Printer profiles

[FMueller]

Not only do I want good printer profiles, I want to understand what's going on. There is an admittedly steep learning curve on this and I freely admit that I am on the early and steep part. Attached are a 2d and 3d plot of a printer profile I've made using i1 profiler. The 2d plot shows some jagged edges, the 3d plot is more of what I expected.

Why the jagged edges?

Thanks,

Fred

[VitOne]

I experienced some bad readings problems with i1Profiler and the i1Pro REV D I was using. Maybe this is the case (not sure about it). In many situations the irregular edges are normal. You can compare the profile from Ilford (here is the plot) with the one you made to have an idea of how the gamut should like.



If you open this file:  X:\ProgramData\X-Rite\i1Profiler\XRi1G2WorkflowSettings.ini  you can try to modify the parameter MeasurementTolerance=0.5 and other ones to improve the reliability of your readings. For what I know the default in earlier version of the software was 0.4, now they set it to 0.5 but maybe this is too high to get good results (sometimes the software will not find bad readings). My suggestion is to try 0.3 (you can go from 0 to 1) so that the software will ask you to re-read the patch more often if you do readings errors.

Hope this can help.

[smilem]

I experienced some bad readings problems with i1Profiler and the i1Pro REV D I was using.

I wonder how you knew you had reading problems, since i1Profiler does not allow to compare data of 2 separate mesurements like PM5 does?

[VitOne]

I simply got “very strangely shaped” gamut plots, with irregularities and different shape from the “custom” ones (and from the plot of a profile made with new readings).

Same paper, same settings, same printer, same inks, very similar conditions (where I work I have air conditioning and lighting is controlled). In red a profile made from “good readings”, in color a profile made from “bad readings”. Not very big differences, but you should see them. This was not also the worst result I could get with i1Pro + i1Profiler, I had some really bad results (I didn’t know how to move correctly the instrument).

[tony22]

VitOne, I find this very troubling. I have used i1Profiler to try and make a couple of printer profiles and they came out only good, not great. I tried using Argyll CMS but frankly I was not inclined to make so many tries at getting an acceptable result (too many variables to play with). I'm worried that if I keep trying with i1Profiler I will wind up eating through the same amount of paper.

Are there optimized setting for i1Profiler and i1Pro Rev D which will get the best possible readings? I know others have used other targets, but I'm not sure that is enough.

[VitOne]

I am not an expert so maybe you’ll want to wait for an answer from some other users. Anyway I’ll give you my 2 cents. I think that i1Profiler targets work very well and I don’t suggest you to use that software with other targets. i1Profiler requires his targets to work properly. You don’t need to print the target more than one time, because you can both re-read them and use them to generate a first profile that you can later improve with new patches (that you can generate in many different ways). I am very happy with the i1Profiler software, but I found the i1Pro REV D to be a not very reliable instrument if you can’t use it properly; you will need some time. My suggestion is to re-read the patches that you printed until you are happy with the result, maybe lowering the parameter I discussed in my previous post to a value of 0.3 (to start). You could also try to vary the size of the targets. I am not using the i1Pro to read the measurements, I just did some tests, so I really can’t suggest you something that will work for sure. Also consider that the reading should be done at the right speed, not too slow and not too fast.

What I would do is this:

1)   Re-read the patches trying to do it in the right way.
2)   Improve the profile I have using new patches.

I really think that every serious color management passionate should learn ArgyllCMS, it is very powerful, customizable, and you can learn a lot about color management while you use and study it… Just my 2 cents, as I told you before.

[FMueller]

Thanks for the input. I grabbed the Ilford profile that they distribute and compared it to the one I generated.

Based on graphing, the i1 profiler profile bested the Ilford profile in every way. Ilford canned profile is shown in single color red. The i1 profile is the grid in the 3d plot.

It prints well also, so I'll consider my query answered with all things proper.

Thanks