Using ArgyllCMS + Colormunki to produce excellent printer profiles

It's okay.

I've typed this and it worked:

targen -v -d2 -G -e8 -B8 -g128 -f576 10x15cm
printtarg -v -ii1 -a0.9 -A0.6 -T360 -m2 -p100x150 10x15cm
chartread -v -H -T0.4 10x15cm
colprof -v -qh -S AdobeRGB1998.icc -dpp -D"Tecco High Gloss PHG260" "10x15cm"
pause

But I've got 6 x 10x15...

Is that right?

Edit 1: After counting each partch, I only have 108 patches/sheet...
I also have a big white space on the left...
How can I center the patches?

Edit 2: If I make targen and printargs only one time, I can store the tiff files and use them everytime I want, without generating new tiffs.
I only have to make chartread and colprof to profile to generate the icc.
Am I right?

Okay, I've understood...

You used the -L parameter in printargs to remove the white space, and now I only have 4 sheets.

I think you used the -B parameter in chartread?

Just got my first profile with argyll.
Woaw.
It´s at the same level as the spyder3print, maybe a little bit more detailed because it´s less dark (it was a black and white picture).
The advantage is that it´s super fast compared to the spyder3print.
First I was affraid because of the size of the patch was tiny and It was the first time I used argyll, so, I was not aware about how to slide the colormunki.
My patch was scratched because I scanned it several times...
But finally, there's no problem even if Argyll told me that my scan was non monotonic.
I'm very happy of my colormunki.
Thanks for all.

I hope this thread will not die...

It's impressing what you guys are doing and I'm very thankful. I have got a Colormunki "Design" (the meter is the same as the Colormunki "Photo" but not the same software but works the same with Argyllcms). What bothers me is that it's an UV-cut instrument that I suppose will not get correct readings on papers without optical brighteners (OBA). I try to mainly use papers without OBA:s. So I got a very good deal on an EyeOne Pro that's supposed to be without UV-filter. If I profile papers with a lot of OBA it's anyway possible to compensate for that in software (Argyllcms). I do usually use Linux but because there is a lot of problems using two screens with my computer (Lenovo W520 laptop with NVIDIA Optimus) I've decided to use Win 7 for printing. Linux is faster and more stable in use but what to do... For me it's much easier to work in the terminal in Linux than in the Windows command prompt. After a lot of trouble I did manage to make the targets in Argyll under Windows (it was much easier with Linux) so I hope I later on will manage to make the profiles with Argyll if I follow your excellent instructions!

Another reason i did make a bid on the EyeOne Pro on Ebay was that Serhat Abaci has showed that's it possible to get 870 patches on one single A4 sheet that's possible to read with the EyeOne Pro. What I can see he hasn't written back what the quality was with the profiles he made but I suppose they were good!?

The meter I bought on eBay (I haven't got it yet) is without any software. Most meters sold on eBay seems to be without any useful software. Also if you buy the more affordable EyeOne Pro Basic it's without software to profile printers. Then comes ArgyllCMS (andDispcal GUI) to the rescue! If you buy one of the very affordable Dell Ultrasharp monitors (I've got the U3014) you have the possibility to use the "Dell Calibration Solution". With this program it's possible to calibrate the monitor with the X-rite i1 Display Pro (a colorimeter). It gives good results calibrating the monitor. With the program it is also possible to make printer profiles but not with the Display Pro colorimeter. In that case I think you need the EyeOne Pro Spectrophotometer that's supposed to be supported. I don't know if it's possible to use this device also for calibrating displays in this program (I will report back if it's possible). This program doesn't support the Colormunki though because of restrictions from X-rite that has licensed the program. EDIT: No it's not possible to save the profiles becdause of license restrictions!

What's interesting is that when profiling papers with the Dell program to get 800 patches you have to use two A4 sheets (compared to Serhat Abacis 870 patches on one A4-sheet with ArgyllCMS). Anyway if you have or buy one of the Dell Ultrasharp monitors and don't want to learn how to use ArgyllCMS then you have a profiling solution with the EyeOne Pro (I suppose also the Basic version vill work). EDIT: No this is not possible!

Sorry for this rant but I just wanted to give back some of my recent learned experiencies!
Per

I've just got the EyeOne Pro. It says nothing about UV-cut so it must be without UV-filter. It came with no software what so ever. It installed automatically in Win7. I have downloaded the diagnostic test from X-rite and it passes the test. This instrument must never have been used as the lamp has been burning about 200 seconds! It has no software licenses so I have to buy licenses from X-rite but I'm not going to do that. I opened the Dell Ultrasharp Calibration Solution and the instrument is recognized and it seems to be possible to profile displays, projectors, scanners and printers with the program. I will later test to calibrate the Dell U3014 and see if it is as good as the i1 Display Pro Colorimeter. If it is I can sell that instrument. EDIT: No I have to keep the i1 Pro Display to make calibrations in hardware LUT with my Dell monitor because of license restrictions!

The i1 Pro Spectrophotometer comes with a ruler for reading patches. It's very nice but seems to have a problem that it can't read full length strips as the one Serhat Abaci made with 870 patches. I have installed Argyllcms following the instructions on its website in a separate directory. The learning curve is a little steep in the beginning but then you work with the command prompt in the "bin" directory. You have to navigate to it. You can save the commands in a text editor and then just copy them to the command prompt and also make eventual changes. By changing the argument "A4" to the size of the paper in mm and reverting width and height I've managed to get the strips in the other direction (there are more and shorter strips to read) and they are still 870! I've also tried to make patches on A5 and I managed to get 840 patches on two A5. That's actually no advantage compared to the former but I will try to first make a profile from 420 patches on A5 and then use this profile as first calibration and then use the 870 patches. If this works well it is possible to use one A5 and one A4 for a calibration. To get the instrument to work with Argyllcms you must use the Argyll driver. So if you want to profile your Dell display with the Dell program (I think you want to do that as it's a hardware calibration) you must use the X-rite driver. I don't think Dispcal GUI with Argyllcms can do a hardware calibration on these monitors. EDIT: Because of license restrictions is it not possible to do a calibration with i1 Pro Spectrometer wit hthe Dell program!

I have a reservation: I have not yet tried to use these small patch sizes. I have calculated the patch witdth to be about 6.3mm. The diameter of the light source according to X-rite is 6mm and the measurement aperture is 4.5mm. You have to position the meter carefully to read a single strip. I will report back if it works! It could be that this is to stretch things a bit to far! The i1 Pro2 can read smaller patch sizes than the original version according to X-rite. The ruler that comes with the i1 Pro seems to be rather precise so it could be possible. I think it's impossible without the ruler. EDIT: I don't think I will experiment with smaller patch sizes than 7x10mm that I got with my 720 patches chart. It gave good results!

EDIT: If you have one of the Dell Ultrasharp monitors you will need the i1 Display Pro to make a hardware calibration (LUT). It could be a better idea to buy a used i1 Pro spectrophotometer (or maybe even a new "Basic") instead of buying a Munki for paper profiling. Try to get an i1 Pro without UV-cut!
Edit:
I found this very helpful link: http://www.ericsantiago.com/eric_sa...9600-with-an-x-rite-i1-pro-and-argyllcms.html It's very good instructions how to start up with Argyllcms!
He has a more conservative approach and made 700 patches on a single Letter size paper. I made a target with 700 patches but for A4 instead. The formats Letter and A4 differs a little and also this time it can be a little hard to read A4 lengthwise with the ruler (it may be possible though). Also this time it's easy to replace the argument with 297x210 and that works well in the ruler. The patches are 6,5x8,3mm.

I have printed the targets on a laser printer to see how they work in the ruler and I think I will take the "safe" approach and start with this 700 patches target with the strips over the short side of the paper. I may experiment later with A5 targets with maybe a little more than 300 patches to make a first calibration.

Edit: I wanted the patches a little bigger and by lessen the margins to 2mm and a0.9 I get 720 patches with the size 7x10mm.

Per

Well this is what have come out of this so far
The Dell calibration solution does NOT support the i1 Pro Spectrophotometer. It lets you do both printer and display calibrations but then says: "Profile calibration is not licenced, please connect the right dongle". This is simply idiotic from Dell and X-rite! Why support an instrument and then not let you use the profile. This comes as an emabarrasment AFTER you have gone through the lengthy and costly process of trying to calibrate your papers and display! So I have to keep my i1 Display Pro colorimeter to be able to make a hardware LUT with my Dell monitor.

The good news is that the profile I made with 720 patches with ArgyllCMS and i1 Pro turned out to be at least equal (actually I think a little better) for the SIHL Lustre Photo Paper 300 compared to the manufacturers profile for the same paper. As this is a paper with a lot of OBA I used the FWA-compensation in Argyll (-f D50M2) when I made the profile as the i1 Pro I have has no UV-filter. Profile check gives this: peak err = 1.38, avg err = 0.38, RMS = 0.45. I think this is a very good result! So I will sell my Munki but keep the i1 Display Pro until Dell changes their idiotic licence restrictions (or is it X-Rite?). I will edit my above postings so they are correct! Another interesting link is this one: http://www.ludd.luth.se/~torger/photography/argyll-print.html
It's possible for example to use the paper manufacturers profile as a preconditioning.icc as a startpoint. I think this could be useful even when you use third party inks as they may not be too dissimilar from OEM ink. When my cartridges with OEM ink are empty I will profile the Inktec replacement.

I'm really happy I've found the ArgyllCMS program. It's very important to follow the instructions in the documentation but then it's very easy to use! It's fantastic and I will of course donate to it!

Per

This is the chart that I ended with. I have used the template from the first link above (ericsantiago) but changed some things to get 720 patches and the format from letter to 297x210. The first attempt with chartread gave 10-15 faults during reading but the second only two. It's important to have a steady and consistent movement when you read the patches. It's selflearning because Argyll protests when anything goes wrong and it's easy to repeat the step. If someone is interested I can post the commands that I've used and also some instructions how to install ArgyllCMS.

Per

If anyone is interested in why Argyll comes up with the comment non-monotonic, see below.

I've never seen it not say that about my profiles.

Hi,
Really funny looking profiles! I haven't seen anything like that when I profile with Argyll.

I'm trying out some papers and compare mfg:s profiles with my own profiles. I think my own profiles with the 720 patches chart look better than the manufacturers profiles. I dont know if they are more correct but I think they look better. Deeper colours and more life. The difference is small but possible to see. My wife can see the same thing as I do (blind test). I've made my own profiles for Canson Platine Fibre Rag and Canson Baryta Photographique. Of these I like the Platine best. They are very similar but I think the Platine has more "depth" the motif stays out more from the paper. I've now also profiled the Museo Silver Rag in the Letter format (the test pack came in that format). I managed to do a chart with 720 (actually 722) patches that I could read without misread. Also in this case I did change the orientation to make it easier to read the strips. I think I like the Silver Rag even more than the Platine but they are similar. The Silver Rag is more "neutral" but the Platine may be more "3D".

I'm now also trying out matte papers. After reading a lot of comments on the net about paper choices I thought the way to go was with papers that used the PK-ink (glossy or satin/pearl). I printed the Museo Portfolio Rag, the Museo Max and the Canson Rag Photographique all matte papers with the manufacturers profiles. But wow! I do really like these matte surfaces. The Museo papers has the deepest blacks and are more contrasty. The Canson has the smoothest strucure that I like. I am making profiles of the Museo Portfolio and the Canson. The charts are now drying over night and I will make profiles tomorrow. I like the US formats (Letter and 17"x22") better as they fit my 4/3 cameras proportions better. When I make my final profiles I will try to make the profiles in two steps. First a "preprofile" on half letter or A5 and then use that profile on the full Letter or A4 chart.

I have been using a lot of ink in the printer during the trials so soon it's time to change to the third party inks.

Per

All your colour profiling is certainly fascinating but excuse my ignorance for a moment, if you intend changing over to 3rd party inks then why in heavens earth are you profiling OEM ink when you have no intention to continue using them.
Isn’t that a total waste of time, talent and effort not to mention good OEM inks..?