A basic guide (see post #1) to setting up ARGYLL CMS profiling on your computer

[RogerB]

I looked closely, but I can't say I identified any diagonally split patches, is it possible that the uploading process has a limited colour step range? Or is it me!

Interesting experiment, the results are better than the standard in general use.

Exactly! I'm sure it's nothing to do with the uploading process. The fact is that the errors here are, for the most part, simply not detectable to the eye. Here's and enlarged portion of the patch with the biggest error, which is detectable.



And just as an example this is how the comparison looks with big errors.

Incidentally the CAP stands for Create Argyll Profile, so I could add it to post #1 as CAP1058. Is it in its finished form or are you still working on it?

The command lines worked well for me. The batch file is also working well, although it is (of course) a work in progress. For example, I am just adding a option to select the illuminant which could be useful for dye inks that show poor colour constancy. So, if you want to add anything go ahead.

 

[Emulator]

Yes, when you point it out it is visible and the big errors demonstrate the effect well, which program are you using? There is a somewhat similar display in Spyder, but buried in the profiling software.

 

[RogerB]

I use MeasureTool which is part of the x-Rite ProfileMaker suite. You can download it (free!) from
http://www.xrite.com/product_overview.aspx?ID=1225&Action=support&SoftwareID=931
When you run the installation program you can select just MeasureTool to install. The main modules need a dongle to work, but MeasureTool works as a stand-alone program. Unfortunately it doesn't support CM for measurement but the comparison tool works well with .ti2 and .ti3 files. Very useful when you're trying to assess the repeatability of chart readings.

 

[Emulator]

I downloaded the whole package and installed it, so I will have to remove the bulk of it and retain Measure Tool. With regard to the .bat file, I will leave copying the file to post #1 until you are finished, anyone who wants to try it out can work from post #128 for the time being. Thanks for the information.

 

[RogerB]

Just for completeness I thought I would look at the errors for the Low quality setting in profgen. Not surprisingly, the accuracy is not as good, although I suspect that many people would be perfectly happy with this accuracy in "real" prints.

 

[RogerB]

I've had enough of batch file writing for now, so the attached is my final effort. It now has two additions:
a) An option to specify a different illuminant when generating the profile. If you use dye inks and you know that your prints will be viewed under tungsten illumination then selecting "A" as the illuminant should get rid of any red cast in B&W prints.
b) A facility to compare two .ti3 measurement files and report the differences, patch by patch. Useful to check the repeatability of chart readings. Can also be used, with a suitable test image, to assess the accuracy of a profile from a test print.
As before, I'm not claiming any great quality in the batch file, but it works for me. If anyone cares to try it I would love to hear if it works for someone else!

 

[Emulator]

Roger, I have copied it across to post #1. Hopefully it has transferred OK, I see it has reset your views from 8 to zero, it is the first time I have tried this. Anyone please report if you have problems, or if it all works!

 

[RogerB]

If anyone is interested in running profile verification tests using my batch file ArgProf#2 I thought I should provide suitable image and reference files. I use a 288-patch chart which is a modification of a profiling target. It is modified so that all of the colours lie within the gamut of my Epson Pro3880 - no use trying to check accuracy for out-of-gamut colours!

The procedure is quite simple:
1. Print the image with your normal colour-managed workflow using the icc profile you want to verify.
(Use either Relative colorimetric or Absolute Colorimetric rendering.)
2. Allow the print to stabilise for the same length of time that you use for your profiling charts.
3. Measure the chart using chartread (option 1 in ArgProf#2) to generate the .ti3 file.
4. Use option 4 in ArgProf#2 to compare your .ti3 file with the attached 288_Reference.ti3. If you
specify the profile that you used, colverify will ignore any colours that are out-of-gamut for your
printer.



Edit: After posting this I realised that the image size was changed on uploading. The actual jpeg file and the .ti3 file are in the attached .zip file.

 

[pharmacist]

Wabi Sabi: Less is more....

Could someone try this set and report back? It is based on only 480 patches and the results are impressive, especially B&W on non-K3 printers

 

[Emulator]

Wabi Sabi: Less is more....

Could someone try this set and report back? It is based on only 480 patches and the results are impressive, especially B&W on non-K3 printers

I've downloaded the files and printed the tiff. I'll leave it to mature before reading it tomorrow.

The patch sizes look large as I've been playing with a 957 patch on A4, which I find I can read on an updated version of my reader (7mm square patches).