The most visited post on JimRaffel.com in 2007 is #16 written October 2005 and addressing metamerism. Because of the popularity of this post I have decided to revisit the topic. I will start with some basic terminology and definitions. Then, I am going to present a real world example we are dealing with right now at ColorMetrix.
Definitions: First, metamerism requires a pair of objects. The two objects are often referred to as a metameric pair. In order to be considered metameric the pair must; match under at least one combination of illuminant and observer; not match under at least one combination of illuminant and observer; and have different spectral response curves.
Metramerism Tests: Post #16 shows an example of the GATF REHM indicator used for visual assessment of light sources. This is also an excellent of example of a visual test for metamerism. The two samples may also be measured by a spectrophotometer, and then an instrument test for metamerism can be performed. If the two samples exhibit a small Delta E under one illuminant observer combination, but have different spectral curves that cross at least three (3) times they are metameric.
Our ProofPass.com system was recently used to measure two samples. Below I have provided the L*C*h* values of a ‘Red’ swatch measured on a hard copy proof and then reproduced and measured on an LCD monitor.
| Hard copy | Monitor | |
| L* | 40.56 | 43.02 |
| C* | 60.85 | 61.31 |
| h* | 28.44 | 27.9 |
The CMC(2:1) Delta E of the above two samples is 1.34 in a D50/2 degree environment.
Now, take a look at the spectral reflectance and emission curves of the two swatches. (Library Reading = Hard copy and Sample Reading = Monitor)
Again the instrument test for metamerism states that if the spectral curves differ, and cross each other at least three times, then the objects are metameric. The curves look different and I count at least 10 crosses.
As long as the physical sample is viewed in a proper viewing booth, and the monitor continues to be calibrated and also viewed in proper conditions, the metamerism effects of hard copy vs. monitor proofing above should not be a problem. My concerns about this type of metamerism extend to placing a package on a retail shelf which is seldom a D50/2 degree environment.
Hi Jim,
Hope you and yours are having a wonderful year. Here are my thoughts about the above.
1. sp: MEtamerism (interestingly, spelling for METAMERIC match is correct)
2. Many store environments have gone or are going to the Westinghouse equivalent of D-50 flourescent tubes in customer viewing areas (Target, Macys, Mervyns, and yes even Walmart are some I know of, off the top of my memory)
3. I have a grad student looking for a project in color matching. Can you suggest how we might approach a packaging solution in a store viewing environment and gather data about the metameric quality of color matching and verify with ColorMetrix and instrumentation? Please advise.
Best regards,
Howie
Hi Jim,
Hope you and yours are having a wonderful year. Here are my thoughts about the above.
1. sp: MEtamerism (interestingly, spelling for METAMERIC match is correct)
2. Many store environments have gone or are going to the Westinghouse equivalent of D-50 flourescent tubes in customer viewing areas (Target, Macys, Mervyns, and yes even Walmart are some I know of, off the top of my memory)
3. I have a grad student looking for a project in color matching. Can you suggest how we might approach a packaging solution in a store viewing environment and gather data about the metameric quality of color matching and verify with ColorMetrix and instrumentation? Please advise.
Best regards,
Howie
Hi Jim,
Wonderful info – and great writing.
Okay, so, while I struggle mentally (trust me, not had for me to do, now where did i leave my keys?) on how a color measuring device can possibly remove all the difference between a monitor transmitting light (of one color temperature) and a small section of paper reflection light (of color another temperature) – I am no engineer, so I will buy in to your scenario.
So, lets take this a step further – we are in a professional fashion photography studio, and the model is wearing a fabric that is Pantone 150 (or whatever the equivalent in the new Pantone Goe color is) – but corduroy !
So, lets assume (just for SAG) that we absolutely have a metameric issue (the fabric swatch matches the Pantone swatch in a D50/2 degree environment, but fails in ‘other’ environments – but no matter, we are just trying to get them to match in the press side viewing booth.
We all know the studio lights (scrims, fabric, ect.) and the digital cameras lens and CCD array will introduce new color issues – but if we zero them out of the equation…
Okay, so – in your perfect world – what gadget do we need to capture the color of corduroy and how would you – again – in your perfect world – embed that information so a system might use it (the system could be Photoshop with some fancy plug-in, but it could also be some color managed rip…
Would you use this gadget, or something else;
http://www.xrite.com/product_overview.aspx?id=349
Perhaps this will be Blog post 68-71 !
Michael Jahn
No expertise in particular, but fond of PDF
Hi Jim,
Wonderful info – and great writing.
Okay, so, while I struggle mentally (trust me, not had for me to do, now where did i leave my keys?) on how a color measuring device can possibly remove all the difference between a monitor transmitting light (of one color temperature) and a small section of paper reflection light (of color another temperature) – I am no engineer, so I will buy in to your scenario.
So, lets take this a step further – we are in a professional fashion photography studio, and the model is wearing a fabric that is Pantone 150 (or whatever the equivalent in the new Pantone Goe color is) – but corduroy !
So, lets assume (just for SAG) that we absolutely have a metameric issue (the fabric swatch matches the Pantone swatch in a D50/2 degree environment, but fails in ‘other’ environments – but no matter, we are just trying to get them to match in the press side viewing booth.
We all know the studio lights (scrims, fabric, ect.) and the digital cameras lens and CCD array will introduce new color issues – but if we zero them out of the equation…
Okay, so – in your perfect world – what gadget do we need to capture the color of corduroy and how would you – again – in your perfect world – embed that information so a system might use it (the system could be Photoshop with some fancy plug-in, but it could also be some color managed rip…
Would you use this gadget, or something else;
http://www.xrite.com/product_overview.aspx?id=349
Perhaps this will be Blog post 68-71 !
Michael Jahn
No expertise in particular, but fond of PDF
I don’t believe crossing of the spectral curves has anything to do with it, especially to say at least three (3) times…
Looks like the “library curve” lacks the precision of the “sample reading” and may falsely lead a person to think they have to cross curves; they are just not showing all the ‘ups-and-downs’.
Metamerism occurs due to the different spectral curves in the illuminant being used to say they are a match. They may cross at a wavelength and drastically change after that. Here is where the difference is evident.
Check out pg 23 http://www.triangledigital.biz/triangle/downloads/L11-029_color_guide_en.pdf
by X-Rite
Other:
http://docs.happycoders.org/unsorted/ai/case_based_reasoning/iccbr-co.pdf
I don’t believe crossing of the spectral curves has anything to do with it, especially to say at least three (3) times…
Looks like the “library curve” lacks the precision of the “sample reading” and may falsely lead a person to think they have to cross curves; they are just not showing all the ‘ups-and-downs’.
Metamerism occurs due to the different spectral curves in the illuminant being used to say they are a match. They may cross at a wavelength and drastically change after that. Here is where the difference is evident.
Check out pg 23 http://www.triangledigital.biz/triangle/downloads/L11-029_color_guide_en.pdf
by X-Rite
Other:
http://docs.happycoders.org/unsorted/ai/case_based_reasoning/iccbr-co.pdf
I believe I have seen it with less than 3 crossings.
I believe I have seen it with less than 3 crossings.