More Fade Test Results

[RogerB]

At the beginning of April this year I started light fade tests to see how some 3rd-party inks for the Epson Pro3800 compared to the OEM ink. Out of interest I also included a couple of OEM dye inks from Epson and Canon. The test samples were placed in a south-facing window and have been there now for 4 months. I have not attempted to measure or control the light exposure so I cannot compare my results to other peoples results.

The test sample for each ink/paper combination is a section of the Gretag-Macbeth TC918 profiling target. I have used the first three rows of the target which gives 51 colour patches. The patches are measured three times using an i1Pro and the measurements averaged. I then convert the measurements to real images (TIFF) using the LOGO ColorLab utility and import them into CS5. Using this method the colours are quite accurate.

By masking the patches of the original measurement and overlaying it on the latest measurements I get the image shown below. Each colour patch is divided diagonally, with the top left section showing the original colour and the bottom right showing the colour after light exposure. I feel that this method provides the best of both worlds, since the numeric data is embedded in the image and the visual change can be seen very clearly. Unfortunately the severe compression of the JPEG here has introduced some nasty artefacts, but the general effect can still be seen. I will post a high resolution version on my web site soon for anyone who is interested.

What is immediately obvious, and not really surprising, is that the two OEM dye inks are fading more quickly than the pigment types. What is surprising, to me at least, is that the Epson Claria is fading much more quickly that the Canon. This may well be because of the paper used. The Claria is printed on 7dayshop gloss paper (an "own brand" UK paper) while the Canon is on Canon paper.

Of the pigment samples, the first three are printed on Harman Gloss Baryta since this is my paper of choice. The Lyson ink is on Epson Premium Gloss. What surprises me here is that the Epson K3 ink seems to be fading slightly faster than the 3rd-party inks. The least fading is exhibited by the Lyson, followed by the IJF and then the ConeColor. All seem to be fading less than the Epson overall. There are visible differences in fading of different colours for different inks, so different images may fare better with some inks than with others. Confusing? Yes!

I am not trying to draw any definite conclusions from this. What does reassure me is that the 3rd-party inks are standing up very well on the paper that I chose to use for my "serious" work. Taking a different view, I ask myself if I would see any difference in a real print after the same light exposure. I suspect that all but the Epson Claria would still be acceptable, some might even say "identical to the original", but the Claria on this particular paper would probably be noticeably degraded.

I will be writing this up more fully at a later date, but thought that I would share these preliminary findings straight away.

 

[The Hat]

RogerB
Coincidentally I just finished a fade test with my 3rd party inks on three different papers left in the southern sun for 90 days.
The conclusion I came to and had always suspected is that fading is caused by both the inks/papers
being incorrectly matched together that can cause a bigger problem, not just the UV light.

I wouldnt put or leave anything let alone a print in direct sunlight that I cared about so
if the ink/paper combination that I use fades after only five years so what, Id just print another one.

If on the other hand I was selling prints commercially that's a different story altogether, then to protect my reputation
(if I had one) I would use only OEM inks and papers together to insure the best quality printed products..

 

[rodbam]

I've often wondered about the statements that OEM inks & papers have the word "quality or best" attached to them & the 3rd party consumables are considered to be close but not as good. Do all the OEM ink makers have top secret ingredients in their inks & papers, do they even make their own inks & papers? I think Canons Fine Art Rag paper is made by Hahnemuhle & you can use Hahnemuhle's own ICC profile with it.
I think too many myths get created on the internet & get a life of its own which most photographers seem to accept without questioning.
The obsession with archival properties seems to be the latest internet mantra whereby you are almost cheating a person by selling them a print made with dye inks, I imagine these same photographers were quite happily selling dye ink prints before pigment inks became de rigueur. A lot of this seems more about the ego of the photographer than the worry about the person buying their prints, I'm not talking about the great photographers selling prints for astronomical prices that the purchaser looks at as an investment but the run of mill stuff that sells for 100s of dollars & most of the buyers of these prints probably wouldn't care less if the print has faded in 10 years time as they will probably be sick of looking at it by then. It appears to me that the run of the mill photographer is the only one who is worried that their precious work of art will last for an eternity as we all live in a throw away society & a $300 or $500 print is the least thing we will be worried about in 10 or 20 years time, yes dye inks last that long.
I have a dye ink printer & if I did sell a print & the word archival did come up I would just stick a CD with the print file on it to the back of the print..............With acid free tape of course

 

[Filemonster]

Please don't take this criticism the wrong way because I applaud your efforts here RogerB. I have been through training for color control and been responsible for color control of a massive international project so I know how difficult it is.

The premise of your test (unless I misunderstand it as stated) is faulty, you must print all of the inks on the same paper, not some on one brand/type and others on another brand/type. If you want to observe differences using other paper types, you would need to print all the inks on all of the other paper brands/types to obtain meaningful results.

The ink is not totally opaque and so the substrate has quite a bit of influence on the fading issue.

Just trying to undertake a project like this is a massive undertaking both in financial and physical terms as well as the logistics, and furthermore would never be complete because each ink will have different results (in many ways not just "fading") depending on the specific formulation of each paper brand and type.

In addition (I realize this is off topic a bit and not directed at your test) as mentioned in the thread, I would tend to think that the accuracy of the color as initially printed, with all its' qualities such as hue, brightness, saturation etc. would be more important in most applications. Then again if you are targeting commercial applications I would also stay with OEM inks and paper stock as it will be virtually impossible to know the long term effects of a given ink and paper combination unless you had access to specific testing results.

Again I appreciate your efforts, but your test needs to be redesigned to be of real value/help.

Regards,
Charlie

 

[Filemonster]

By the way folks, printer manufacturers work with both paper manufacturers and ink companies when testing paper handling and how the inks react. They need to decide a range of applications for that printer which is why the have different product ranges such as office machine, photo machines etc. Then they can specify which papers are to be used and then work on the ink formulations.

They don't need to be the actual producer of the ink or paper. It's just like auto manufacturers work with tire companies and shock absorber companies etc. An auto company doesn't make most of the parts in the car. They design the various parts either on their own or in conjunction with suppliers, they set specifications and tolerances etc. and they may even have people at the factory of the supplier to insure quality and production levels. This is the way most manufacturing is done today.

There is a lot of testing involved before the introduction of a new product comes to market, even if the perceived or actual changes appear to be minor. Companies that fail to do this testing risk having inferior products come to market and that can harm a lot of effort that went into generating brand value. This is one of the things that separate good companies from the chaff and is why they can charge more for their products.

Charlie

 

[RogerB]

The premise of your test (unless I misunderstand it as stated) is faulty, you must print all of the inks on the same paper, not some on one brand/type and others on another brand/type. If you want to observe differences using other paper types, you would need to print all the inks on all of the other paper brands/types to obtain meaningful results.

You have misunderstood slightly. The first three samples are printed on the same paper so I believe the results can be compared directly. This is the paper that I use for all my work so this was the main objective of the test. The others were included, as I said, just for interest.

The ink is not totally opaque and so the substrate has quite a bit of influence on the fading issue.

Yes, the substrate does affect the result since the substrate colour contributes to the overall colour of the lighter tones. But, as I said, the first three samples are all on the same substrate. I should also add that the substrate colour is measured as part of this test.

Just trying to undertake a project like this is a massive undertaking both in financial and physical terms as well as the logistics, and furthermore would never be complete because each ink will have different results (in many ways not just "fading") depending on the specific formulation of each paper brand and type.

Not that massive. I am simply trying to compare three ink types, on the same substrate, subjected to the same light exposure.

In addition (I realize this is off topic a bit and not directed at your test) as mentioned in the thread, I would tend to think that the accuracy of the color as initially printed, with all its' qualities such as hue, brightness, saturation etc. would be more important in most applications.

The accuracy of the colour is only important when printing real images. For real printing all of these ink/paper combinations would be profiled and all would give very acceptable colour accuracy. The reason that I have used the colour patches for this test is that I have generated custom profiles for all of the ink/paper combinations so I know that they are all capable of giving good colour accuracy.

Then again if you are targeting commercial applications I would also stay with OEM inks and paper stock as it will be virtually impossible to know the long term effects of a given ink and paper combination unless you had access to specific testing results.

Staying with OEM ink and paper is too restrictive for many people, myself included. There is tremendous interest in the performance of third-party papers and inks, as demonstrated by the popularity of the Aardenberg Imaging work. I think you will find that even the best commercial print specialists use third-party papers, and many use third-party inks.

I use Inkjetfly ink because it gives me a slightly wider gamut than Epson ink, at a much lower cost. As far as paper goes, the Harman Gloss Baryta is the closest I have ever found to a "traditional" paper and that's why I use it. Epson doesn't have a paper to match it in my view.

Again I appreciate your efforts, but your test needs to be redesigned to be of real value/help.

Regards,
Charlie

Charlie, I agree that the last three samples in my test are less relevant than the first three. However, if anyone is using those particular ink/paper combinations I think they would be of interest. The first three are of particular interest to me, and I maintain that they are giving me useful information. I also maintain that my method, which is based on the method used by Aardenberg Imaging, is far more sensitive and reliable than the subjective methods used by most people. Looking at a print and deciding that "it looks as good as when it was first printed" does not convince me.

I should perhaps add that I worked as a physicist for forty years and since my retirement I help people with colour management, especially with printer profiling. So, I am under no illusions about the absolute results from my tests - I can't quote deltaE changes after x Magalux.hours like Aardenberg - but I am happy that they give good comparitive results.

Regards

Roger

 

[The Hat]

RogerB

Staying with OEM ink and paper is too restrictive for many people, myself included. There is tremendous interest in the performance of third-party papers and inks, as demonstrated by the popularity of the Aardenberg Imaging work. I think you will find that even the best commercial print specialists use third-party papers, and many use third-party inks.

No commercial printers or high street digital copy shops that I know of would even consider using anything
but OEM inks, they wouldnt stay in business long it they did..

 

[RogerB]

I wasn't really talking about "high street" print shops, but fine art printers. Take Cone Editions as an example; their web site says that they have moved away from Epson pigment inks, which contain a smalll amount of dyes, to their own Conecolor 100% pigment inks. Check it out here http://www.cone-editions.com/epson.html

It is clear from forums like the Epson Wide Format Group that other commercial printers are using third-party inks for fine art printing. Admittedly they are predominantly in the USA, but that's where most of the fine art printers seem to be.

 

[ThrillaMozilla]

RogerB, it's very nice to see information like this posted. You might have fewer confused comments, however, if you were a little more explicit about the ink and paper types in each test. For example, you compare the OEM dyes with the pigments, but you don't tell us which are dyes, and which are pigments. And again, you write, "Of the pigment samples, the first three are printed on...", but the user has to do some investigation to figure out which those are. It may seem obvious to you, but not everyone is intimately familiar with all brands and types of ink and paper.

 

[Filemonster]

Roger,

Thank you for the clarifications and corrections to my comments. I agree that if the aim of your efforts is for personal knowledge of these paper/ink combinations for your personal use then you are OK, and correct that it won't be a really big undertaking.

Yes your method is far more professional and will give results that are truly useful for the conditions you stated, as opposed to the subjective testing many others do which are not useful at all.

I would suppose that commercial printing companies, unless they have done extensive testing of third party paper/ink combinations would not dare use them if they value being in business and are professional. On the other hand I do understand the very strong desire (I have it as well) on the part of some folks to find a less expensive alternative to OEM inks. In that case I advise doing your own correctly-designed testing. The problem will be that even after you have your results you won't know if they hold true for the next batch of ink.

I can tell you that after having been in charge of a large multi-year international project where color control from batch to batch was of critical importance it is not easy to control color, gloss level, and light fastness etc. Even the smallest detail like the particle size of the TIO2 for example can have drastic consequences. There are UV inhibitor packages and UV absorber packages along with quite a few other components that go into the various formulations. How will we know as end consumers if any of these components have been changed either on purpose or by mistake? We can't know, so the only protection we have is to do vigorous testing on each and every batch of every color we obtain. Is anyone going to do that, I doubt it. Are the ink resellers doing this testing? Does a commercial printing operation have the time let alone the ability to do this testing?

What are we to do? People buy OEM inks because they know the OEM ink companies do this testing and watch over the production very carefully, even if that production is contracted out to ink manufacturers. If you have profiles set properly the actual colors may close enough for your purposes even with the production variances that occur, that's fine, but the most expensive components are the UV additives and that is where the greatest risk lies for the ink manufacturer to "cheat" us or "make a mistake" and is for me the greatest area of concern. As I mentioned earlier the only way to be sure of the performance of any given batch is to test it.

Sorry for the long blab here and it is not aimed at you, Roger, just giving folks something to think about.

Regards,
Charlie