Profile Prism and my Canons

[Grandad35]

CollegeDad71,

I warned you that color management wasn't easy. You will have to do a lot of reading and experimentation to understand it, but here is an explanation for what you see when you print the test chart.

1. The following image shows a comparison of the sRGB and aRGB color spaces vs. the color space of an i9900 with OEM inks and Photo Paper Pro paper. I used www.viewicc.de to generate these images, as it allows you to upload your personal profiles and compare them against other profiles. The top part compares sRGB (colored area) with the Canon printer profile for PPPro and OEM ink (wire frame). There are areas of sRGB that the printer can't reproduce, but there are also areas of the printer's color space that are outside of sRGB. The bottom part is the same comparison, using aRGB and the same printer profile. Since my camera can be set up to generate images in aRGB, the larger color space shows why I work in aRGB, - my prints have the capability of a wider range of colors than if I set my camera and PS for sRGB.

If the printer's color space is larger than the source color space there isn't much of a problem, since the printer can print the colors as they exist in the source color space. Obviously, there is a problem when the printer's color space is smaller than the source color space. This image shows the same test image in sRGB, aRGB and the Canon profile used above (the image has an embedded aRGB profile, and it may need to be viewed in PS {or Qimage} to see the differences in the first 3 charts). However, the 3rd chart clearly shows that the colors are expected to be different for the printer, as the printer cannot produce the extreme colors. I believe that this is what you are seeing when you print this test chart. How do your "standard photo prints" look when they are printed? Since most photos do not approach the extreme colors, they normally look much closer to the display than the test chart looks.

One more point - when you convert to a new color space, there are several options (called "Intents") that change how colors near and outside of the limits of the printer's color space are printed. The following image shows the two most commonly used Intents - "Perceptual" and "Relative Colorimetric" (there are also "Saturation" and "Absolute", but they aren't usually used for photos). It can be seen that this setting also affects how the image is printed.

Read the references in my previous post and chase down the links that they give - it will take a while before all of this strange jargon begins to make sense.

 

[CollegeDad71]

Grandad,

I've been reading through the links you suggested and working with Monaco EZ color and PrintFix to get an acceptable profile. PrintFix includes a CMYK calibration sheet for its scanner, those colors are quite distant from those in the color test image you sent in post #15, possibly explaining why they said their profiler does not support 3rd party inks. I made a new sheet image in Photoshop using the color values from your image and the printed result after profiling was actually a lot closer in every color except blue. Their profiles so far have been heavy in magenta, I haven't started with their profile editor yet. EZ color at least calibrates to a reflective IT8 target, but their profiles so far have had a green cast. The really odd thing to me is that I can get the histogram for an RGB image and its soft proof to be very similar, as well as specific color values on screen, but the print with preview image is always different. The preview identifies the magenta cast from a PrintFix profile and the green cast from EZ color. The printed output is in between the preview and soft proof image. CM from the printer is disabled, honest. I have used Relative Colormetric for EZ proflies and Saturation for PrintFix (because PF says that their profiler is tuned for saturation). Source space is set to Proof, print space to "same as source". I would really appreciate knowing if your preview image matches the soft proof.

 

[Grandad35]

CollegeDad71,

I don't use PS to print (I use Qimage), so I am not all that familiar with the details of its use. I soft proofed the color test image with my printer's profile and then used PS's "Print with Preview" to display the image. It appears that the previewed image is the same as the original, not the soft proofed version. A screenshot of the dialog box (on PS CS2) covering about 1/2 of the original image demonstrates this, as well as the settings that I used to print the test page. The printed test page doesn't match either the original or the soft proof perfectly, but it was a lot closer to the soft proof than to the original. It is always difficult to compare images at the limits of the color space - the important thing is that the colors in my printed photos appear very close to the photos on my monitor.

I am not an expert in this, but I believe that it is up to each profile to contain the necessary information on how to handle each of the available Intents. It is hard to understand how PF could recommend Saturation for the Intent. The following sentence was copied from (http://www.normankoren.com/color_management.html#Rendering_intent):
"Saturation, also called Graphic or Preserve Saturation. Maps the saturated primary colors in the source to saturated primary colors in the destination, neglecting differences in hue, saturation, or lightness. For block graphics; rarely of interest to photographers."

In all fairness, however, the color differences on typical images viewed with the various intents are often quite minor, so don't get too hung up on this point.