Argyll B&W Profiling

[RogerB]

I just came across this http://www.russellcottrell.com/photo/ArgyllColorPatches.htm which gives a .ti1 file for an Argyll chart having a very limited colour gamut - all of the 404 patches are neutral or nearly neutral. I thought I would try it to see how it performed for B&W printing.

I generated a .ti3 and TIFF file using the .ti1 file from the link above. Here's what the TIFF looks llike. This is a single A4 intended for the i1Pro but probalby usable with the ColorMunki.

Using this chart I generated a profile for my 3880 on Harman Gloss Baryta, my normal B&W paper. Comparing the B&W response in Gamutvision suggests that the 404-patch small gamut profile is slightly better than the normal 1058-patch profile. The B&W response of the 3880 is very good anyway so I wouldn't expect to see a big difference. These two screen shots show an enlarged portion of the curve at the shadow end. The difference is subtle but quite obvious.

I would be very interested to see if the 404 grey chart works for anyone that has more problems with B&W response. I'm happy to share my .ti2 and TIFF files if anyone cares to try it. Any takers?

 

[Emulator]

Interesting experiment. I found it was difficult to achieve regular, consistent results when I experimented with 957 patches total of which 335 were greyscale. It seemed to give slightly better dark end greyscale on one occasion, but the next iteration of patches might not bring as good results. I haven't yet found out why.

This is an example of a set of patches and the resulting B&W density response.



I see I need to clean the scanner glass, it has added a vertical line in column A.

 

[VISDATA]

hello,I used the TI1 file linked by RogerB, and then generated TI2 printing, but an error message occurred when generating the ICC configuration file, and the ICC file could not be generated. Can anyone tell me the reason? Can you help me?ti file is placed in the compressed package

 

[nertog]

hello,I used the TI1 file linked by RogerB, and then generated TI2 printing, but an error message occurred when generating the ICC configuration file, and the ICC file could not be generated. Can anyone tell me the reason? Can you help me?ti file is placed in the compressed package

Argyll needs a reference white, and the attached patch set has a maximum L=98 instead of 100. You can either add it manually or, I believe, use the -f flag in Targen to limit the lab gamut.

 

[VISDATA]

Argyll needs a reference white, and the attached patch set has a maximum L=98 instead of 100. You can either add it manually or, I believe, use the -f flag in Targen to limit the lab gamut.

hi,nertog,I still don’t quite understand what you mean. Do you mean that the TI file in the link http://www.russellcottrell.com/photo/ArgyllColorPatches.htm does not define white? What does L=100 mean? I don’t. Use the targen command to generate the pattern, directly using the TI file in the web link.My English is translated by GOOGLE

 

[nertog]

hi,nertog,I still don’t quite understand what you mean. Do you mean that the TI file in the link http://www.russellcottrell.com/photo/ArgyllColorPatches.htm does not define white? What does L=100 mean? I don’t. Use the targen command to generate the pattern, directly using the TI file in the web link.My English is translated by GOOGLE

What I meant was that Argyll expects at least 1 white patch (Lab = 100,0,0). The .ti1 file you use does not contain such a patch. Maybe earlier Argyll versions were ok with that, but the current version is not.

Anyway, I am not sure why you would want to generate patches around an absolute neutral Lab-axis. When you are printing, your neutral greys are (nearly always) adjusted to the white point of your paper. This white point is never completely neutral.

I had a quick look at the Targen documentation and I would recommend to use the -n flag:

The -n parameter sets the number of patches in a set of neutral axis wedge steps. This uses the pre-conditioning profile, to lookup the assumed neutral axis device values. By default, no neutral axis values are generated. If you have a previous profile for a device as a reference, adding some neutral axis values can improve the neutral axis rendering of the subsequent profiles.

This generates neutral patches based on a previous ICC profile from your printer, ideally one using the same paper, inks and printer settings.

E.g.: Targen -v -d2 -c preconditioning.icm -G -n128 -f0 Greys

I attached the resulting output. Notice the greys are non-neutral on-screen but neutral for the specific printer-paper combination used.

 

[VISDATA]

nertog, thank you for your answer. I want to use ARGYLL to generate a special black and white configuration file through the link on the first floor. Currently I use I1studio software to generate a black and white configuration file. The I1studio configuration file needs to be generated twice to generate a black and white configuration file. It cannot be Remote friends generate configurations, so it would be nice if you could use ARGYLL to generate specialized black and white configuration files
Your TARGEN F is 0. I don’t understand why it is defined like this. Isn’t the F value the pattern value we want to define? For example, f is 100. In addition, is preconditioning.icm a pre-set color configuration file? I want to create a black and white one from scratch. Configuration file, not optimized and created based on any existing configuration

 

[nertog]

nertog, thank you for your answer. I want to use ARGYLL to generate a special black and white configuration file through the link on the first floor. Currently I use I1studio software to generate a black and white configuration file. The I1studio configuration file needs to be generated twice to generate a black and white configuration file. It cannot be Remote friends generate configurations, so it would be nice if you could use ARGYLL to generate specialized black and white configuration files
Your TARGEN F is 0. I don’t understand why it is defined like this. Isn’t the F value the pattern value we want to define? For example, f is 100. In addition, is preconditioning.icm a pre-set color configuration file? I want to create a black and white one from scratch. Configuration file, not optimized and created based on any existing configuration

To get accurate patches for monochrome profiles, you should use a preconditioning profile. This would be one of your previous icc profiles of the same printer-ink-paper combination. As I wrote before, printing monochrome images does not equal printing a completely neutral grey scale. Your white point has to be adjusted to the white point of the paper + lighting conditions or things will look wrong once printed.

If you want a more "general" b&w target, you could try to use the -g flag and this will generate patches with R=G=B values only.

A third option you can try is to modify the .ti1 file from the original post and add the needed white patch. I attached a version with an extra 9 pure white and 9 pure black patches (just extended the original Lab series).

I would be interested in seeing a comparison between full RGB, the -n + preconditioning profile, the -g and finally the small gamut patchset proposed above.

And finally, you only need the -f flag if you want to generate (extra) random color patches.

Let us know the outcome!

 

[VISDATA]

To get accurate patches for monochrome profiles, you should use a preconditioning profile. This would be one of your previous icc profiles of the same printer-ink-paper combination. As I wrote before, printing monochrome images does not equal printing a completely neutral grey scale. Your white point has to be adjusted to the white point of the paper + lighting conditions or things will look wrong once printed.

If you want a more "general" b&w target, you could try to use the -g flag and this will generate patches with R=G=B values only.

A third option you can try is to modify the .ti1 file from the original post and add the needed white patch. I attached a version with an extra 9 pure white and 9 pure black patches (just extended the original Lab series).

I would be interested in seeing a comparison between full RGB, the -n + preconditioning profile, the -g and finally the small gamut patchset proposed above.

And finally, you only need the -f flag if you want to generate (extra) random color patches.

Let us know the outcome!

Hello, nertog, I used the test.ti in your attachment, then printed it, and generated a black and white configuration file. Then I printed with the black and white configuration file generated by argyll, and compared it with the configuration file printing generated by I1 studio. , I prefer the configuration file of i1 studio, which feels closer to the color of the original image, while the configuration file generated by argull looks greenish to my eyes when printed

 

[Ink stained Fingers]

I'm only reading this thread on the side since I'm not using ArgyllCMS for profiling nor do I focus much on the specifics of B/W printing, but I nevertheless have some questions.
@VISDATA restarted this thread with posting #3 , I don't see any comment which printer is actually the one which should be profiled for B/W printing- and on which type of paper. There is only a reference to a L805 and L801 in the personal data, the L805 is a 6 color dye ink printer - without a gray. I see principal limitations to get very good B/W prints with such a dye ink set. B/W prints may look good enough for most users - but you won't be able to get away with hue shifts of the blacks and grays under different light conditions, and it would get more complicated if your paper carries optical brighteners which cause a different paper white point if your viewing light carries some UV - like from halogen bulbs or small neon tubes.
It's a pity that the L805 driver does not support the adanced B/W mode of other higher end Epson printers which could help to get away with a greenish cast instead of trying to profile the printer.