Mike Chaney's Tech Corner
July 25, 2021, 06:17:08 PM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News: Feb 2013: Qimage Ultimate Challenges... have fun and explore features!
   Home   Help Search Login Register  

Professional Photo Printing for Windows
Print with
Qimage and see what you've been missing!
Pages: [1]
Author Topic: Lab 3D gamut profiles look different.  (Read 7458 times)
Full Member
Posts: 223

View Profile Email
« on: August 27, 2013, 03:24:16 AM »

I made a ICM profile with Profile Prism 7 and Vuescan on Canon Matte Photo paper on the Epson 3880 with non-OEM dye ink.

I also made one using the "ColorMunki Photo" as well as one using the "i1 Photo Pro 2" on the same paper and printer above.

Looking at the 3D Lab gamut, they are quite different.  The small one within the wire-frame gamut shown is Profile Prism and the larger wire-frame one is the gamut off the x-rite "i1 Photo Pro 2."

Don't know why the big difference.  Huh  Could be HP GP3110 scanner maybe?  Vuescan?  Not enough targeting info off the IT8?  Dunno.

A not-so fun test:

1. Make a new 8.5x11 photo page in your photo editor of choice.  "Bucket fill" the entire page with an RGB of 128 for each Red, Green, and Blue.  Save image.
2. Open QU to print using generated ICM profile.  Pull the edges in an inch or more, and set the "Border" color to the gray (or silver, showing 128 in all RGB windows) in Qimage.
3. Print it.

If you see a difference (color or luminance) between the Qimage "Border Color" 128 value, and your "Bucket fill of 128" image in the middle, you got a profile problem to deal with.

Forum Superhero
Posts: 3315

View Profile Email
« Reply #1 on: August 27, 2013, 01:29:09 PM »

Your comparisons were for which transform?  Perceptual?  RC?  You can never depend on much more than the general shape of the hull when comparing profiles.  Gamut edges will never match between different software.  You have to remember that ICC is not "smart", meaning that it cannot evaluate the colors in the image first and, for example, map them all to exact (accurate) colors if the image doesn't have any bright (out of gamut) colors.  It has to treat every image as if it could contain out of gamut colors.  For this reason, ALL profiling software does some smoothing and stretching of the gamut beyond the actual limits.

Also, your 128,128,128 test is not really valid.  At least if I'm understanding your process.  128,128,128 is always relative.  What color space (gamma) was used when you saved the image with the 128,128,128?  In Qimage Ultimate, when you pick a border color, your chosen color is assumed to be in the monitor profile's color space (sRGB if monitor profile is OFF) because you are picking the color visually or from your screen: and your screen is profiled (hopefully).  So if you pick 128,128,128 as your border color, QU will actually profile that color to the current printer ICC profile just like it does the image.  So if you make a 128,128,128 image and save that in your monitor profile's color space, then make a 128,128,128 border, the two will always match: because they start from the same image color space and are going through the same printer profile.  If they don't match, it's only because the color space of the chosen 128,128,128 doesn't match... not because the profile is bad.

« Last Edit: August 27, 2013, 01:34:39 PM by admin » Logged
Full Member
Posts: 223

View Profile Email
« Reply #2 on: August 27, 2013, 04:43:12 PM »

Thanks Mike.

Fwiw, I also ran the i1 Photo Pro 2 against the ColorMunki Photo using their respective software.  The 3D gamuts are very similar with the i1 favoring the blues a bit better and maybe due to its UV filter for optical brightners in papers.  The one I got in Profile Prism is very "blocky" in appearance and much smaller overall.  I fear the blockiness of the Profile Prism 7 generated ICM is pushing some colors off into a different coloration in some portions of a colors 0-256 density scale (especially the green/yellows where I can see density issues in the color ribbon of the green vertical ribbon in the Northern Light test image.).  I'm suspicious of IT8 target as the culprit since the x-rite's scans use 100 colors for the ColorMunki Photo test charts (2 pages) and 2033 colors (4 pages) for the i1 Photo Pro 2 where the IT8 has less colors for its initial scan base?

As to the 128 gray, I just don't get why 128 in a border in QU is different than a bucket fill of 128?  This has been driving me insane and boxes of paper and ink (Reminds me of the college assignment in the color lab of printing a Kodak Neutral Gray card in CP5/Ektaprint chemistry which took weeks!  Teacher was a sadistic SOB for that!).  I used sRGB as the colorspace in PS when I saved the image (ProPhoto is another matter!), but I cannot get a "QU 128 Border" and a "PS 128 Bucket Fill" to match out of the printer without some fine tuning in QU and altering the border 128 values - and QU is very good for that fine-tuning part.

I can also see the visual difference on the (x-rite calibrated) monitor as well.  However, I can alter the 128 in the QU border value settings to make it 128 in the print and agree with the bucket fill, just that it's a lot of work (e.g. QU might take a 132, 124, 116 for a Border, and PS needs a 127, 126, 121 for a match out of the printer - if that makes any sense?  At which point the spectrometer will read 128 for everything on the print.  Huh  )

Attached are the i1 Photo Pro 2 and ColorMunki Photo combined, and the Profile Prism 3D gamut alone (extracted out of prior image above).

I'll add that an image made with the ICM profile from the ColorMunki Photo tends to the warmer side where the i1 Photo Pro 2 tends to the cooler side.  The i1 also provides a smoother transition from 0-255 in the grays too.  The two x-rite devices also tend to be about 5 points different in agreeing with each other too.  The HP scanner reading the IT8 who knows?

Mack  (...Talking to myself this AM as I might get back to this in the future and have an "Ah Ha!" moment - I hope.  Wink )
« Last Edit: August 27, 2013, 04:53:14 PM by Mack » Logged
Sr. Member
Posts: 440

View Profile
« Reply #3 on: September 17, 2013, 01:43:03 AM »


Your step 1)  (of post 1) are you assigning or converting the colour space when you are saving the image?
Pages: [1]
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!