Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "john" <deanwork2003@...> wrote:

> The  beauty of your method is that you can SEE what is  going on in real time right on the website by  observing the changes in all the hues and values at specific intervals. And anyone else can submit the same ink/paper combination to back that those results up.
> 

Yes, although it helps to have a calibrated monitor, and even then the monochrome target pushes the limits of typical display technology, but nonetheless, I felt that resaonably accurate colorimetric reproduction of the fading results in the test reports would be helpful for everyone including myself. I miss that confirmation in the numbers that we get from the industry-sponsored tests.


> So, is  Wilhelm's end point a fade rate of aprox 35% as I believe Jon Cone has suggested in the past?  I mean he's got to have some objective criteria other than - oh well  my 60 something year old eyes see a definitive change therefore we stop here? What percentage is Wilhelm using to stop the test and declare it an end point. I still don't understand that.
> 

As I noted earlier in my previous remarks. the WIR 3.0 densitometric criteria set has numerous individual endpoints. The one that triggers first is the limiting factor that stops the test. However, merely testing for the limiting weakest colors can be misleading when it comes to total overall performance, so both weakest colors and average overall response should be evaluated in any thorough test.  This is a basic testing principle that so far been lost on many of the researchers in the field, ie. the notion that the system's weakest colors may not even exist in some image content, so we need to test for more than just the limit factor and express the total system rating as an exposure range rather than a single value.  

I actually misspoke earlier by saying there were 13 individual endpoint criteria in the WIR endpoint criteria set. Actually, it's 17, and you can get the itemized list here:

http://cool.conservation-us.org/coolaic/sg/emg/library/pdf/wilhelm/wilhelm_paper_feb_2002.pdf

Also, bear in mind that when it comes to monochrome print testing the number of relevant patches in the WIR test boils down to just three: media white point, 0.60 density neutral, and 1.0 density neutral. No full tone scale to exercise the blends nor any evaluation of dmax.  Recall that the original test design was for chromogenic color prints. One can infer the behavior of the whole tone scale by tracking just these few patches when evaluating traditional chromogenic color prints like Kodak Edge or Fuji Crystal Archive. Not so with modern inkjet which is why I concluded a completely new test design was needed. And hence....the I* metric and method was conceived, and AaI&A became the vehicle to get the new testing started.

Yikes, this thread has gotten so long, I'm having a hard time following all the branches.
I think its time for me to finish!

cheers,
Mark
http://www.aardenburg-imaging.com

> 
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Mark" <mark@> wrote:
> >
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > --- In DigitalBlackandWhiteThePrint@yahoogroups.com, David Kachel <david@> wrote:
> > 
> > > On a semi-related topic: I have seen "Wilhelm years" mentioned here several times and could use a reference to an explanation. Apparently Wilhelm's tests are being "translated" into some other number that in some cases seems to imply that prints are virtually turning to dust as they issue from the printer <grin>. Anyway, it makes it difficult to compare the numbers for some of the B&W ink sets discussed here to Wilhelm's numbers for OEM inks. The implication that Wilhelm's results are overly optimistic leaves me with no point of reference.
> > 
> > True "Wilhelm display life years" involve two major factors that most people simply don't take the time to understand (i.e., they are happy just to accept the predicted years at face value):  1) the WIR display prediction is derived using a "normative" light level in the display environment of 450 lux for 12 hours per day, yet real world display light levels can actually vary around this normative level by three orders of magnitude.  2) A criterion to decide "end of life" is required (i.e., are we talking about a little fade,  moderate fade, a huge amount of fade, or image totally gone!). WIR describes the endpoint of the test as "easily noticeable fade". To understand just how easy is easy and where these densitometric criteria go wrong with modern digital printing media, one needs to study the WIR 3.0 "visually weighted" endpoint criteria set which has 13 different threshold values.  Unfortunately, even if you take the time to understand how the ratings are calculated, the determining threshold factor is not published, so we don't know how the sprint sample is actually failing. The specific failure mode can have significant bearing on what system is best for a particular image. For example, a print that tends to show highlight yellowing (e.g., loss of optical brightener activity) would be more problematic for the white dress in a bridal portrait, not nearly so visually apparent in a dark and brooding landscape scene.
> > 
> > When Paul described the AaI&A rating for the K6 neutral sample as 18.5 Wilhelm years, he was using the AaI&A table in the report in order to convert the 37 megalux hour rating into "years on display" by selecting  the same WIR value of 450 lux for 12 hours per day. Thus, the 18.5 year AaI&A prediction at that light level is on the "WIR display time scale", but it definitely does not use the WIR densitometric criteria set to establish the amount of fade which occurred in test up to that point. The AaI&A Conservation Display ratings are based on "little or no noticeable" fading rather than "easily noticeable" fading, and they use a colorimetric rather than densitometric model for judging changes in image color and tonal appearance. Thus, the apparent disparity in ratings between AaI&A and other testing labs is caused not so much by the basic accelerated nature of the testing, rather the overall criteria used to define how the scores are achieved.
> > 
> > It's important to realize that the current industry-sponsored tests were developed originally for consumer-oriented photofinishing applications.  I can assure you that the types of hue and tonal nuances that are being debated between ABW and other monochrome inkjet printing methods in this forum are way too subtle to have any meaningful impact on a consumer-oriented "easily noticeable" fade rating. Thus, "easily noticeable fade" is, IMHO, too liberal a fading endpoint for fine art, but it's a totally valid goal to evaluate how a print is faring in mid-to late stages of deterioration. Because systems often fade non linearly, early stage fading may not be a good indicator of late stage deterioration and vice versa. Full system characterization needs at least three points to define the fading curve which is why AaI&A reports results at multiple exposure intervals and doesn't stop the tests when the AaI&A conservation display limits are reached. The AaI&A conservation display ratings estimate early stage fading performance only.  
> > 
> > Lastly, I want to add that I personally don't view print permanence testing as a game of "winners and losers".  It's about helping the end-users of various processes to make informed choices.  As others have eloquently said in this forum, the artist should first and foremost choose a process that imparts the aesthetic he or she visualizes for the created image. I've seen beautiful ABW prints, and I've seen exquisite B&W inkjet prints made with single channel, three channel, four channel, on up to seven channel monochrome ink sets,  but in each instance the artist made intelligent choices in image selection, image edits, ink, paper, coatings, etc.   Sometimes, the best choice involves a compromise in print permanence, but sometimes the best choice also happens to be highly durable as well.  We can certainly ignore print permanence testing and lull ourselves into a false belief that if it's a "pigment print" it must be "archival" (whatever that means), or we can take the time to discover the print durability strengths and weakness of our chosen processes. Adding print permanence information to the list of variables that should concern the artist helps to cut the bull$*#& and to choose wisely... even if the final choice is to use a fragile but aesthetically beautiful process.
> > 
> > cheers,
> > Mark
> > http://www.aardenburg-imaging.com
> >
>