How to Make an accurate profile for inkjet/solvent printers?


I’m a newbie to this forum, just getting into color profiling!

I have a small print house with a few inkjet printers (both solvent and aqueous). I’ve purchased I1Publish (with I11Profiler) and I run a WaSatch SoftRip for both types of printers. I’m struggling with getting accurate color prints. Can anyone provide me with instructions on how to make a VERY ACCURATE color profile? Here is the process I’ve been doing so far:

  1. Print a 6 color (CMYK + O + G) unrestricted channel target that I’ve created in PS CS5 (with squares for each % from 1 to 100%).
    1a) Print a total ink limit chart (0 to 600%) and evaluate how the inks are looking combined.
  2. Determine and set what I think is the proper channel ink limit.
  3. Print a 6 color linearization target from my WaSatch Rip PS files.
  4. Scan with my EyeOne Pro the densities of the targets and create color curves in my Rip.
  5. Print a total ink limit chart again and evaluate what the total ink limit I should use is (I use this limit in my profile making software i1profiler, NOT my WaSatch Rip - I leave that limit at 600%).
  6. Create patches to print and scan for generating my ICC profile.
  7. Scan into I1profiler
  8. Use results and settings to generate a ICC profile.
  9. Add profile back into imaging configuration in WaSatch Rip.

This process results in reasonable profiles, but nothing that is “Dead On” or a “true Match” when I compare the prints in my Judge 2 light booth with my Ezio CG2330 monitor (which auto profiles itself once a week).

Any one see something I’m not doing that will really make my profiles accurate?

Any advice helps!



Hi Matt,

First I have to get this quibble off my chest: Anytime one is dealing with a RIP you need to decide where to come down on the continuum between maximum gamut and stable color. Either one of these extremes can be considered accurate color, depending on who you’re talking to. Most people want to be somewhere in the middle.

Your steps sound right in line with good RIP setup workflow. I might take exception to #5 - If you set your total ink limit in the RIP, then you have less chance of over-inking your profiling target, and the patches in your target would be spread more uniformly over the range of printable colors on your machine. But I don’t think this total ink limit choice would be a very big deal either way.

I would be more interested in hearing if your paper white in the viewing booth matches the white of your calibrated Eizo monitor. This is what we refer to around the office as the “white paper test.” In order to get a screen to print match, you have to start by making sure that the whites of a blank piece of your printer paper match the white background of your monitor. All the other colors are scaled to those whites.

More on this whole subject here: … e_of_paper


Thanks for the quick response! One of the reasons I do leave the ink limit on 600% when printing my patches is because I change the print mode during patch printing to uni-directional, media/printer width (instead of data width) and add a 3 second drying time delay.

I’ve been doing this since I watched a training video from Colorburst/Epson employee Roy Bohnen (who I’ve heard referred to as “The Godfather of Color”). In his presentation he explains that while you would never run a printer at 600% ink limit in a production environment, if you choke your ink limit down in your patch printing, you will reduce the color gamut possible because the ability to reach some patch colors/builds will be reduced. So you want to build your profile with as much accuracy as possible and choke the ink limit down after the patches are printed to ensure the largest gamut ICC profile but also reduce over inking on your production prints.

Just a thought, I’ve tried both ways and feel I get a larger gamut by choking the ink after the profile patches have been scanned.

Also, I noticed you replied to the post on: Ink limiting, linearizing, total ink limit [, where another contributor asks about scanning densities instead of using a visual inspection. Could you use an I1 as a densitometer and scan to find where the optimal ink limit is per channel by comparing densities and finding where the highest reading is before it starts to go back down? I use a chart that as 0-100% ink patches in single percent increments so I feel that if I do this I can really dial in on my individual channel ink limits.

Could I also do this for my 0-600% ink limit (total ink limit) by scanning the density (with this I use a 10% increment).



So you come down in the direction of maximizing your gamut. That’s a perfectly valid option and you’re following the correct path to get there.

The MeasureTool function in the ProfileMaker suite will give you density values calculated from the spectral readings in your i1Pro. This can only be done with CMYK readings (not RGB), and you need a dongle to activate this function.

Otherwise, you could do this sort of by hand by plugging your Lab measurements (actually just your L*) into the companding calculator at Bruce Lindbloom’s site:

I’m not that familiar with Wasatch. Do they not offer an adequate procedure for getting a good linearization for their RIP? While the idea of using spectrophotometer measurements is a good one, density does not always tell the whole story. Some RIPs will use the measurement of Chroma to calculate ideal linearization, since that tends to reflect the size of the gamut. It is possible to have ink become less ‘colorful’ the more of it that is laid down.

Here’s a spider graph of a 6-channel printer, showing the primary color ramps. The green and the magenta actually hook around and become slightly desaturated at the heavier patches. But if you can isolate the point of maximum chroma, that would be more useful for linearization.