What can one do about metamerization?

[Michael H. Cothran]

First, I hope I'm using the appropriate word. If not, please correct me. I'm concerned with, and am having an issue with my prints metamerizing, ie, changing hues under different lighting situations. I understand this is a fact with inkjet prints, but I've never encountered it, had an issue with it, or noticed it, running Epson drivers with canned Epson profiles in the past. However, a few months ago I purchased the Image Print RIP, and metamerization reared its ugly head. Prints that are neutral indoors turn green outside, and prints that are neutral outside turn pink indoors.
FWIW - I'm running an Epson 7600 thru a new Mac Pro, OSX.6.
IP has not been much help, pretty much telling me this is the way it is.
As a test, I ran the same file thru both the Epson driver and IP driver. Inside they match as closely as you could expect. But walk both of the prints to a window, and the IP print turns noticeably green, while the Epson print remains neutral. I simply love the management conveniences of IP, but cannot live with the noticeable hue changes in my prints.
I'm really getting frustrated with this.
Any help or suggestions would be appreciated. Thank you.

[Ernst Dinkla]

Selected rendering is the same?

Is the profile in both cases made for the same color temperature?

Another thing could be a difference in the profiling to the paper white. A paper with OBAs may act different when light is changed if the profiling has been made to neutralise the paper white or keep the original paper white. Rough explanation, it can be more complicated if a UV-cut or UV passing instrument has been used. Check for ink dots in the lightest spots in both. Paper sample printed at the same time from the same batch, so not one print that has lost its OBA effect in time?

Use a microscope and check how much of the colored inks are used in the neutral greys in both the Epson and IP print, the more the composite grey mixes  (CcMmY inks) are replaced with grey inks the better to reduce "metamerism". That is most visible in the neutral greys but should affect desaturated colors too.


met vriendelijke groeten, Ernst

330+ paper white spectral plots:

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

[madmanchan]

This is the nature of the original UltraChrome inks (e.g., as used in the 7600).  You'll get different results with the Epson driver and RIPs that use their own ink-mixing methods, because some of the inks are more prone to color shifting under different lights, than others.  (As I recall, the original UC Yellow was a culprit here.)

Have you tried Bill Atkinson's profiles that he made for the 7600/9600 printers? 

[Michael H. Cothran]

This is the nature of the original UltraChrome inks (e.g., as used in the 7600).  You'll get different results with the Epson driver and RIPs that use their own ink-mixing methods, because some of the inks are more prone to color shifting under different lights, than others.  (As I recall, the original UC Yellow was a culprit here.)

Have you tried Bill Atkinson's profiles that he made for the 7600/9600 printers? 

Yes, I have both the Epson canned profiles plus Bill Atkinson's profiles. There are no specific profiles in this bunch for different light sources. I presume, but could be wrong, that Epson's and Bill's are both daylight balanced. Both work well and show no metamerizing.
It's the IP profiles that show a distinct color shift. And IP does have profiles for different lighting. They advised that their F3 profiles are designed for mixed lighting, and are their most commonly used. I've tried both IP's F3 and daylight profiles, and both are bad with color shifts. Just the opposite on the color wheel. A neutral F3 print turns green outside, while a neutral daylight print turns pink inside the house.
Someone else suggested that it might be the way the IP driver lays down the ink.
I recently had to decide between having my 7600 overhauled for $1300 (including a new head), or spring for a new 7900. I, obviously, chose the former. Would the IP profiles work better with the newer inks in the 7900?

Something else of a nuisance - Last year I also upgraded my old G4 (running 10.4) to the new Intel based Mac Pro (running 10.6) Since then, the Epson files don't print the same, and I now have to lighten them drastically in order for print and monitor to match. I found the best way to do this is to get the image the way I would "like" it to print, and then increase the brightness with a new layer. I found it easiest to dupe the background layer, then change the blend mode to Screen, and reduce it to between 50%-75% opacity. With this "method" the print looks close to my file "without" the Screen mode layer. What a mess!...

[madmanchan]

Well, the whole driver system got overhauled starting with 10.6, and my experience is that older profiles (made for the older OS versions) don't necessarily work under 10.6.  My personal resolution was simply to rebuild profiles from scratch for 10.6, but I realize that's not for everyone ...   

[Damir]

I know it sounds conservative - but when I got my system working I stay with it. Therefore I still use old Windows XP, old Photoshop SC3 and outdated firmware on my Z 3100. It run perfectly, when I made update to Photoshop CS4 I lost 2 weeks to tune it and never got results as with CS3 - therefore I just get back and stay with system I am satisfied.

[ippolitois]

I had a similar problem as you. What rendering intent are you using with IP? when I tried the relative color intent, I had all kinds of issues with a magenta shift but after I returned to perceptual, I was back to a neutral situation. F3 is your best bet for the best profile.

Hope this helps.

Paul

[Michael H. Cothran]

I had a similar problem as you. What rendering intent are you using with IP? when I tried the relative color intent, I had all kinds of issues with a magenta shift but after I returned to perceptual, I was back to a neutral situation. F3 is your best bet for the best profile.

Hope this helps.

Paul

I'm already using Perceptual along with the F3 profiles, so that's not the culprit. My monitor is a new Apple 27" Cinema, and is calibrated with Color Munki, using a brightness level of 1.20. I don't think the monitor could look any better, or any more neutral. Color is not an issue with F3 prints indoors, as they do come out quite neutral. The problem occurs when you take the F3 print into daylight. Wham!! It's surprising at how much it can go from neutral to green. It could almost be a "magic trick". Brightness differences, between monitor image and print are extreme under any lighting condition. It's amazing at how dark the prints come out, even when I think the monitor image is way too bright or over exposed.
So, between metamerizing and brightness differences, I'm going looney !
I certainly agree somewhat with Damir - when it works, don't fix it. Printing with the 7600, Epson driver, and Mac G4, I never had problems. Now everything is discombobulated.

Maybe there's just something not clicking between the old school 7600, and the new Mac Pro. FWIW - I've got the latest Epson driver for the 7600 installed, but I'm still running CS3.

[JohnHeerema]

If your prints seem to be dark in comparison to the monitor, you might want to review your monitor calibration. There is a lot of advice to be found elsewhere on this forum, but my first thought is that your monitor brightness is set too high. Most colour management workflows are based on a relatively dim monitor (100 cd or less).

You should be able to calibrate your Apple monitor to something around 100 cd, although some newer LCD screens can't be calibrated this dim. In principle it should be possible to map the perception of a brighter monitor to the perception of a print, but I have yet to see a colour management methodology that addresses the perceptual difference between a bright monitor and a print that will be viewed under typical lighting conditions.

Your finding of metameric failure using the ImagePrint RIP is quite interesting.

[Michael H. Cothran]

If your prints seem to be dark in comparison to the monitor, you might want to review your monitor calibration. There is a lot of advice to be found elsewhere on this forum, but my first thought is that your monitor brightness is set too high. Most colour management workflows are based on a relatively dim monitor (100 cd or less).

You should be able to calibrate your Apple monitor to something around 100 cd, although some newer LCD screens can't be calibrated this dim. In principle it should be possible to map the perception of a brighter monitor to the perception of a print, but I have yet to see a colour management methodology that addresses the perceptual difference between a bright monitor and a print that will be viewed under typical lighting conditions.

Your finding of metameric failure using the ImagePrint RIP is quite interesting.

Thanks John. I've tried adjusting the brightness on the monitor, but the perceptual difference does not change. I've darkened it past the point of being acceptable for any kind of viewing. Color Munki does a good job calibrating, but the difference between print & monitor is far from acceptable. I think the problem is something else. As for the metameric failure, that is what I'm seeking help for in this thread. It is definitely a "deal breaker" issue for me and IP. But with $1200 invested in the software, also offering great printing and layout management, I'd like to find a solution.

[Geoff Wittig]

First, I hope I'm using the appropriate word. 

The most accurate term is 'metameric failure', wherein a print's colors look fine under one light source, but wrong under a different source.
Cutting to the chase, the Epson 7600 has quite bad metameric failure when using any gloss, semigloss or satin paper. In fact I found it to be so severe that I used my 7600 exclusively with matte cotton rag papers, where the problem was much less of an issue. Since swapping from matte black to photo black was also exhorbitantly wasteful with the 7600, it was an easy decision to stick with cotton rag papers.

I don't think there's any good solution to the problem using the 7600 inkset, other than sticking with matte papers. Newer generation printers are much better in regard to metamerism/metameric failure

[Ernst Dinkla]

The most accurate term is 'metameric failure', wherein a print's colors look fine under one light source, but wrong under a different source.

Metameric failure, wherein two color patches differ under one light source, but match under another light source. The last is called a metameric match.

The most accurate term is color inconstancy, wherein a print's colors look fine under one light source, but wrong under a different source.

To quote Berns:

" Related to metamerism is COLOR CONSTANCY, the tendency of an object to maintain its color despite changes in illuminating conditions. Each object forming a metameric pair has a different DEGREE OF COLOR INCONSTANCY. For example, the right-hand square will turn pinkish when illuminated by incandescent illumination, whereas the left-hand square tends to remain grayish. Although both samples match under daylight and are thus a metameric pair, only one sample of the pair exhibits objectionable color inconstancy. If both formulations are candidates as a color standard, the sample that minimizes the degree of color inconstancy would be preferred."


met vriendelijke groeten, Ernst

330+ paper white spectral plots:

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

[Michael H. Cothran]

Cutting to the chase, the Epson 7600 has quite bad metameric failure when using any gloss, semigloss or satin paper. In fact I found it to be so severe that I used my 7600 exclusively with matte cotton rag papers, where the problem was much less of an issue. Since swapping from matte black to photo black was also exhorbitantly wasteful with the 7600, it was an easy decision to stick with cotton rag papers.

I don't think there's any good solution to the problem using the 7600 inkset, other than sticking with matte papers. Newer generation printers are much better in regard to metamerism/metameric failure

I certainly wish this were the case, but sadly, at least for this response, it is not. I AM using matte paper, only matte paper, and have done so for almost 10 years of owning my 7600. Also, matte black only. Never ran Photo black thru the machine. Currently, I print on Hot Press Natural, Moab Entrada Natural, and Enhanced Matte for economic jobs. So this is not the culprit, as my metameric failure is with these matte papers. But only with the IP profiles. The Epson driver plus canned Epson profiles or Bill Atkinson profiles don't display this metameric failure. Thanks for the response and possibility though.