The New Light Fade Report Format and Some Actual Test Results

When I make color and tone corrections to my digital images in Adobe Photoshop I always use the “Proof Colors” feature so that Photoshop can simulate on my monitor how the colors will look when printed. It works very well on a calibrated monitor, but for very subtle color and tone edits, I have also learned to interpret the LAB numbers shown in the Photoshop Info tool. It takes time to familiarize oneself with edits by the numbers, but my basic philosophy is “Trust your visual senses, but verify by the numbers”, especially for very delicate changes. I’ve used this same philosophy to design the AaI&A light fading report format. The light fading test results are presented both visually and numerically. You can trust your senses to evaluate the visual fading patterns, and you can ultimately learn even more about system behavior by verifying what you see with the numerical results. The I* metric presents rational and precise scores for color and tonal accuracy, whereas the classic Delta E color differences (∆E) may be more familiar and helpful at examining just noticeable or noticeable differences as the samples fade.

To achieve the visual “Proof Colors” effect in the PDF file, I take the actual colorimetric measurements of the color patches in each test image, both before light exposure and after, and then convert them to RGB equivalent numbers in sRGB color space so that the fading can be as faithfully represented on your monitor as today’s 24 bit RGB color and computer display technology will allow. It helps that the AaI Standard 30 patch color set exercises a color gamut that is essentially contained in the sRGB color space. Significant gamut compression is not required to render the colors to your monitor, otherwise you wouldn’t be able to see them as they exist on the original print or after light fading. It does help your viewing experience if have a calibrated monitor so that you can observe accurate color representation, but for less than optimally calibrated monitors, the display errors tend to be nulled out on a relative basis across the “before” and “after” views of the test images in the PDF file. In this situation, the color changes can at least be reasonably judged in relative terms if not in absolute terms.

Today I have posted four PDF light fade reports for two papers and two inks printed with the Epson R1800 printer. The papers are either glossy or luster finishes, and therefore the gloss optimizers as supplied by both ink formulations were also used to make the test prints. The reports contain only the 10 Megalux-hour exposure level so far, but they will be updated as more exposure accumulates in test. Click on the sample #s below to download the PDF reports.

The samples made with Epson OEM ink were printed at Aardenburg Imaging whereas the samples using the MIS ink were provided by a contributing artist in Aardenburg’s Digital Print Research Program. The ability to have participants easily make the test samples using standard printmaking practices rather than having to undertake special densitometric calibration procedures is one of the advantages of the I* metric and method. Another advantages of using the I* metric and method is that it logically scores the performance at every step along the fading curve. A single failure criterion is not required to be reached in order to post test scores. One can follow and report on the progress of fading all throughout the aging cycle of the print.

The AaI&A light fading report is designed so that you can view the “before” and “after” samples in both side-by-side comparisons and also in an “overlay” mode. The overlay mode on page four of the report takes advantage of PDF layers which can be accessed by Adobe Reader 8 (a free download at the Adobe site). Toggling between layers quickly superimposes before and after test exposures and can help you visualize the magnitude and directions of change in the image colors. The utility of this feature will become more apparent as the results of additional Megalux-hour exposure intervals are added to the test reports.

Sample #: AaI_20071008_SN002
Printer: Epson R1800
Ink: MIS Associates (www.inksupply.com) R800/R1800 Ultrachrome Equivalent
Paper: Red River Ultra Pro Gloss Plus (www.redriverpaper.com)
10 Megalux-hour Exposure: I* Score = 95.4color/98.3tone

Sample #: AaI_20071008_SN018
Printer: Epson R1800
Ink: Epson OEM for R800/R1800
Paper: Red River Ultra Pro Gloss Plus (www.redriverpaper.com)
10 Megalux-hour Exposure: I* Score = 94.1color/98.9tone

Sample #: AaI_20071008_SN007
Printer: Epson R1800
Ink: MIS Associates (www.inksupply.com) R800/R1800 Ultrachrome Equivalent
Paper: Epson Premium Luster Photo Paper
10 Megalux-hour Exposure: I* Score = 88.7color/98.1tone

Sample #: AaI_20071008_SN013
Printer: Epson R1800
Ink: Epson OEM for R800/R1800
Paper: Epson Premium Luster Photo Paper
10 Megalux-hour Exposure: I* Score = 99.8color/99.0tone

Let me conclude today’s post by making a couple of observations about the test results to date. Ten Megalux-hours of light exposure is enough noticeably to fade the BW#3 patch on the Blue Wool scale of 1-8 textile patches. The #1 BW patch is extremely fugitive and #8 is highly stable. 10 Megalux-hours of light exposure is also equivalent to 5 years on display using the 450 lux/12 hour per day, cool white fluorescent, glass filtered illumination criteria of Wilhelm Imaging Research, Inc. Thus, a 10 Mlux-hr exposure is not insignificant, but even by museum standards it is not a large dose, and any artwork exhibiting appreciable fading at this exposure level would be classified as a fugitive material. By examining the I* scores and ∆E color differences and by making your own visual assessment of the images in the AaI&A light fade test reports, you will be able to judge for yourself the relative merits of the test samples. Here’s my take on these test results so far: First, the Epson ink with Epson Premium Luster paper has completed this exposure interval with no visually discernible changes. Second, The MIS ink on Epson Premium Luster Photo Paper is showing measurably more change than the other three samples. The loss of yellow colorant is not severe at this point, but it is noticeably affecting the neutral gray scale and the light skin tone colors in the image. However, here’s the interesting part of the study to date. Both ink sets are running “shoulder to shoulder” on the Red River Paper. The MIS ink is performing better on the Red River paper than on Epson Premium Luster Photo Paper, while the Epson OEM ink is performing worse on the Red River paper compared to Epson Premium Luster. Some of the changes are due to loss of optical brightener activity and not the inks themselves. Paper does make a difference, even for pigment based inks! Measurable differences can start very early on in the “life” of a photograph.

Now, onwards to more exposure doses where we will discover whether the fading rates are sustained, slow down, or speed up in subsequent 10 Megalux-hour doses. The next measurements will take place at 20 Megalux-hours total exposure and should be posted on or about May 1, 2008.