More Fade Test Results

[Filemonster]

This response is not meant specifically toward Cone Editions, they may do fabulous work, I have no idea, however that is marketing talk on their site. I am sure that like so many others the switch away from OEM ink is driven by cost. They say they switched to 100% pigment inks for even better longevity. Firstly ALL testing for colorfastness is artificial so a rating of 100 years or 75 years or whatever is only an estimate. Secondly when have you ever seen the time period connected to a Delta E value of color shift? Who will care or know if the color has shifted, degraded, changed or whatever in 75 or 100 years.

Aside from that the colorfastness of an ink or stain or paint is determined by other additives, not just whether or not it is dye-based or pigment-based. It is easier to have a wider color gamut with dye-based ink than with Pigment-based ink and that is probably the reason Epson has some dye in their formulation. You can achieve excellent results with either approach and I have yet to see any definitive test results that show a clear advantage of one type ink over the other.

Charlie

 

[Filemonster]

I wanted to let folks know something else about testing for "fade" in the sun. I think it is probably obvious to everyone that the angle of sunlight is different depending on your location. Severe testing is done using closed containers with glass panels in the Arizona desert in the USA and also in Australia. If a product can pass testing in Australia along with instrumented testing it will be fine to use anywhere.

My point is, an ink that has performance deemed acceptable in England or Germany or Canada for example, may fail horribly in Florida or Spain or some other location, while another ink tested side by side that "seemed" to perform comparably to the "acceptable" ink may still perform well in those more demanding locations.

This why we used a minimum of 1500 hours in a xenon arc machine with specific filters to match as closely as possible natural sunlight. The colorfastness must not be discernible from the original to the eyes of trained professionals under controlled conditions or to a photospectrometer or it is deemed a failure. This test is far more reliable and accurate than a test in a QUV machine which is not able to create nearly as realistic a test, the available bulbs just have too small a spectrum.

Charlie

 

[qwertydude]

Even a xenon arc lamp won't recreate all the damage possible. Ozone testing is a major test done by the big labs like Wilhelm Imaging Research. Sometimes a UV stable ink will badly fade when exposed to ozone. Ozone is usually present in smoggy cities, or also in places where a lot of high voltage equipment is used or hospitals with ionic air cleaners.

 

[RogerB]

ThrillaMozzilla
Apologies for the lack of clarity. For anyone else who is confused, the top three samples are all "K3" pigment inks printed on Haraman Gloss Baryta paper. First Epson, second Inkjetfly and third Conecolor. The fourth sample down is Lyson K3 (pigment) ink on Epson Premium Gloss. The last two samples are dye inks - Epson Claria on an "own brand" paper and then Canon ink on Canon paper.

Filemonster
Your comments about product consistency from batch to batch are very relevant. I do check the colour when I top up my cartridges and I have not seen any significant changes as yet. Not a rigourous test, I agree, but one that will show up any changes in pigment concentration etc.

Of course all testing of light fading is "artificial" - any accelerated life testing is artificial - but that doesn't mean that it's not useful. As far as light fading is concerned, there is good evidence to support the premise that fading is dependant on exposure i.e. Intensity x Time. You ask where I have seen DeltaE changes documented. They are fully documented, not against time but against exposure, on the Aardenberg Imaging web site. Take a look at http://www.aardenburg-imaging.com/ If you look at the test results you will also see evidence of the more rapid fading of dye inks compared to pigments.

No accelerated life testing is absolute but it can give pretty accurate predictions of performance in many cases. And as a comparitive tool, it can definitely help to rank the performance of products tested uder the same conditions.

quertydude
It's certainly true that light fading is not the only mechanism at work, and the effects of ozone can be very severe. But, again, all the theory and the evidence points to the superiority of pigment inks in withstanding the effects of ozone. It is also clear that ozone fading of dye inks is less if swellable coatings are used, but that's another story.

 

[Grandad35]

Here is a 5 year old report on fading that touches on some of these subjects.

 

[RogerB]

An interestng report that clearly shows the difference between dye and pigment and micro/nano-porous and swellable papers. At the time that report was written I was using Lyson dye inks in my Canon S9000 with Ilford Classic Gloss. The results were very good, but prints still faded too much for my liking so I switched to pigment.

The fact that pigments fade much more slowly led me to look for an analytical method of testing that would be more sensitive to early changes in colour. Hence the method that I described.

 

[PhotoSci]

It is true that a pigmented ink is no guarantee of superior image permanence, but it increases the likelihood. A critical factor is the stability of the inherent colorant molecule (to light, ozone, thermal degradationall of which may be different) independent of the way it is dispersed. However, once that molecular stability is established then dispersing that colorant in pigment form will almost always improve its stability, sometimes very significantly.

Predictive image stability testing which purports to provide a years rating is so fraught with certain and potential errors that it should be viewed as only the very roughest of guides. Not only is there the significant errors that accumulate in the test procedures, but then the greater errors inherent in both the environmental and assessment factors cause any prediction of years to be close to what engineers call a WAG (the G stands for Guess). If you doubt this, I suggest that you read Peter Masons paper in the Journal of Imaging Science and Technology (Mason is a chief scientist at Torrey Pines Research, where they do image permanence testing for a living.) Anyone who lists a years prediction without large error bars is either nave or dishonest.

It is true that a product rated at 25 years is very likely to be less stable to the stressor being tested than one rated at 150 years, but trying to compare a product claimed to last 106 years vs. one said to last 143 years is probably close to meaningless.

Thats why organizations like Aardenburg Imaging prefer to report (for light fade) a megalux hours exposure to a given fading endpoint. That number still contains significant experimental errors, as Aardenburg would freely acknowledge, but does not heap upon that additional errors and assumptions.

Xenon filtered to mimic sunlight is not the ideal light source for light fade testing as it is not what most indoor images see. Rather they see window-filtered daylight which is altered by window glass, reflection, and absorption on room surfaces. This has been documented in the literature by monitoring light in homes around the world (check out Bugner, et al. in JIST). Xenon is a good start, but it needs to be filtered to more closely mimic that light spectrum.

Its also important to test for degrading factors other than light. Ozone was mentioned. To that should be added heat and humidity. Well over 90 per cent of images are stored in the dark where light is not a factor. As yet, no one knows how to combine these in any meaningful way.

Ozone tests are relatively quick to perform which makes one wonder why some laboratories (such as Wilhelm Imaging Research) still display reports where the ozone test is said to be in progress years after the light tests are completed.

The bottom line is that meaningful predictive testing is a difficult task. The wrong conditions or assumptions can give not just quantitatively wrong answers, but actual inversions of relative stability for real world use.

 

[Filemonster]

qwertydude
Even a xenon arc lamp won't recreate all the damage possible. Ozone testing is a major test done by the big labs like Wilhelm Imaging Research. Sometimes a UV stable ink will badly fade when exposed to ozone. Ozone is usually present in smoggy cities, or also in places where a lot of high voltage equipment is used or hospitals with ionic air cleaners.

This is true and I didn't mean to give the impression that there aren't any other factors involved, but the jist this thread was started with was in regard to how light affects the ink/paper combination so that is what my comments were aimed at.

RogerB
Once again I find myself congratulating you on your personal efforts and care. Please accept them as true compliments, I am not trying to be sarcastic. I am a bit of a perfectionist myself and many others think I am being careful to the point of being anal, I really do admire your organization and care. I agree with you that the artificial testing we have access to is of great value. The "fading" issue is absolutely related to exposure, intensity, and time as I have written in a previous post, I am in agreement with the folks at Aardenburg. The testing we have done has shown dye-based materials to be inferior to pigment-based materials as well. I was simply trying to point out there are other factors related to the stability than just the colorant itself. We have seen some dye-based products fare better than other pigment-based materials.

PhotoSci
Your first statement is exactly the point I was trying to make, well said. I wouldn't go as far as you in saying the testing is fraught with certain and potential errors, however I do agree it's not perfect. Yes the proclamations of "x years of stability" are a guess but at least they are relevant to compare one sample to another within the same test cycle, the same idea as RogerB is trying to accomplish with his testing. Aardenburg is doing the same basic thing as most of the test labs do as far as I can tell; expose the material to a light source at a certain intensity, distance, for a set time under set circumstances such as humidity, temperature etc. this is standard procedure. There are of course many factors to be aware of such as the light output reduction over time of the test light source etc. I am not questioning them.

There are simply too many variables to take into account for test purposes of all the indoor lighting circumstances so a standard must be used. That standard is sunlight and the best reproduction of sunlight is the xenon arc machine with proper filters. If someone wants to pay for a specific test to mimic the light source their material will be exposed to that's fine but the results will be of limited use to the public at large.

I don't disagree there are other factors involved, I think we all know that, but the subject in this thread was the factor that light plays. At the end you state the same thing I have said previously, that testing is difficult and it is expensive when done comprehensively and correctly.

 

[PhotoSci]

Filemonster
How many people display their prints in direct, unfiltered sunlight? And whose sunlight? Sorry, the best accumulated data for home display is not sunlight, but rather a much modified form of that power spectrum. That's where people hang their prints.

And saying that it doesn't matter as long as all materials are tested under the same light source is not correct. Different colorants have absorption peaks at different wavelengths with different bandwidths. They will not react to different light sources in the same way, often not even relatively. You need to come as close as you can to what images really "see" in the home. Compare the power spectrum of sunlight to the average found for home studies world wide. Will you be spot on for everyone's home? No, but you'll be a lot closer.

Other evironments require different light sources. For example, offices are best simulated by cool white fluorescent. Most museums attempt to use light sources that limit UV. It all matters.

Sorry, we'll just have to diagree on this one.

 

[Filemonster]

PhotoSci I don't think we disagree, you are missing the point I was trying to make. If you are going to display something inside with specific light and environmental circumstances then yes of course you should try to set up the test as close to those circumstances as possible, however in many cases that is either not practical or too expensive or not possible for other reasons.

As far as I am aware there isn't a standard to do testing of indoor situations with all the various light sources, if there is and I am not aware of it please let me know. That's why the standards that do exist for lightfastness are done relative to sunlight. Indoor situations include so many factors like the wall colors, type and gloss level of the paint on the walls, floor surface, ceiling surface, room size, temperature, humidity and so forth.

You are correct that different colorants react to different wavelengths of light in varying ways is correct. I never said they don't. Of course nobody displays things in direct, unfiltered sunlight, but direct unfiltered sunlight will most of the time be stronger than indoor lighting and have a much wider spectrum to boot which will include the spectrum you will encounter indoors. If something can stand up to sunlight it will in most cases do just fine indoors unless you have some unusual circumstances.