Hello Matthew,
I've no specific experience with the Wasatch Softrip, but have made quite a number of profiles for printers using various other rips. Here are some suggestions from my side.
- Do your initial profiling on neutral white media. Optical brightener agents will not make it any easier if good color matching is your aim. Depending on the amount of OBA's, you can expect different measurement results from either uv-cut or uv-included spectrophotometers
- Changing the rendering intent has a big impact on the output. The perceptual intent is where profiling engines differ most, and which you can influence most.
- When limiting the separate ink channels, do not be mislead by density readings, but limit the ink where the chroma no longer rises significantly. For the black channel, use L*. You can use Measuretool for this task, but there are other tools as well. If possible, include some secondary colors in your test files, to find out what the impact of your ink limiting is on them.
- For a CMYKOG printer profile, more patches are needed than for CMYK output. I use 1400-2000 patches for CMYK, so I'd certainly go for 2000+.
- A high level of GCR has a positive impact on the neutrality of the gray balance and ink usage. i1Profiler has a 'super GCR' called Full Scale Black, which yields very good results.
- When generating you profiling test chart and you intend to use a high degree of GCR, in i1Profiler you can anticipate this by giving preference to patches containing more black.
- Test your output profile using a known reference. For visually assessing the outcome of RGB files, I always use the Printer Evaluation Image file from OutbackPrint because it contains a lot of memory colors, has gradients, RGB gray, and patches to check highlight and shadow detail (see
http://www.outbackprint.com/printinginsights/pi048/essay.html ). If the gray image in the middle does not look neutral, then your output profile may need to be rebuilt using other settings or remeasured. For RGB sources, I mostly use the perceptual rendering intent. There should be a good visual match between the printout and Photoshop.
- Another nice tool to test the output profile, is the Colorchecker 24 target (CC24). You have one in your i1Profiler box. Download an L*a*b* version of the file from brucelindbloom.com and print it. Compare the output to your CC24 card. You can also measure your own CC24 and recreate it in Photoshop. L*a*b* files print absolute colorimetric in most rips, so measuring the printed colors gives you an idea of how numerically accurate your profile is. You can use similar techniques for (Pantone) spot colors. Many customers in the wide format printing business claim not to need 'proofing grade' output profiles, but then they say they want accurate Pantone colors
- For CMYK source images, I do the visual testing with the AltonaVisual file (
www.eci.org), of which I have good offset printed reference sheets, from standardized presses. For numeric accuracy, you can use either the Fogra MediaWedge, or the Idealliance 2009 strip. The idea is that you print this CMYK strip absolute colorimetric with the right reference profile (Fogra39 based, Gracol, ...), and then compare the L*a*b* values of the reference and sample. You can do this with i1Profiler too. Most solvent printers with their commonly used media will not pass the test, but the average delta E might give you an idea how reliable your output condition is
- Use the profile optimization scheme in i1Profiler. Although it's not the same as profile iteration techniques used in e.g. EFI Colorproof XF (= Fiery XF), this may be just what you need to get there
- If this all fails, try building a CMYK profile, or CMYKOG with less orange. I've read an in Agfa report on profiling CMYKOG printers that the use of the orange ink can cause visually too reddish skin tones, brown bricks that are too red ALTHOUGH the measured proof verification passes. Try matching an Epson x900 to an Epson x880, and you'll get the idea.
Should you have anymore questions, we're here to help
Regards,
Yann
