CHROMiX

Disabling color correction for target

I am attempting to follow to the letter the guidance provided here http://www.colorwiki.com/wiki/Maxwell_ColorValet_Pro#The_yellow_square_test regarding validating that nothing in the print path is influencing the colors going to the printer. I understand the concept that we want our targets to flow through with no correction.

Where I am stumbling is on actually making this reality. First, I should comment that the instructions in the ColorShuttle popup helper window do not apply for my combination of operating system and printer driver. Those tabs and choices simply do not exist. :slight_smile:

I am using OS X 10.5 with a Canon iPF8100 and the regular 8-bit operating system CUPS driver.

What I do have are two separate interface tabs, one under the title “Color Matching” that presents two radio button choices: ColorSync or Vendor Matching. If the former is selected, an additional dropdown to select a profile is displayed, and it defaults to Automatic and displays “sRGB v3.0 (Canon)”. If I choose “Vendor Matching”, no dropdowns are available, and the second dialog box which I will describe in the next paragraph becomes available in a different portion of the print interface.

The second screen is in one of the Canon-specific subtabs labeled “Main.” Under Main there is a Color Mode -> Set button that brings up a very promising looking dialog that includes a Matching tab. Under this is a list of choices among which one must be chosen. The choices are:

a) Driver Matching Mode (Kyuanos)
b) ICC Matching Mode
c) No Correction

Obviously it looks like © is a winner, but when I print with this option I see plenty of other ink colorsbeing laid down in that pure yellow square.

I have also tried selected ColorSync in the first dialog, which disables this second Matching option. I see a different pattern of colored dots, but dots nonetheless amidst the sea of yellow ink.

I am left wondering if anyone has found a way to get color correction completely off with this printer.

[Side note: Interestingly, I did another target print on a different type of paper and selected a different Media Type in the Canon driver. (Media types influence head spacing, vacuum pressure, and total ink lay-down, among perhaps many other undocumented things, on this printer series.) I see a very different dot pattern on that paper target using the same supposedly-uncorrected settings. This makes me wonder if perhaps the Canon driver path always enforces some ink/color mapping management no matter what is set in terms of profiles and ColorSync.]

I will cross-post this to the user-supported Canon IPF Wiki and talk to Canon Tech Support this week to see if they have any ideas, but I thought I would post here first.

Thanks for reading!

Ben

I had one more thought after I posted the description above.

The iPF series has a built-in linearization/calibration routine that is supported directly on the hardware. This routine prints a bunch of color patches and strips to plain paper and then, presumably, uses some kind of sensor to relinearize the device. It does NOT generate profiles, but merely stores a calibration factor to the printer.

This is a very useful feature because it allows any shared profiles, including canned ones from Canon, to work well across large swaths of printers installed in different environments with varying printhead performance.

I am now wondering if maybe that linearization process is the “unavoidable step” in the color path that makes the pure yellow square, not.

If this is the case, I am not worried about requesting profiles in this scenario. It just means my profile would be reusable by others (using the ColorValet Pool feature) with a higher degree of fidelity anyway.

I am interested in what the Chromix experts think about this.

Ben

No worries. And, I’m making progress so I probably should have posted later rather than earlier.

I figured out how to turn off the hardware calibration in the printer. This now results in a pure yellow patch in that top left corner. Ostensibly, that seems like exactly what I want – and I can certainly do all my profiles and leave the printer calibration off.

However, this seems like it has two disadvantages ignoring the Color Pool one:

  1. The Canon IPF Wiki suggests that profiling on top of the linearized printer will make it easier to keep the profiles accurate over time as the print heads age, get replaced, etc. There’s no data there to back that up; it’s just stated as assumptive fact.

  2. The linearized targets look MUCH different. They are much smoother, the ramps are better, there is less blocking in the dark areas, and it is pretty easy to see the “intended” color ramps through each section of patches. Some of this is also presumably because I selected a Media Type that others have tested and found to lay down the optimal amount of ink for the darkest tones on this particular paper.

On the other hand, the pure targets printed with the linearization bypass have much more saturated, intense everything, especially in the pure yellows. The raw yellow is just qualitatively much more vivid, and I don’t exactly have super color-critical eyes or years of experience looking at color patches. Sadly, most of the rest of the color ramps seem really out of whack – the perfect candidate for profiling, I suppose.

So I am not sure which will work better – profiling the raw device or profiling the calibrated, less intense, “prettier” device.

I suppose I could submit two profiles and then judge the results myself, but I’d love to hear the expert opinion.

Thanks again.

Ben

p.s. Rick, your posting times are weird. It’s like it’s got the wrong time zone for your posts, and so they’re inserted ABOVE mine even though they must be chronologically after. It makes the thread hard to read because it’s “backwards.”

Well, most modern inkjets do pretty well without a linearization process. Linearization is most useful when you are dealing with an unusual paper type or surface that will not accept ink in the usual way. In that case, you would expect to see some areas of the gamut not well represented in a non-linearized target. For example the target might be overall very dark, and suddenly ramp up to the bright patches, or the other way around: A rather light target will have few shadow patches. If you’re seeing something like this then you should consider using the linearized version.

Generally linearization is not necessary if you are running through a printer driver - as the media settings handle a lot of the linearization and ink limiting. Now, you say your linearized targets are smoother, with better ramps. I guess the question is: Do the non-linearized targets do the job of sampling a good range of the colors available to the printer? As long as there are no obvious gaps, or sudden changes of color in the middle of a pattern, you’re probably in good shape to use the non-linearized ones. Especially since, as you say, the purpose of profiling is to take raw behavior and define it and make it look pretty.

Be aware that if you’re used to seeing our target printed with color management on, then it might look rather dark with CM off by comparison. That is normal. Our target also tends to sample shadows a bit more because that is a critical area to get right for many people.

Thanks Pat.

Another Canon owner ran some tests on his printer and is seeing different results that call into question my entire analysis above.

Is there any way you can send me the raw target image, or a piece of it with that yellow corner, so that I can try printing it in Photoshop first? I want to eliminate ColorShuttle from the mix for a moment while I try to isolate what part of the print flow is making the extra color changes. (My fellow Canonite found that he got perfect yellows even with the printer linearization turned on.)

Ben

Sure thing.

I will email you a target with a similar yellow patch test.