Digital BW, The Print

I have just completed my first curve for Ilford Smooth Pearl, using 
Lyson Fotonic ink printed on an Epson 2100. Quite an effort !! I 
think I need to lie down for a week to recover. However I would 
appreciate some comments when evaluating the tone of a print and the 
corresponding Lab values, a and b. I have, I think a grasp on how the 
inks vary these values, and the effect that the paper has. However 
what do you look for when evaluating the figures ?

Just want to know if I'm doing it right.

Thanks

Peter.

Peter,

> I have just completed my first curve for Ilford Smooth Pearl, using 
> Lyson Fotonic ink printed on an Epson 2100. Quite an effort !! I 
> think I need to lie down for a week to recover. 

If you start from an existing profile, it should be a pretty quick 
process.  Of course, you are using a different inkset, so it would 
probably be a bit more effort.  Try starting with a profile that 
comes closest to your inkset and paper combination.  That should 
eliminate most of the work.

> However I would 
> appreciate some comments when evaluating the tone of a print and
> the corresponding Lab values, a and b. I have, I think a grasp on 
> how the inks vary these values, and the effect that the paper has. 
> However what do you look for when evaluating the figures ?

When evaluating a profile, I use my spectro to measure color (a* and 
b* values) at different points on the curve.  If I want a neutral 
profile, I try to get the a* abd b* values as close to "0" as 
possible.  I often read the color of the paper base before building 
my profile, and if it has a somewhat creamy or slightly yellow base 
color, I will often let that bias carry through from step 1 thru 26.  
But that is an artistic judgement call.  If you want all steps the be 
dead neutral, then that is another choice.  At any rate, if I measure 
my grayscale, and step 5 has a b* value of +2.5, but step 15 has a b* 
value of -3.2, then I have an obvious color shift going on.  If you 
don't want that color shift, it is time to adjust one or the other, 
or both.  Depending on the inks and curve shapes, you may be able to 
do this with ink limits, or more likely, by adjusting the curve 
shapes.  It's really pretty easy once you play with it a little.  
Make a tweak, print a new chart for linearization, then remeasure.  

I also try to get the densities for the various steps pretty close to 
the target values.  This way, the software applies smaller 
adjustments during the auto linearization process.  Most papers have 
a paper white L* reading of 97 to 99 (density of .01 to .03).  

Glossy papers have a Dmax that varies from L* 4 to 10 (density of 
2.35 to 1.95).  Matte papers have a lower Dmax, L* 15 to 20 (density 
of 1.72 to 1.52).  

If I am far from those numbers, I check to make sure I have a good 
Dmax and adjust the dark inks to get the best I can.  Once I have a 
good Dmax, you can enter the white and black points into the 
linearization screen, and IJC will give you targets for all the 26 
steps.  If a few spot measurements show I am way off, I will probably 
adjust my ink limits or curves to get closer.  If not, I just 
linearize.

This sounds like a lot of work, but once you have a good base 
profile, the others come very quickly.  If not, there is no way I 
could have created roughly 50 profiles in a month's period.  Once you 
have a good neutral profile for your ink and a matte paper, you can 
use that same profile as a starting point for other matte papers.  
Minor tweaks quickly deal with color casts and ink limits.  

Hope this helps.  Once you build a few profiles, it becomes pretty 
fast and easy.

Lou

Lou

First of all thanks for such an informative reply. The design of the 
Black and Light Black curves was pretty easy because as you pointed 
out I based them on the UC ink curves for the paper that I was 
profiling. I used the EyeOne to measure the L values and adjusted the 
shape and ink limits to get a nice smooth curve. That was the easy 
bit. The harder bit was tonning. Understanding the relationship 
between a and b values, and the fact that there were no suitable 
curves to use. The Fotonic inks are cooler than the Epson inks. 
Initially I tried to get both a/b values zero, but after a check 
print decided that the result was too warm, yellow for my taste. I 
guess this is because of the base colour of the paper. Ilford Smooth 
Pearl gives me values of 96, -0.15 and -5.36. So I cooled things down 
a bit with b values averaging -6. I have no postive to negative 
shifts in values. The values do increase slightly. The bigest 
increase is from steps 24 to 26. The b values are pretty constant 
ranging from step 5 -6.02 to step 23 - 6.86. The a values are not as 
good they range from 0.15 to 2.97.

Do these values look sensible to you. Would you tweak the a values ?

The print looks darn good. The best I have achieved yet. I really do 
like OPM/IJC. The more I work with the program and try to understand 
it the more I like it.

Once again Lou many thanks.

Peter.

Peter,

To me, numbers are simply a means to a desired end.  The measured 
numbers help me to quantify what is truly neutral, what color is 
causing a problem (e.g., is it a yellow or green cast), or where 
crossovers occur.  I let my eye be the final judge, but I do look at 
the results under a variety of lighting sources to be sure I don't 
have any weird shifts occuring due to metamerism.

For me personally, I design my "neutral" profiles to be a little cool 
for cool papers, and a little warm on naturally warm papers.  That's 
my judgement call and is what I like.  

Also, I tend to be a little less concerned about the deep blacks if 
they measure with a color cast, since I have a hard time seeing it.  
I try to keep it in line with the overall tone of the profile, but I 
don't obsess about it.  Having said that, my profiles are pretty 
close.

In Lab, positive a* values are more magenta, and negative a* values 
are more green.  Positive b* values are yellowish and negative b* 
values are more bluish.  Lab space isn't particularly intuitive, but 
you get used to it if you work with it for awhile.  So, you readings 
of Ilford Smooth Pearl (96, -0.15 and -5.36) show a very cool paper 
white.  My tendency would be to keep the b* value cool.  

For the a* value, if you are happy with the look of the print in 
various lighting situations, I'd leave it alone.  It is easy to get 
carried away with the numbers and forget that the final result we are 
after is a great print.  If you like it, it's done.

Lou

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "bwbonkers" 
<PeterDLevis@a...> wrote:
> 
> Lou
> 
> First of all thanks for such an informative reply. The design of 
the 
> Black and Light Black curves was pretty easy because as you pointed 
> out I based them on the UC ink curves for the paper that I was 
> profiling. I used the EyeOne to measure the L values and adjusted 
the 
> shape and ink limits to get a nice smooth curve. That was the easy 
> bit. The harder bit was tonning. Understanding the relationship 
> between a and b values, and the fact that there were no suitable 
> curves to use. The Fotonic inks are cooler than the Epson inks. 
> Initially I tried to get both a/b values zero, but after a check 
> print decided that the result was too warm, yellow for my taste. I 
> guess this is because of the base colour of the paper. Ilford 
Smooth 
> Pearl gives me values of 96, -0.15 and -5.36. So I cooled things 
down 
> a bit with b values averaging -6. I have no postive to negative 
> shifts in values. The values do increase slightly. The bigest 
> increase is from steps 24 to 26. The b values are pretty constant 
> ranging from step 5 -6.02 to step 23 - 6.86. The a values are not 
as 
> good they range from 0.15 to 2.97.
> 
> Do these values look sensible to you. Would you tweak the a values ?
> 
> The print looks darn good. The best I have achieved yet. I really 
do 
> like OPM/IJC. The more I work with the program and try to 
understand 

> it the more I like it.
> 
> Once again Lou many thanks.
> 
> Peter.

Lou many thanks.

Peter.