relating CMY values from pattern print to RGB

In Canon's Print Studio Pro, you can print a "pattern print" in an attempt to find out where and how much the color is off in your prints. I've seen similar tools in other programs but can't recall where, right off the top of my head. It gives you the amount of change in each little image in C,M, & Y values. But I can't find anything to tell me what to do with that CMY info in order to change the color of my image so it'll print the way I want it to. Does anyone know?
Thanks

By the way, if you don't know what a pattern print is, it looks like this.

That kind of output in days of darkroom printing could be of value in adjusting the color filter values used in the enlarger to get a better balance compared to some starting value.

In the context of digital, programs offer Yellow control vis the WB, but not in color filter values. And you have Magenta (vs Green) control, BC but again the adjustments may not be in color filter density vajues.

They are not used for correcting the image, but for correcting the print, i.e adjusting the values at the printer end.

Are you talking about in the printer driver? The "manual color adjustment"?

ncjohn wrote in post #18030564
Are you talking about in the printer driver? The "manual color adjustment"?

Depends on the printer, and how you are driving it, but essentially yes.

You don't adjust the image to match the printer, you adjust the printer to match the image.
The test print shows you how far each value needs to be adjusted, since the printer works in CMYK, it gives you CMYK values.

You're diving into the great depths of printing, which has as much knowledge to be gained as do the great depths of photography.

Well, that's interesting. I never related that part of the printer driver to those old enlarger filters.
Thanks

I would suggest that you print a reference print used to establish a color managed workflow so that you can identify which device is not cooperating in the workflow chain. Reference prints are available here:

http://www.northlight-images.co.uk …le_pages/test_i​mages.html

for example.

A reference print is one in which the colors are established to be "correct" such that, if they look incorrect when produced by a specific device (your display, or your printer), you can characterize the device's issues and attempt to manage them.

I would suggest that your printer is probably "correct" and your display is not "correct." If this is the case, but you attempt to use the print driver to "correct" your print, you are essentially introducing an error upstream (the display) only to have to undo that error downstream (the printer-print driver). Also, if you are using the print driver to manage the color, it would appear that you are not using software color management ("let photoshop manage colors" or something similar) but letting the printer determine the output color - that is fine, but will probably require the kind of gymnastics you appear to be inviting with the test pattern print.

If your device is well-characterized and your printer is using a well-characterized process, then maybe you are viewing your prints under lighting that makes them appear "off" requiring some tweaking after the fact?

What I do not understand is that that pattern print appears to be using an ICC printer profile - maybe you can describe your workflow a little bit more so we can understand why you would need to use a print pattern to "correct" your print after passing the image through the printer/paper profile- or is the "correction" more for aesthetic reasons?

kirk

Hi Kirk. I'm borderline sick so I may be misunderstanding but it seems that you're saying that a well-color-managed workflow will always give you prints that are color correct. But if that were the case we wouldn't have softproofing.

In general, I'm happy with my display-to-print matching, but in the case of this specific image, the print comes out a tad red or magenta and when I try to softproof it, getting the subject right results in a cyan-y background. That's why I wanted to use the pattern print. I can usually get pretty good results with softproofing but this time the softproofing problem is probably user-error because adjusting colors based on the pattern print did give me a pretty good print.

But the fact is, although I'm mostly satisfied with my print results, I'm not too thrilled with Spyder 3 Express, which is what I use to profile my display. I'd really like to find instructions for dispcalgui that I can understand, but what I've found seems to require a PhD in color management in order to set it up. Any sugggestions in that regard?

Thanks

Have you downloaded and used the newest incarnation of DispcalGUI- now known as DispCAL? It seems like it has been made easier to use and there is a Quick Start guide to follow that might make things easier.

http://displaycal.net

take a look and see if it is useful.

Kirk

kirkt wrote in post #18032442
Have you downloaded and used the newest incarnation of DispcalGUI- now known as DispCAL? It seems like it has been made easier to use and there is a Quick Start guide to follow that might make things easier.

http://displaycal.net

take a look and see if it is useful.

Kirk

Actually, I found it about an hour before you posted this, and I was so excited to find it (after so many years of looking for something like it) that I didn't have the presence of mind to come back here and post about it! I really hope you didn't put a lot of time into looking for it.

Anyway, after running it, I loaded up that same image from the pattern print into PS Elements and I could tell immediately that it was different from before. It now matches the prints that I said before were "too red." So, with no proofing or adjustment of any kind, the screen image is now a good match with the print.

Of course, I LIKED the screen image I had before! So what this means is that almost all the portraits I've been working on for the last 2 months (since I got Portrait Professional and the NIK collection) are really, really horrible! REALLY horrible! (I don't know why but it seems that the only images affected are ones that I used those 2 programs on.) I haven't found any other images that look bad except portraits, and I guess that's because the Red component is a lot higher now.

And, even though I found it myself, I wouldn't even have looked for it if you hadn't brought up color management, so thanks for that. I've always felt kind of iffy about the profiles that Spyder was creating and I feel really good about this one. And the Quick Start guide did do the trick. Just as a matter of interest, the Spyder calibration takes about 15 minutes; the DisplayCAL calibration took 2 hours!

So I have lots of portraits to fix now but that's pretty minor considering that my display is probably really right for the first time.
Thanks again.

ncjohn wrote in post #18032262
Hi Kirk. I'm borderline sick so I may be misunderstanding but it seems that you're saying that a well-color-managed workflow will always give you prints that are color correct. But if that were the case we wouldn't have softproofing.

That is exactly what he is saying. If everyone in the process has every step of the process properly calibrated to a known standard, then you do not need to softproof.
I used to send thousands of prints to a print lab that did keep their machines properly calibrated, and they consistently matched my calibrated monitor, despite them making hundreds of ink and paper changes.

Softproofing is used when you have a printer that is calibrated, but not to a known standard, or when it is only capable of producing a limited gamut. e.g. you're using an aRGB wide gamut screen, working in software with an aRGB image, then soft proofing can simulate how the image will look printed in sRGB (which is what most commercial photo printers and Labs use).