Please help ! Printer calibration disaster

[Anadrol]

Hi all,

As strange as it sounds, while I can calibrate easily dye based printers,
it's a total disaster with pigment based ones !
This happens with PLAIN PAPER, I don't have this issue on glossy paper.
And yes, I absolutely need to print on plain, I print on 100g plain paper.

I did contact Xrite but they aren't of any help !

I own:
Pixma PRO-1 with OEM inks.
Pixma PRO-10S with OEM inks.
Pixma PRO-100S with OEM inks.
Canon MG7150 with Chinese inks and a CISS.

I calibrate using Xrite i1Profiler with a i1 PRO 2 spectrophotometer.

I was very careful to disable color management when printing the test charts, with highest print quality.
I did select RGB when profiling, and made dual scans for each line, I did optimise for best quality.
I was careful to not have marks that occur when the paper becomes bumpy when wet,
and the print head scrapes the paper (you can see that on the green color of the PRO-10S result chart).
I did use 2033 patches because if I use 6000 patches, well i1Profiler crashes ! (does it happen to anyone else ?).
Of course, I did print the result charts with ICC profiles enabled.

By the way, please note that pigment inks make a much more evenly result on plain paper,
contrary to what I've read online, that you need dye inks for plain paper !

Here are the results:


Red is especially terrible on the PRO-10S, while it has a dedicated red ink cartridge !
Exactly the same happens with the PRO-1.

One can notice that Chinese inks have gorgeous colors on the MG7150, seem better for me than the PRO-100S with OEM inks !
I bought these inks for LESS THAN 5 USD PER LITER ! Yes per liter !
These guys deserve publicity http://www.aomya.com.cn/ perfect customer service from A to Z !

One very strange thing that I noticed, is that when I make the profile on glossy paper for the PRO-10S or the PRO-1;
then colors are great on plain paper ! But well, this shouldn't be the right way to do it.
The guy from Xrite doesn't believe me ! He insists that I should make the profile on plain paper if I intend to print on plain,
but this just doesn't work ! Even if that seems logical.

Please help me, I really don't have a clue there, I tried hours and hours to do it, making different profiles endlessly...

 

[stratman]

One very strange thing that I noticed, is that when I make the profile on glossy paper for the PRO-10S or the PRO-1; then colors are great on plain paper !

Why not use that profile with plain paper? If it works, don't fix it.

 

[Ink stained Fingers]

The guy from Xrite doesn't believe me ! He insists that I should make the profile on plain paper if I intend to print on plain,

That's the game with profiling - you print the target on that paper you want to print your pictures on finally. It's not clear from your description what really is going wrong, at which step of profiling. If the 600 sport target doesn't work, why not trying a 800-900 spot target then, or with 300 spots. The gamut on plain paper is small , vers small, I don't think you would see a difference to prints made with a profile based on 2000 spots, it just takes much more time. And if a profile made for another paper/surface works as well with plain paper, o.k.. that may be a visual jugdement, the colors may be balanced, but whether they are colorimetrically correct is another thing.
Pigment inks on plain paper, Epson is doing that since a long time with most of their printers which are not specifically made as photo printers - with 6 Claria dye colors. That 4 color pigment inkset by Epson is running under the Durabrite name.

5$ per litre of Chinese ink, that's about the price for their fast fading ink, it prints, it may even have a decent gamut, but no longevity, UV/ozone stability, it's your choice, and some people claim it will deteriorate the printhead faster than with OEM inks - I don't know, nobody ever did some comparative testing like that.

 

[Emulator]

Some of the patches on the Pro 10 are so far off there must be a major glitch. Looks as though you need to simplify your profiling process and reduce your profiling patch numbers into hundreds rather than thousands.

 

[Anadrol]

Thanks to all for your answers,

Why not use that profile with plain paper? If it works, don't fix it.

Yes, it's what I will do if I can't find a solution, but that really doesn't sound logical, I'd like to understand the flaw in the process !
And as "Ink stained Fingers" points out, "that may be a visual jugdement, the colors may be balanced, but whether they are colorimetrically correct is another thing.".

5$ per litre of Chinese ink, that's about the price for their fast fading ink, it prints, it may even have a decent gamut, but no longevity, UV/ozone stability, it's your choice, and some people claim it will deteriorate the printhead faster than with OEM inks - I don't know, nobody ever did some comparative testing like that.

Well I did print 4000 pages already since 2 years (I got actually the UV resistant version).
Please don't forget that OEM inks are manufactured in... China, really not hard for the Chinese to get and copy the exact OEM formula !
I've read many times that Chinese inks are terrible and shouldn't be used, but that's simply not the truth, they work flawlessly, at least from this supplier for the Canon MG7150.
No aging problem, and colors are gorgeous as you can see from my charts.

Some of the patches on the Pro 10 are so far off there must be a major glitch. Looks as though you need to simplify your profiling process and reduce your profiling patch numbers into hundreds rather than thousands.

Yes spot on, there is a major glitch indeed !
I will try again with 400 patches instead of 2033.

 

[Ink stained Fingers]

o.k., this thread is about profiling issues, not about Chinese ink in the first place. But just a word of caution - I got various 'UV resistant ' inks from China, they all are fast fade inks. Ink UV resistance was/is subject of other threads and lots of testing, there are only a few state of the art good and stable inks - the OEM inks of Canon, Epson, HP, and inks used by Epson, Fuijfilm, Noritsu in their Surelab/Drylab photofinishing equipment/printers. And China substitute ink for Epson Surelab printers runs at 150$/litre, and that's much more they even charge for typical pigment inks, they can fly around the moon but not do miracles. Most of all the other inks will print, will print fine, I don't have much of a concern with their performance, but they are not UV/ozone stable. The problem for Canon users is that they won' be able to use those inks from the Drylab printers, inks for piezo and bubble jet printing are not compatible, and I'm not aware of any other good source for Canon inks except tapping bigger cartridges from their larger format printers

I'm not sure yet to really understand what is happening with your profiling, are you starting new with this SW/HW, did it work originally o.k., but suddenly you get odd results, did you use another profiling package before ? Do you have profiles to compare with ? Until some, I guess principal issues are worked out you won't get better results with 2000 patches than with 300 patches. Those additional patches will give you some better accuracy at the borders and edges of the gamut, darker areas, regions of higher saturation, but the overall color balance gets fixed with even less patches. What is the rendering intent you are typically using for printout ? Is your monitor profiled/calibrated ? Are you getting a 'good' profile for one type of paper but not for the other ?
I'm using an I1io/i1pro with the previous software version by XRite - Profile Maker, so I'm not familiar with the i1Profiler s/w in detail, with default settings , options etc

 

[Anadrol]

Truth to be told, I didn't have my prints made with Chinese inks exposed to light a long time,
anyway that's not the issue there, I use OEM inks with the PRO-1, PRO-10S and PRO-100S.

Well what is baffling me, is that I successfully manage to make great profiles for dye based printers, and for some odd reason it doesn't work with pigment inks, the dye/pigments test charts look similar visually though !
This is not something new, I always had this problem, I didn't try with another hardware or software (well yes with Canon supplied profiling software I tried a bit 2 years ago, but results were not good, will try again now).
I tried with 2 different pigment based printers (PRO-1 and PRO-10S) and 2 different dye based printers (PRO-100S and Canon MG7150).

In my examples, I did use relative as the rendering intent to print ColorCheck charts, but also tried with saturation, perception, absolute.
I did specify to i1Profiler that the test charts were printed on plain or glossy paper accordingly.
Yes my monitor and scanner are calibrated, but that's not the issue there as I don't alter the files.

Yes I get a very good profile if I use glossy paper, even if I print after on plain ! The color balance is right this way, but this isn't really colorimetrically correct I guess (as correct as it can get on plain paper).
So this is not about pigment inks making weird results on plain paper, because a glossy based profile makes great colors on plain paper !
If I don't find a solution, I'll just use the glossy based profile, but I'd like to understand, Xrite hardware and software is very expensive, but they don't have a clue about it, they just tell me to make the profiles using the same source and destination paper, which is logical but just doesn't work !

I'm waiting to have more answers to this thread and I'll contact Xrite again to show them.

 

[Ink stained Fingers]

O.k., I understand your situation better now, you are unhappy with your profiles generated for pigment based printers but profiles for dye based printers are great. Forget the plain paper for a while, e.g. compare the profiles and print outcome for the same paper on both the dye and pigment ink printer. I think that somebody proficient with the use of Gamutvision may be able to analyze the differences between such profiles , Gamutvison as a program with the capability to show that in graphical format. I'm not actively using it at this time, I'm overall happy at this time with the outcome of Profilemaker for various dye and pigment ink based Epson printers, and a few other people. Or you run a trial on Gamutvision for yourself, you need some reading to understand its functionality and how you get what you are looking for. The existence of such program as a tool shows that not everything is easy and straightforward with profiles.

 

[The Hat]

@Anadrol, no it’s not you and I reckon you’re not going mad or doing thing wrong at all.

I have found the latest Pro model printers out of the Canon workshop all do very strange things when it comes to printing, especially forced borders and the use of black ink.

To get a satisfactory result when using certain type of paper even on gloss, I often have to set the Media to use odd setting to get the best set of colours.

The Pro 9500 is not as effected by this anomaly as the newer models are so I have a constant battle to remember which media setting I used the last time I printed some repeat prints that I’ve done.

So if you can't get the results that you want by using the correct Media setting then by all means use completely different ones, otherwise you could be there till hell freezes over trying to convince yourself you’re not going crazy.

There are a number of member who won’t agree with this idea and say hogwash and they are entitled to their opinions and rightly so, but as you have found out for yourself not everything works that way in reality, go with the flow and use what works for you that gives you the colours you expect.

P.S. I don’t use self made profiles only the ones that Canon loaded onto my system from every machine installed and I get surprising results working that way..

 

[stratman]

Have you tried a different brand of plain paper? Could optical brighteners or something else particular to the paper cause an issue?

Is it possible that one or more of the inks in the printer are changed up somehow resulting in a color shift the X-Rite cannot compensate?

Does your i1Pro 2 need to be factory re-calibrated? Example: http://www.colorforums.com/viewtopic.php?p=3548&sid=6caee61e7126eed1f36fc381d14cfe8e