|
 |
Quick Find |
 |
|
|
|
Categories |
|
|
|
|
Information |
|
|
|
|
| 
Controlling Metamerism
on the Epson Stylus Photo 2000P
Because the Epson 2000P uses pigments inks, metamerism has
shown to be a significant issue when balancing color for each
print's final display environment -- in other words, one color profile
will not fit every illumination situation.
What is Metamerism?
Metamerism is an effect created when objects having different spectral
distributions look alike under one light source but appear different when
viewed with a dissimilar light source.
Metamerism is most frequently seen when two colored objects match in
daylight, but differ markedly in color when viewed in tungsten-filament light.
This arises because the visible absorption spectra of the two objects differ
significantly, although the tristimulus values in daylight are identical (the
colors have the same set of color coordinates but different spectral reflec
tance curves).
Pronouncing "Metamerism":
me*tam*er*ism (just in case you wanted to talk intelligently with others on this subject -- hey, we
never knew how to pronounce it until just last week ;-)
Examples of Metamerism in 2000P Output:
John Custodio of New York City sent us some great examples of metamerism. John supplied us with a grid of his images and color patches that were printed on three different
output devices:
- A Lightjet print (a digital printer similar to a Durst Lambda that exposes regular dye-coupled silver halide photographic paper, i.e. Kodak Ektacolor or Fuji Crystal Archive)
- An Epson Stylus Photo 2000P print (Epson Premium Semigloss Photo Paper).
- An Epson Stylus Photo 1200 print (using standard Epson inks on Epson Photo Paper).
|
All photos © John Custodio
|
When these three prints are viewed together under varying light sources, there is a considerable change in color on the 2000P prints, but only moderate change in the color of the Lightjet and Epson 1200 prints.
|
Here's an example of a southwestern "redrock" formation that John photographed, and the way it appeared to us,
when viewed under 5000 degree Kelvin lighting (a graphic art industry standard for color matching): |
 |
Here's what the CCD scanner (Umax Powerlook 3000) saw (which uses a cold cathode lamp). You'll notice that
there was little change on the Lightjet and Epson 1200 images, but the Epson 2000P image changed quite dramatically: |
 |
Since most of us don't view our prints under "cold cathode lamps", we decided to do a more practical
test. The six color patches (red, green, blue, cyan magenta and yellow) in the bottom right corner of
John's test print were photographed with a digital camera, under three common light sources:
- Tungsten filament (3200 degrees Kelvin)
- Reflected daylight (open shade or "north light")
- Florescent (common "cool white" tubes)
All three print types were photographed together under the same light source (one image group for each light source) using a 1.3 megapixel digital camera. The digital images were acquired, separated from each other and adjusted in Photoshop's "Levels" to set the same white point for each image. The purpose of this demonstration is not to show how accurately each print reproduced the six colors (after being printed and rephotographed), but how much the colors shifted under the three different light sources:
Our Conclusion: The Epson 1200 images shifted very little in color, going from one light source to another. (Although the 870/1270 printers use a different lightfast ink, it is still a dye-based ink like the Epson 1200 inkset, which could give similar results -- however, we have yet to test the 870/1270 ink.) The Lightjet (or regular photo graphic "C" prints) were not without change. Although there was a large density shift in the blue when going from tungsten to reflected daylight, the blue's hue remains fairly consistent. The Epson 2000P prints, however, showed a huge shift in the magenta when going from tungsten to reflected daylight, and a little less pronounced in the florescent lighted print. (There appears to be little red in the 2000P/diffused daylight/magenta patch -- making it appear almost "blue", with a little more red appearing in the florescent illuminated "magenta" patch.) There was also a slight shift to the green in the diffused daylight/yellow patch. Both of these shifts also tend to make the "daylight" and "florescent" red patches appear a little darker and less saturated. These test results seem to coincide well with our visual observations. Prints from the Epson 2000P appeared most pleasing and neutral when viewed under tungsten illumination, and shifted to a greenish-cyan color when viewed under reflected daylight (and slightly less green when viewed under cool white florescent lighting).
In Our Opinion, the Epson 2000P Print Driver is Profiled for Tungsten Illumination.
Epson 2000P
Print Driver
Dialog Box
|  |
|
Aliza and Dave Vanderlip of Sterling Images in Ashton, Oregon first brought this opinion to our attention. They pursued Epson for a confirmation of this opinion and were finally able to get one of Epson's "Techies" to give them a correction when making prints for reflected daylight and florescent illumination. The correction Epson gave was to add magenta (or remove green) in Photoshop's "curves", by using an input of 192 and an output of 187 in the green channel. (We think that works fine for cool white florescent tube illumination, but for diffused daylight, we prefer going a little further, with an input of 195 and an output of 187.) |  |
Putting the Epson 2000P to Work:
I'm pretty happy with the 2000P output. The saturation is not quite as strong as the 1200 or the 1270, but close enough. The blacks are not as strong the 1270's black (about 90%). But for a pigment ink, its not bad. As long as you don't compare 1270 prints side-by-side with the 2000P prints you'll be happy with the 2000P output. I think what Michael Greer had to said on the Epson-inkjet LEBEN discussion list (17 Sept 2000) bears repeating:
When I was running MIS archival inks in my 3000, I grew to love the output. I did a test against Epson OEM ink and the Epson ink definitely showed its superiority with its wider gamut. But for my subjects (portrait studio shots), the MIS Archivals produced gorgeous output. Other people hated the relative lack of gamut and stopped using the ink. The key word is "relative". When we look at prints for enjoyment, we're not holding other prints up against them to check gamut. When viewed by themselves, my MIS prints are extremely pleasing.
Bottom line: All pigmented inks suffer from inferior color gamut relative to dye inks. But that does not mean that impressive prints can't be made. One must test for oneself.
Knowing how the customer plans to display your prints is an important issue when printing with the Epson 2000P. With the 2000P, it is good to know what KIND of LIGHT your client plans to DISPLAY your prints under. Due to the higher "metamerism" effect on the 2000P pigmented inks, a print bal anced for tungsten lighting will look quite greenish-cyan under reflected daylight. With a 1270 dye -based print, there would be minimal apparent color shift -- it would look a little cooler (bluish) to the eye, but not a more extreme shift to the green/cyan, like the 2000P print.
You can live with these metamerism shifts -- you just have to be prepared for it (custom made prints for each client situation). Whereas with the 1270, you'd be able to make generic prints to sell at an art festival (not knowing where [the illumination type] the client was going to hang the print under). In my opinion, the 1200 or the 1270 is an excellent "portfolio" printer -- where one is keeping their prints in a book or a box. The 1270 is a great printer where one might be selling open edition prints at an art festival (matted prints in sleeves or shrink wrap) with a warning that the prints should not be taken out of the sleeve until they are ready to be framed (due to possible cyan fading). The 2000P prints I'd reserve for more serious custom prints for portrait clients and limited edition fine art, where you have more control on matching the print for the customer's final print environment.
Despite what I just said, you should realize that I and many other photographers are quite picky when it comes to color balance. What I and other "perfectionist" think is unacceptable color, looks just fine to the average art buyer! I also find that prints with deep, rich colors do not show the effects of metamerism as much as prints with subtle or more mute colors. I've also found that if I print for dif fused daylight or cool white florescent illumination, and I display the print under tungsten, the resulting shift to the red is often not as objectionable for most subjects as the shift to greenish-cyan (which comes when balancing for tungsten and then moving the print to reflected daylight or cool white florescent). I just completed a series of watercolor prints for a well-known wildlife photographer, using the Epson 2000P and balancing for cool white florescent illumination, and found the color balance of all the prints to be quite acceptable when viewed under other lighting conditions.
No printer will fill all your needs anymore than one camera lens will cover all a photographer's shoot ing situations, but the Epson 2000P comes darn close for me. --Royce Bair
Controlling Metamerism - balancing color output to your print's display environment
|
Tracking
My Account
Cart Contents
Checkout
Home page
Forum
Contact us
Links
Newsletter
Finishing