Warning: The profile check resulted in some deviations.

I have only recently started making printer profiles, with an Eye-One Pro and Match 3.0.1. The first attempt was for an Epson R800 on a cheap everyday matt-coated paper.

I printed the two-page RGB target, which looked ok, and was measured well dried in strip mode. When Match came to calculate the profile however, it threw up the alert: “Warning: The profile check resulted in some deviations. If you decide to continue please test the profile before you use it in production…”. I re-read the targets with the same result, and then repeated the procedure with the one-page target. Same again. The profile seems ok in practice as far as I can tell.

Mike Cummings reassures me that Match often throws up these spurious alerts, especially with matt papers, but I am posting here partly to inform others who might be in the same boat as me, and partly to ask for other peoples take on this problem. What sort of deviations is Match finding? If it is not impressed by the gamut achievable with a particular printer+paper combination that is understandable, but the alert implies a problem with the profile. Id like to know what is going on, or what Match thinks is going on.

A couple of comments might help here:

Matte paper is, by its very nature, bumpy and scatters light more than glossy paper. This can cause a number of things to happen when measuring.

  1. You may not get very good repeatability. If you measure the target multiple times you may find that there is more variation in the measurements than if you did the same thing with glossy paper.

  2. The effects are more pronounced in the darker shadow colors. Light scattering on a bumpy surface tends to throw light around that has yet to be filtered or absorbed by colorants on the paper. This causes dark colors to appear lighter to the eye and instruments. A polarizing filter can be effective in cutting this down as scattered light tends to be polarized and can be filtered out. Unfortunately there’s only on mainstream instrument that has a Pol filter: the Spectrolino/Spectroscan that we use for our ColorValet profiling service. You also can’t just get a polarized piece of plastic or glass to read through as the calibration will fail.

  3. Finally, the Match software “expects” colors to behave within some limits and will warn you if your readings fall outside of those limits. Strictly speaking, there should be no expectations of how a printer behaves, just samples. In practice they can put limits on it to help inexperienced users out. They have been requested to change the internal limits in Match as too many people are getting too many warnings and starting to just ignore them, which defeats their purpose.

  4. You can take multiple readings, pull the data into Excel & average it & then import it back into Match if you want. ProfileMaker’s MeasureTool will also do this. This can help get around some of the scattering issues…

hope this helps




I was concerned about the warnings so I contacted Gretag Macbeth support.

I got a VERY prompt response to the question -as follows:

"I1 Match is validating the profile and give a warning depending mainly on linearity.
As some printers may not be fully linear either by generic linearity settings (by paper type selecting), either By not correct customized linearization or if having a thrid party printer paper where you may select either an option like “other substrates” or selecting closest papertype of the original printer manufacturer.

Anyhow the profile in 99% will do a good job, the warning is there if you still could influence some parameter such as calibration.

(Another) reason could as well be that manually you might have done some miss measurements where then you may remeasure."

All the profiles I have generated with this warning have been OK when tested, but I have seen VERY non-linear responses from Epson printers (which is why we need the profiles!)