Image stability summary

[Photo Engineer]

There has been a lot of argument and disinformation about image stability floating around. I thought I might summarize some facts. The data below is generic to B&W photos, color chromogenic (analog), dye bleach (Ilfochrome) and digital printing.

Image stability tests fall into the following classes which are rather simplified here.

Type I

1. High intensity, high temperature, high humidity
2. High intensity, high temperature, low humidity
3. High intensity, low temperature, high humiditiy
4. High intensity, low temperature, low humidity

Type II

1 - 4 above with medium intensity

Type III

1 - 4 above with low intensity

Type IV

1 - 4 above with ultra high intensity

Type V

1 - 4 above with no light

Type VI

All of the above but with atmospheric contaminants including ozone, sulfur dioxide, hydrogen sulfide and other common pollutants.

In addition, these can be done with simulated daylight or tungsten as well as flourescent, wich further increases the breadth of the experiment.

This involves at least 20 x 4 minimum conditions if all tests are done. It requires a large light cabinet with calibrated light source.

Most testing companies run only one of the clasess above, or mixtures of one with another to try and damp out fluctuations or any reciprocity effect. This reciprocity effect is the failure of a high intensity, temperature or humidity to project back to ambient conditions for one reason or another, such as the diffusion rate of atmospheric oxygen or other effects.

As a result, almost every company testing these materials differ on the effective lifetime of their products. And, since products continue being improved, it depends on which generation is being tested. For example, since one of these tests might take up to 2 years to complete, by the time the results are published, the product no longer exists or has been upgraded substantially.

In the end, the fade results are judged based on ANSI standards which are presently undergoing revision. The standard ANSI test measures several factors but the basic measurement is a JND (Just Noticeable Difference) in which the observer(s) first detect a change in density or a shift in color.

Museums generally are kept low light constant temperature and humidity conditions. Homes are low light and varying temp and humidity, and offices are generally high light and constant temp and humidity (depending on the availability of air conditioning).

At Wilhelm and Fuji, both publish data from high intensity light and several conditions of temperature and humidity. Kodak tests at both low and high conditions of light, temperature and humidity and publish data from the low light. RIT uses as many of the above as possible but add atmospheric pollutants. Recent tests from Kodak, Wilhelm and Fuji have begun to include those pollutants. (I actually don't know which organization first started the testing with air pollutants, so the sequence may be reversed here)

Kodak tests for average home and museum conditions, while Wilhelm and Fuji test for office conditions. Therefore, data differs.

Kodak had been urging the ink jet companies and Wilhelm to test for dye smear in inkjet prints due to their experience with image transfer images. It was not until recently that those tests have begun to appear on the Wilhelm web page, but they do not feature a prominent position. Analog prints do not smear unless done by some diffusion process. Exceptions exist such as dye transfer which was not known to smear due to the unique mordant.

The Kodak justification for testing at low light levels comes from measuring picture storage conditions in over 10,000 homes in just about very corner of the world. This test took years to accomplish, but gave them a base line for the biggest customer group. Fuji and Wilhelm based their tests on business and advertizing customers which is the most visible.

The bottom line is though that the complete battery of tests needed is huge and very very costly. Most companies only use one or two types and therefore end up with differing results.

Due to the rapid evolution of digital, some data would suggest that currently hanging prints in museums will last less than 60 years, but prints going up now will last longer (dye digital), and the data suggests that the current chromogenic prints will last 100 - 200 years.

Pigment inks are different.

Here: http://www.macworld.com/weblogs/creative/2006/12/inktype/index.php

you will find this quote:

"Right now, many dye-based prints, kept under glass and away from direct light, can last for up to 25 or 30 years, which is more than adequate for most of us."

Which is less optimistic than my comments above, but again it depends on conditions.

A pigment ink is different than a dye ink. Going by common defintions, pigments are made from heavy metals and are used in paints, so we have red lead, cadmium yellow and etc. Since no manufacturer of print materials publishes information on their dye sets, we cannot say with certainty what these pigment inks are comprised of in detail, other than the fact that they contain no dyes. The article quoted and referenced above gives pros and cons, but since we aren't considering digital this is just incorporated here for reference.

Kodak and Fuji both upgraded their chromogenic papers within the last few years such that the expected lifetime is > 100 years again depending on who is doing the tests and how they were run. There is continual work on these upgrades which will continue to give analog prints longer and longer lifetimes. It is probably that the digital world will catch up with analog in stability, but in terms of image smear it may be a while before digital catches up.

References include:

"Image Permanence: Understanding, Measuring and Predicting Print Life" by Jon Kapecki.

ICIS Proceedings '02 for improvements in Kodak Endura paper dyes

ICIS Proceedings '06 for improvements in Fuji CA paper dyes.

The articles in the last two references include structures of the new dyes for those 'geeky' enough to want to see them.

B&W image stability is becoming less and less of interest, but the article by Ctein is by far one of the best on this subject.

I hope this helps you all.

PE

 

[Mr Bill]

Thanks PE. This is a topic near and dear to me, and I have a few comments to add.

In the interest of simplification let me say that I think of image stability initially in two categories, "light stability" and "dark stability". Then a third category would be for the special conditions or contaminants. So the image degradation is broken into 1) damaged by light energy, 2) an inherent chemical degradation of the dyes or substrate or 3) some external source of degradation.

Using these, your "Type I-IV" would be under the category of light stability. "Type V" is dark stability and finally, "Type VI" is the special condition/contaminants. The special condition/contaminants is for things which we do not explicitly expect. For example, we don't always expect images to get wet, or stay in high humidity, or be exposed to high levels of ozone or sulfur compounds, etc. But if the images ARE sensitive to those things, we'd like to be aware of it. For example, most books or magazines have a big problem if dropped into the bathwater, yet most people are ok with that; they know not to let it happen.

The dark stability tests generally use what is known as Arrhenius testing, see Wilhelm's book (available free on his website), Chapter 2, pages 85-? He discusses the initial application to photography by Larson of Kodak. Additionally, the method is explained with a sample graph reprinted courtesy of Kodak. The method typically uses a range of five or six temperatures (one sample for each temperature) in the general range of 140-200 deg F, holding humidity constant. So testing a different humidity value requires another complete test run. On completing the test sequence,one can extrapolate how long it would take to reach the specified end point under different temperatures.

Wilhem has a very good overview of current (2004?) test methods here, including some history of the ANSI/ISO standard AND some discussion of light levels used by various testers:

Dead Link Removed

BTW, regarding your references, it looks as though Kapecki's document might be strictly handouts from his tutorial? Or is there a publication available? Also, can you say a bit more about Ctein's article (title or whatnot) ?

Thanks!

 

[AgX]

Rita Hofmann (head of Ilford R&D before the splitting) puts emphasis on Ozone. She called Ozone the the most detrimental factor to porous ink jet materials.

Concerning chromogenic prints she considers the PE layer the limiting factor, giving it 50 to 80 years.

(I’m referring to a script of a lecture she held in Barcelona in 2003)

 

[Photo Engineer]

Bill;

I agree with your statements.

One fact you must recognize is the reciprocity effect in Arrhenius tests. Diffusion of oxygen for example, affects dye stability, but diffusion is non-linear with temperature and therefore has a built in reciprocity. There are others that I could cite, but this is enough.

The Kapecki booklet is available for about $80 US, from ICIS. At least it was last I looked. It is packed with information.

The Ctein article is referenced (IIRC) on his web site. I can't find my reference here but I'll try to get it for you. He did B&W stability tests after having some complaints about fading and had a lot to say on the effects of keeping conditions on B&W. It is probably one of the better articles on B&W stability recently.

I would like to add additional references such as:

1 hour tour of the RIT Image Stability lab.

4 hour conversation 1:1 with Henry Wilhelm, and individual for whom I have great respect.

I hasten to add that it is difficult for him or anyone to have 'THE' answer to image stability. What is the definition of a 'bad' print? One that has changed color uniformly, one that has spots, one that has just faded uniformly to a certain degree? These are questions asked by all dye stability researchers. I have seen prints in my tests that went nearly totally blank at the end of a very long test. But, in real life, prints on this material still survive on my wall in full illumination and ambient home temperature and humidity.

So, tests are not everything.

PE

 

[PhotoSci]

Kodak has long advocated that any meangingful attempt to estimate print lifetime be based on the four primary factors: heat (dark stability), light, humidity, and pollutants (most notably ozone). And in fact, its papers on image stability going back at least 20 years included the first three. (Until the advent of ink jet, ozone had not been a primary degrading factor. Silver halide dyes are well-protected by their gelatin/organic phase matrix. When ink jet started to become impoprtant, Kodak proposed the addition of ozone testing to the repertoire.)

But equally important for meangingful estimations is that the four factors be in balance relative to the way they are likely to be found in the real world. For example, if you peg light at a very high level while keeping the other factors at the more probable values, you will not only predict an incorrect estimate for lifetime (based on first mode to fail), but you might well rank a truly less stable product as longer-lived than a more stable one.

Ozone is usually the fastest of the tests to perform, by the way, so be suspicious if you see light/heat/humidity data reported, but ozone "test still in progress." One interpretation might be that the product didn't do so well in that test.

Other pollutants, like sulfur or nitrogen oxides, may become significant as research develops, but so far ozone is the worst actor. In environments where indoor kerosene stoves are used, sulfur and nitrogen oxide effects are seen.

Humidity may be the hardest test to do. It has at least two effects: colorant degredation and smearing--and in human perception they are interactive. Further, extrapolation from high test humidities to lower real word values are not simple, due to phase transitions and other factors.

One last point, suggested in the paragraph above. It is very important to do human factors (a/k/a perceptual studies) to make sure that what you measure correlates with what you see--the eye-brain is a tricky recording device. Too often people publishing test results skip that expensive, but critical link, with the result that their predictions do not correlate with the human visual evaluation.

 

[z-man]

looking in the light for what you lost in the dark?

all the above notwithstanding-

the paint industry has been reacting to customer complaints of fading for several hundered yrs-they have data on pigments and dyes

the issue of smearing is closely related to dot gain in the printing industry and they also have data re pigments and dyes

i can see from many posts that there are many from both industries here among the membership-i hope this will catch the eye of those who can give us the benefit the experience and data from those and other industries can provide

mechanised gum printing of fabrics by the brits goes back 300 yrs-printed fabrics undergo the most stress re fading i can think of

kodak's gathering of data by polling customers goes back to the begining and we have the benefit of pe and his close association with kodak

so how about the rest of you out there?-i owe pe for pulling my coat to the latest data on ultrachrome aging-not what i wanted to hear but what i needed to know

vaya con dios