I covered the basics of inkjet printer profiling with X-Rite i1 Profiler in a previous post. This time I will cover a few tools i1 Profiler has that let you take things a little further, checking the quality of a profile and squeezing a little more accuracy out of the printer. I will assume you’ve successfully created an profile already and output a test image and are generally happy with the results. If a printer profile doesn’t pass the very basic test of outputting a test image with an acceptable level of accuracy then you want to revisit the basics of how it was made and applied. The tips here are for the final minor tweaks, not for correcting a bad profile. We’ll be using our trusty Epson Stylus Photo 1500W again for this article.
i1 Profiler has a couple of different quality checking procedures you can follow. The first is done with the ColorChecker Proof. This chart is a version of the standard 24 patch Macbeth ColorChecker chart that has holes punched in to each colour square. i1 Profiler has a digital version of the chart built in and if you click on the ColorChecker Proof workflow you can select your profile and print out the chart. The list of profiles shown is dependent on wether you have RGB or CMYK Printer selected in Device Selection on the home page of the software. The Show Out-of-Gamut Indicator puts a line through any colour in the target outside the gamut of the printer and hence unlikely to be printed accurately.
Once the print is output and allowed to dry for a few minutes you can then place the ColorChecker Proof target over the print and see how different each colour is. You’ll need to do this under daylight or colour correct lighting, obviously. If you deem any of the colours to be too far out then you may want to think about using a different media type in the inkjet driver, reprofiling or maybe even choosing a different paper.
The second check you can do (if you have the full Publish license version of the software) is designed for CMYK proofing printers and has the advantage of being a measured and not visual assessment of profile quality. If you select CMYK Printer from the Device Selection you’ll see that a Quality workflow becomes available, don’t worry if you created an RGB profile it will still work. You can choose from a number of common pre-press industry standard test wedges, but I’d recommend the UGRA/FOGRA Media Wedge CMYK V3 AED. Advance to the wedge layout page and then click Save as.
Open the resulting tiff file in Photoshop and make sure you assign it the Coated FOGRA39 profile as this is the most common one used for European printing. Print the chart using your printer profile but use Absolute Colorimetric rendering intent. This will attempt to match the paper colour of the FOGRA39 profile on your printer and result in a better match.
Once the print has had time to dry advance to the Measurement page and measure the wedge with your i1 Pro or i1 Pro 2. Select the FOGRA39L reference on the Reference page and then advance to the QA Report.
The QA Report will give you a pass or fail, based on tolerances from FOGRA. The measurements are DeltaE values – measurements of colour difference – and lower values are better. I got a pass, which is pretty impressive given that it’s only a standard photo printer. Passing a FOGRA standard is all well and good if you are proofing work for print but most people outputting on RGB inkjets will be printing photos onto standard photo or fine art papers and so won’t get a pass because the paper colour might be different or the colour gamut too small. However, if you save the reports you can compare accuracy of different papers, the stability of the printer over time, the effect of optimizing the profile, etc. It gives you a numeric way of judging a profile.
The profile optimization workflow in i1 Profiler isn’t wonderfully useful. As long as you use a decent amount of patches to make a profile adding a few more as a second pass isn’t going to make any difference. But, a while ago I did find online a set of 2500 grey patches you can use to optimize a profile and it can lead to a small but significant increase in accuracy. I’ve tried to go back to the website I found it on to give the guy the credit but it isn’t active any more.
Click on the Optimization workflow and then select your profile. Then click on the Pantone swatch icon (the one that isn’t the grid or the head and shoulders). Drag and drop the TC_2502_LAB-GRAY_SCO.cxf file onto the window. I’ll email this file to users on request.
Advance to the Test Chart screen and then Print or Save as the file to output it, using the same settings as the original profile, and still with no colour management. Then measure the prints and re-save the profile. Print out a test image, ColorChecker or FOGRA wedge and you should see some increase in accuracy. In the case of my profile for the Epson 1500W optimizing with the grey patches shaved an average of just over 0.5 DeltaE from the QA Report test.
By default i1 Profiler assumes that the output using a profile will be viewed under a lighting standard called D50. D50 is a daylight standard so that assumption makes sense. However, if you are creating a profile and know that the prints will be displayed under lighting quite different from D50 then you can use the Lighting page to create a profile optimized for that lighting.
You can choose from a list of CIE standards, dial in a daylight colour temperature or even measure the light source. This last option can be useful for galleries. The changes are often very subtle when compared to a D50 profile, especially as modern inks tend to respond fairly consistently to various light sources, but if you are aiming for absolute accuracy and the prints will be displayed under a known and consistent light then it could be worth trying.
The other element of compensation for a light source that i1 Profiler offer is the OBC Profiling workflow. Optical Brightener Compensation is a process of correcting for light sources that have a large element of ultra violet in their output and papers that contain bleaching agents that fluoresce under UV.
OBC Profiling puts an additional step into the profiling workflow. After the initial colour patches have been measured a test chart of gray samples is generated. After this is in turn printed you can use the ColorChecker Proof grey squares or OBC standard card to assess visually, under the light source the print will be viewed under, which greys are the best match on the print. You then enter the patch numbers into i1 Profiler and the correction necessary to counteract the UV effect is calculated when the profile is made. Again, the difference can be subtle but it depends a lot if the light source has large amounts of UV and if the paper has lots of optical brighteners. The software will tell you if OBC is not needed for your paper, i.e. It doesn’t fluoresce in the UV spectrum.
The final way you can tweak a profile is by adjusting how colours are created when the Perceptual rendering intent of the profile is used. The default values are, for some reason known only to X-Rite, labelled as Custom. Two other settings available; Colorful and Saturation. However, you can also adjust the three sliders yourself. I’d recommend creating the profile first with the default settings and assessing the results of a test image printed using perceptual before making any changes.
The Contrast and Saturation sliders do what you’d assume from the names. When you profile a paper with a warm or cool tone the greys in the profile are calculated from the colour of the paper so that they appear neutral to paper white. The Neutralize Grey slider allows you to force the greys to become more absolutely neutral by dialing in a lower value or match the paper more by setting a higher value.
These adjustments can be useful to enhance the appearance of the prints, but there is always a chance you’ll make the print less accurate in the process. The changes will have no effect when saturation or either of the colorimetric intents are used.
None of these workflows and procedures is usually necessary as the profiles i1 Profiler makes are very good, but they are useful tools if you want to achieve the very highest level of print accuracy. As ever I’m more than happy to answer any questions so please comment on the blog.