Changing Canon OEM colors with Hobbicolors UW-8 * ONE BY ONE * scans

[jonalava]

From my experience, you can get a better out-of-the-box color rendition (similar to OEM) form Image Specialists ink than from Hobbicolors. I don't know who makes Hobbicolors ink, but it is way too off for my taste. I know that profiling can correct that problem, but I like to have a good quality ink, that matches OEM right away. Maybe you could give it a try.

 

[Tin Ho]

Interesting test. I downloaded your scans and analyzed them by Photoshops. I merely wanted to see the effect of changing one ink cartridge at a time. But I saw some problem that I could not explain.

I used Photoshop Color Sampler Tool and chose a spot on the upper right corner of the brown colored block on the center top of the scans to measure the RGB values of the spot. The Color Sampler Tool read the RGB values of (211,127.65) on the 3rd scan and (209,138,66) on the 4th scan. The difference is in the value of G component, not in R.

Changing PC should have an impact only to the value of R. But the scans showed an impact to G. I am afraid this test is not very conclusive. I can not explain why the result is as such.

From the 3rd to the 4th scan, where the most significant color shift occurred, you said you replaced only OEM PC with Hobbicolors PC. This should only affect the R component of the RGB. If the Hobbicolors PC ink had any impact it would be the R value in the scan to either increase or decrease. I am not judging by the colors rather I am measuring the RGB values all over the scan to see the quantitative impact of the change of one single color of PC. PC is a light C(yan). R is the complementary color of C. So by changing PC there should be an impact of the quantitative value of R in the scan.

From the 1st scan to the 3rd the RGB values were very close. Each was within single digit with each others. From the 3rd to the 4th the G jumped significantly. For this change I would think it was either M or PM ink being replaced.

Well, do you have another set of empty cartridges to test another set of ink from another source? You will be equally surprised, if not shocked, to find a similar result but different color shifts. I have done such test on an i960 long time ago with Canon BCI ink cartridges. The inks I tested included some of the most expensive ones to many Chinese made that were available on ebay.

My conclusion was, it is impossible to match Canon OEM colors. None of the ink I tested was perfect. I gave up my quest for finding an ink set that really resembles OEM inks. I realized that it was silly in fact hoping that some vendors would have inks that really matched Canon closely so that I could avoid the expensive and sophisticated Color Management route to achieve the goal of MATCHING CANON's COLORS.

Why is it silly? Very simple. Why do I want to match exactly with Canon's colors? My prints of portrait always have a reddish tint on the skin tone. I hate that but it is always there. In addition, Canon's colors change when I change papers. This is the same as if the ink was changed. They both cause the colors to change. Unless I use the only few choices of Canon photo papers my colors will always change. The effect of changing papers is not much dfferent from changing inks.

I am not exactly surprised by your test result although I can not explain why my Photoshop Color Sampler Tool gave me conflicting readings. The test should give you an idea that you will never (I could not) find an ink set that will resemble Canon colors closely. This explains why Canon has not sued any ink manufacturers for copying their formulation of their ink yet. I am sure Canon's inks have many layers of protection from many patents. All the inks are clearly formulated differently. No 3rd party is sued for patent infringement.

If your graphic iamges require exactly accurate color reproduction that closely resembles Canon OEM colors the only way to achieve that is to custom profile the ink, printer and the paper as a whole. It is silly though. I would profile my ink, printer and paper for the best photo colors that I like. The profile will not make my printer produce colors that resemble Canon's.

 

[Tin Ho]

Marcel, there are significant color changes from the 4th to the 5th scans. But your note said they had the same Hobbicolors PM, Y and PC with the rest being OEM. Is it a mistake? Your last scan had BK, C and M changed at the same time. And the gray became muddy. BK should have no effect on colors. C and M change at the same time is equivalent to changing Y. It should have impact on blue or yellow. But the gray turned muddy. I can not make sense out of this. Interesting...

When a whole ink set is changed it will be difficult to analyze or measure the quantitative change caused by each color of the ink set. When the ink is changed one color at a time you can find the impact quantitatively from each individual color of ink. You can make some conclusions of the ink set tested. Be aware that the analysis depends on the scanner's color accuracy as well. Not every camera, scanner, printer can capture/render all colors correctly.

The test I did was too long ago so it is meaningless to disclose the brand names of the inks I tested because they probably are all different today. But the brand names included the top and under dogs. I will not name them as it may be interpreted as being biased. Some of them already disappeared from the market and some exist today.

 

[marceltho]

It came never in my mind to match the Canon colors exactly, well.....the results with the OEM ink were very nice, but I knew I would switch to another ink, and in combination with using non Canon paper it was of no use to expect that also. I did chose for Hobbicolors after a search on this forum, that's the only reason.
I was planning to switch all OEM colors at the same time, so I could see the difference immidiately, and try to make the necessary changes, but only PM was getting empty, and I thought it would be a waste to purge all that good OEM ink, so I decided to change the carts one by one, and keep track of the changes, and share that with the forum.
I do not comment so much on the changes, as it is not really important for me, only the end-result is, as that end - result will come back over and over again, and that will be the result that I have to change.
I have a box full of empty purged Canon carts, but whatever test I would do with an other ink, using different paper will never give an OEM result, and at the end I will always need to adjust, but that is OK.

I did the same with looking in Photoshop to measure color changes, but I did not comment on it, as it is each time a temporary phase. But I always used the same computer, same scanner, same scanner settings, and same printer & settings, and same paper.

I am not un-happy with the end result, I think I can make changes myself for my brochures, and when I have made my choice for good photo paper, I will see again. I am not that far yet. Probably next week, or this weekend I will do a follow-up on the lastest changes, just check on the link.

Marcel.

 

[marceltho]

Tin Ho

But your note said they had the same Hobbicolors PM, Y and PC with the rest being OEM. Is it a mistake?

No, it is not a mistake, I kept a log and it was exactly as I stated under the pictures.
I scanned with an Epson Perfection 4180 Photo scanner, for me, a real good scanner. I admit that the scans came out a tiny bit darker than reality, but I decided not to make any change in the settings, as I did not know what effect that would have on measuring color changes, perhaps a good decision, perhaps a bad one, I don't know.

Let's not make too much of a deal out of this, this post was created as I was curious what would happen, think it is not wise to try to solve the problems that occur during the steps in between, as nobody would do go this road normally.

 

[Tin Ho]

Marcel, it's an interesting test which allowed you to witness the wonders of a 3rd party ink set when you switch one color at a time. When I did the test for the very first ink set I was a bit upset quickly. After testing a few more I realized that there was no reason to be surprised. None was close.

Not all inks are created equally. You may want to compare if the ink has a better reputation of not getting your print head clogged among other things. No clogging is absolutely the number one consideration. The next may be the gamut and color accuracy of the ink. Fading characteristic is the 3rd consideration. Cost is another to be compared. The next may be the vendor's customer support (if that's important to you).

 

[Tin Ho]

No, it is not a mistake,

Why are there color differences if both prints (4th and 5th scans) were printed with the same set of inks (Hobbicolors PM, Y, PC and the rest were OEM)? There is probably a typo or a line of text is missing or something there.

 

[marceltho]

There is probably a typo or a line of text is missing or something there.

No there is no typo, my best guess is, that during those 25 prints, the ratio between OEM PC and Hobbicolors PC ( last color that was changed ) in the cart, has changed, and that gave the change in color.

 

[Tin Ho]

Marcle, look at your note at the bottom of your 4th and 5th scan. They both said:

Above :

* Hobbocolors UW-8 Photo Magenta
* Hobbicolors UW-8 Yellow
* Hobbicolors UW-8 Phote Cyan
* Rest still Canon OEM colors

But the colors on the 4th and the 5th are different. I don't think you meant to say they were printed with the same inks. Or did I really misunderstand your notes?

I am trying to analyze what change of ink causes what impact it has on colors. I need to know what ink you changed from the 4th to the 5th.

 

[marceltho]

This was written on the Internet :

It was my plan to change all colors at once and having a fresh start with all Hobbicolors UW-8, but I decided to re-fill each cart that got empty and see what the changes will be.

When I started printing, all carts were the OEM Canon carts, filled with OEM Canon ink.
Each time a cart got empty, I refilled the cart ( with the German method ) with Hobbicolors ink. This means that, after the refill, there is a certain small amount of OEM ink, and a rather big amount of Hobbicolors ink.
Let's say it was 20% - 80% each time.

What happens ......

1. do they mix equally immidiately ? that would give a change in color, but all the time the same color change...
2. is the OEM ink not mixed equally immidiately and the OEM ink goes first out of the outlet, than the Hobbicolors ink....that would make a change in color during printing ( I think that is what happened )

Perhaps I turned over the first carts more during filling, and that gave a better mix, perhaps I waited less long before to put the PM cart in....and that gave a mix only during printing.....

It must be someting like that, because I am still getting the last color as on the pic, and I am sure there is NO DROP of any OEM Canon ink in the carts any more, as I purged all the used carts, and use a spare set of carts with pure Hobbicolors ink in them.

SO....PIC 4 and 5 ARE INDEED PRINTED WITHOUT ANY CHANGE FROM MY SIDE.....something must have happened inside the cart, and this change happened during 25 prints or so.