Separate names with a comma.
Discussion in 'Color: Film, Paper, and Chemistry' started by Iluvmycam, Dec 24, 2013.
Something I did in 2012.
Interesting. I'd like to see the ink-jet prints in 60 years.
Some current inkjet prints will do great! It all depends on the media used and the ink system used.
As always, see Henry's site for much more info. http://www.wilhelm-research.com/
The inkjets look quite good, I assumed they were all pigment based, which claim higher stability. But the lower range of inks might not be it...
Modern C-type (Fuji CA) holds quite well. The kodak royal is current? Perhaps Endura is the most comparable of the Kodak range to Fuji CA. On C-types, apart of the 1992 Wilhelm book and some figures given here and there there isn't much about it.
Too bad about the vintage original Dye transfer prints... As they had to be sacrificed for the tests, right? Well, sometimes a few casualities in the name of science have to be taken...
To me it is an interesting topic, not that much is talked about stability. Well, in an instant satisfaction seeking society archival topics aren't a priority.
As always, the other factors missed are "Summer or Winter" as the hours of daylight vary. Also "Heat and Humidity" as these affect results very much. Also, the absolute illumination level is important as there is reciprocity in all dye fade tests just as in imaging itself.
So, unless the tests were all run side-by-side at the same time, they may be badly skewed.
I've been saving this comment/opinion for a thread like this...
Does anyone remember when Kodak and other companies release an improved color paper that would not fade or yellow like the previous generation?
How about some years after that, when another improved paper was released that would not fade or yellow?
Then again, a number of years later?
Then they decided it was the album pages causing the problem. New materials were released that would not fade or yellow your prints.
Some years later, improved album pages were released...
I do not doubt that each improvement was truly better. I'm just pointing out how things are not always as they are presented.
Just about any color print, ink-jet or RA-4, will fade with time. I don't fully believe the marketing about how good the ink-jet pigments are.
In a short period of time we can give a paper the amount of light it would reasonably be subjected to after 20 years, for example. The same with some other variables. However, other things will happen in 20 years that are not easily simulated, and this may confound the results. Accelerated aging tests are fine, and offer a standard for comparison, which is the important part. However, I do not know how well they correlate with the real passage of time.
Of course, now I have to search APUG for information on accelerated aging, as I suspect PE has probably addressed this before, and I am interested in learning more about it.
It's way more complicated than meets the eye. Broadly, fading can be separated into two types, light-fading (due to exposure to light) and dark-fading (long-term chemical reactions which occur even in the dark). But other factors can be involved, such as temperature, relative humidity, and atmostpheric pollutants, such as oxidizing gases etc.
I've compared predicted vs real image-stability characteristics for DARK-storage on some Kodak VPS film, on and off over more tha a 20 year span. I found that it correlated well with predictions made in Kodak publication E-107, Image-Stability Data: Kodak Color Negative Films. So my confidence in the methods has been supported by real world examples. These dark-fading "Arrhenius methods" are based on a series of elevated temperature tests which allow predictions of time-to-fade at lower temperatures.
Regarding light-fading stability, this is a different story. PE raised the issue of reciprocity (failure). And I don't know that there is any obvious way to get around this. It's easy enough to do comparative testing between different materials, but the only strictly valid result is that material 'A' is better than 'B' when exposed to such-and-such an intensity of light at a given temperature and humidity. If you change intensity of light, the results won't necessarily be the same. If this seems far fetched to you, consider what would happen if you tried to predict who would be the best long-distance runner based on results of a 100-meter sprint. You'd probably guess the wrong person because different characteristics come into play.
If you really want to learn more about this, I'd suggest Henry Wilhelm's book, available for free download. Or have a look at some of his free on-line papers. This one, in particular, discusses methods of dealing with the reciprocity failure issues of accelerated light-fade testing: http://wilhelm-research.com/isj/WIR_Japan_Hardcopy_2004_HW.pdf
The light intensity example doesn't seem far-fetched at all - I figured as much. The effects of reciprocity and "dark-fading," however, are items I find intriguing in a purely academic sense (in other words, I like to learn things even if I do not apply that knowledge). I know these things occur, but have much less information on them. I should probably assume that researchers have used real-time data (over years) to come up with a decent factor for this.
I skimmed the lit review you linked to, and will read it fully later - need to play with my Christmas "toys" first. Typically, when I get a research article I first read the Abstract, jump to the Results section for correlations and alpha levels, then move to the Methods. I am fairly bad at math, but can interpret a study, lol. I do eventually read the whole article, though.
I'll have to check out the book you mentioned.
Here is an example of "reciprocity" in dye fade. Fuji CA is better when faded at 500FC and Kodak Endura is better at 200FC. The former represents lighted office areas and the latter represents home and museum lighting. Henry uses 500FC for his fading tests so which paper looks better in his hands?
What is happening in these accelerated tests is that Oxygen diffusion is important in light stability. The faster the test, the less Oxygen getting down to the bottom of the emulsion layers and so fast tests are different than slow tests.
So, there is "reciprocity" and then there is "reciprocity". I have described 2 kinds here.
The ICIS (International Congress of Imaging Science) gives a short course in image stability taught by Dr. Jon Kapecki, who was a co-worker of mine at EK. I took the course some years ago and have his textbook which I recommend. It has a lot of examples of all kinds of color paper tests including digital which also has "reciprocity".
As for new versions of color paper, we at Kodak kept a standard chart of each new version of paper with the image stability plotted on it. There was never a claim of "no fade" but rather one of better image stability with every new version.
With Ektacolor 70 paper, many complained that it turned red with keeping but I have samples dating from 1970 that look just fine. The thing was, at that time, among other effects, the magenta and yellow became so much better that the change in cyan became obvious. And, with good processing, they do keep well without reddening as my 1970 samples show.
All these accelerated fading testing methods have their flaws, and none is a real substitute for real time (decades, at a minimum) under a variety of real-world display and storage circumstances, but only a set of clues. Extrapolating inkjet results is inherently difficult because the potential ingredients are so complex, and are certainly not all true pigments! (as if all pigments were truly permanent either). Then you've always got the inevitable BS coefficient put in there by the marketing departments. The interest industry have taken in improving permanence in all these various color media is certainly encouraging. But I think it's pretty naive to ever extrapolate any brief testing standard into claims of so-many-years (let alone centuries) of permanence.
Hi, I've never seen any real flaws in the Arrhenius testing method, anything specific you could point out?
For those who don't know what it is, this method uses a handful of test samples "aged" at different temperatures. As each sample reaches an endpoint, for example a 20% density loss, the results can be plotted on a graph, using scales = 1/absolute_temperature vs the log of time. If results fall on a straight line, this can be extended to any (lower) arbitrary temperature, and the predicted time to reach the endpoint can be read off the other axis.
And, it works for light fade as well provided that there is no oxygen diffusion problem or other such odd effects.
My take on the situation has never changed. If you want something to last for the ages, use optical-method conventional silver B&W, toned. For color, the test of time still has Kodachrome as the world's champ. I don't listen to much other claims. I fully realize an E-4 or C-22 is still out there, pretty as ever. But life's too short for me to look for one somewhere. Inkjet, schminkjet--that ain't photography.
Sorry Bill, I really can't get into this much without getting long-winded. I'm pretty familiar with this and that, not only because of years of color printing experience, but also due to an analogous background with industrial pigments. Tests help, but there are far far too many variables out there to make simple extrapolations dependable. And like I said, there is almost always a BS coefficient in there somewhere form the marketing side of things too. Wilhelm obviously did everyone a favor by getting the conversation in high gear, but I found a lot of his conclusions rather simplistic, and not really matched to my own actual experience with prolonged tests - at least going forward in contrast to observing how older media had actually held up. Lots of his idea fall apart badly when very complex formulations like inkjet get involved (which should never be called true pigment prints to begin with). And few things fade in a straight line. Aardenburg has formulated some testing methods which fill in a number of the gaps left behind by Wilhelm, and do address the complexities of inkjet in a more realistic manner. But nothing is perfect. Even the Sphinx of Egypt isn't truly archival. If it's original architects viewed it today, they'd probably go into shock.
We have to be satisfied with "improvement". ... or else accept the fact that following generations will inherit more truly mediocre images
than ever before. Fortunately, many of these will exist only on discs, which will come in practical for skeet shooting.
I am finding the information here interesting. I know full well that marketing spin is just that, so it is good to read concepts and real information relating to the subject.
Marketing and advertising material can be true or misleading. The mere fact that it is composed by people trying to sell me something is enough to take it with a grain of salt and think for myself. Even if my thinking is flawed, at least _I'm_ doing it and not letting others think for me.
One problem I have in real life is many people I talk to believe any marketing material they are exposed to, and sound like a TV commercial when we discuss things like photography, cars, food, etc.
Drew, you understand that I have to reject your position, that the Arrhenius testing method is flawed, if you can't give any hard examples. I know there are possible issues with inkjet, but as APUG is specifically not a digital site, I think we have to stick with chromogenic dyes. With that limitation, I'm not aware of any flaws with the Arrhenius test method.
I worked for a large finishing chain for quite a few years, where I set up an image-stability testing program, so I have some experience with this. Although we did NOT do Arrhenius testing ourselves, we had such data from some of our suppliers. (Konica's SR-100 "Century" paper's projected lifespan was based on Arrhenius tests.) As part of our our own testing program, we also held long-term room-temperature test samples, and I don't recall ever seeing any of these fall outside a predicted tolerance range.
I don't believe there is any such thing as a "BS coefficient" for Arrhenius testing.
Drew, I wish you would be more specific instead of making vague innuendos. The only thing I can think of that you might be speaking of is the so-called "I* Metric." Mark McCormick-Goodhart (Aardenburg) says on his site that he invented it. But strangely, his 2004 paper describing it is also coauthored by both Wilhelm and Dmitriy Shklyarov. (It's called A "Retained Image Appearance" Metric for Full Tonal Scale, Colorimetric Evaluation of Photographic Image Stability.) Also, in McCormich-Goodhart's "About the Author" section, it says "From 1998-2005,he collaborated with Henry Wilhelm of Wilhelm Imaging Research, Inc to develop new image permanence test methods for evaluating modern digital output media."
Apparently the two companies, Aardenburg and Wilhelm-Research, have different goals, but I've meandered on long enough, I think.