conditions depending on a specific set of inks and paper for a given printing process and are designed to streamline and optimise print production. In other words, optimise quality, reduce waste and save time. Your printer would be far better off ensuring they are meeting these standards by targeting the process control parameters which underwrite them. You could use a custom profile generated from the freely available characterisation data based on these standards or use one of the freely available profiles that meets the criteria of their specific printing condition. All the work has been done and is constantly refined and updated either by IDEAlliance and/or Fogra/BVDM/ECI. Why re-invent the wheel?

 

[shorty83]

I have been involved in many fingerprints/ICC profile builds. Do you have software to even build an ICC profile? Most software will require a certain chart to be printed on the press and then read in (usually around 5 different sheets to get an average) and then an ICC profile is built. Once you get the ICC profile set up then it can be saved and put into the proofer's RIP. You will need to document which stock, ink, press, line screen, sticky back, plates and inks, etc. were used. These are all variables when running a job on a press that can affect color and a good way to cover your butt when the press guys come back saying the color on the proof is wrong.

 

[John Reveley]

If the profile is Opencolor then this is a new technology from GMG, it will not be able to be read and converted by any other product. Your best bet would be to contact GMG who will be able to advise you fully on what you need to do, they will also be able to confirm if the chart is indeed Opencolor, and what your options are.

 

[Stephen Marsh]

If the profile is Opencolor then this is a new technology from GMG, it will not be able to be read and converted by any other product.

Doesn’t sound very “open” to me! Sorry, I couldn’t resist!


Stephen Marsh

 

[mastegman]

Doesn’t sound very “open” to me! Sorry, I couldn’t resist!


Stephen Marsh

So much for 'standards'.

Mark Stegman

 

[jimbo0270]

Hopefully this is resolved by now but to add my two cents, GMG's Open Color software is incredible powerful but not cheap and does not create ICC profiles. It make a proprietary profile for their rip. It simulates spot colors (solid and tone) very well and can constantly iterate a profile with just a few patches placed in dead areas of a print.

That said, from what I have seen in xRite software and GMG; patch sizes, the layout of patches and the number of patches vary greatly depending on the variable you put into the software (devices like an IO table can read smaller patches than a i1pro2), the physical size of the chart, etc. This is why you need the reference data, so you can make sure that you are scanning the same color patches in the same location for the software to evaluate.

If the chart is output by GMG, it's not specific to Open Color. Most GMG software versions I have seen output their charts in the same way. It's most likely built to the size of the proofing device connect to the rip in a size and pattern that the onboard spectro can read them. Or it could be one of GMGs standard charts that is selected by the way the profile builder would scan the chart.

 

[mikestr]

Gordo is certainly right about verifying the integrity of the press before attempting a profile; if anything he understates this. Profiling a press is not a trivial exercise, and at the risk of sounding dismissive, the basic nature of the questions being asked about procedures indicates to me that the OP should get some experienced help in this. Gordo touches on an overarching question: Why is the press being profiled at all? Offset presses and inks are highly standardized, and with a competent operator, some basic process control, and minor adjustments in plate curves should be able to produce color that corresponds to one of the well known output conditions represented by one of several well-known "off-the-shelf" profiles, which would also server as the proofing reference. There are exceptions of course: unusual substrate, inks, or screening type, extended color sets such as hi-fi 6 and 7-color, etc. But most of the time when "fingerprinting" and custom profiles are demanded for CMYK it is simply an indication that someone hasn't been controlling the press or using a proofing system correctly, or both.

 

[jimbo0270]

Gordo is certainly right about verifying the integrity of the press before attempting a profile; if anything he understates this. Profiling a press is not a trivial exercise, and at the risk of sounding dismissive, the basic nature of the questions being asked about procedures indicates to me that the OP should get some experienced help in this. Gordo touches on an overarching question: Why is the press being profiled at all? Offset presses and inks are highly standardized, and with a competent operator, some basic process control, and minor adjustments in plate curves should be able to produce color that corresponds to one of the well known output conditions represented by one of several well-known "off-the-shelf" profiles, which would also server as the proofing reference. There are exceptions of course: unusual substrate, inks, or screening type, extended color sets such as hi-fi 6 and 7-color, etc. But most of the time when "fingerprinting" and custom profiles are demanded for CMYK it is simply an indication that someone hasn't been controlling the press or using a proofing system correctly, or both.

Well, he never specified it was offset. I assume that if a GMG profile is involved that it was Flexo or Roto print as that is where GMG dominates as a standard proof in the market today. Also, there are offset printers out there that are not using the standard ink sets or want a profile that simulates the white point of their printing stock. Just a couple of the reasons I have seen offset printers do a fingerprint.