Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "gulstenek" 
<keving@d...> wrote:

> 1. Are there more systematic approaches to partitioning these quadtone 
> inks than than the recursive print and twiddle that I use now?



Kevin,

it's always good to use an existing profile as a starting point. IJC ships
with several. Use the one closest to your paper/ink/printer combination.
Sometimes all you have to do is grab an existing profile, print the
target for linearization, read it in, save the linearized profile and you're
good to go!

The basic principle is that you want to bring in the next darkest ink when
the "previous" lighter one has reached its peak.  The tradoffs are that
if you bring a darker ink sooner, you may see dots (mostly with older
printers) - or that if you wait until your lighter ink hits its max density 
 before you bring in the darker ink, you may be unnecessarily flooding your 
paper.

A shortcut to forever reading-in densities is to compare your printed
target to an ideal 26 step printed target with approximately similar 
dmax to the paper your are trying to profile.  
It's easy to spot the trouble spots by eye this way and adjust.
When you get where all patches separate well, but the overall
densities don't quite match up, you can read it in and let the linearization
take care of the rest. When you get good at it you may only need to read
the whole scale once or twice at the most to get a very nice profile.
Be sure to calibrate the 810 with its own reference plaque.




> 
> 2.  Is there a way of pre-determining what the density of a mix of two 
> inks will produce?  More generally how do the densities of two inks 
> mix.   Does 50% light grey and 50% dark grey lighter or darker than 
> 50% dark grey.  I suppose the answer is "it depend" but what does it 
> depend on?

Theoretically if a "light" gray is - let's say - half the density of a "darker"
gray, at 50% it will only be the equivalent of 25% density compared to the
darker ink.
So, just doing the math....   50% of such a light gray added to 50%
of the darker gray will make 25+50=75%  That means that their
combination will result in a density equivalent to 75% density of the 
darker ink.

However, none of that is useful in practice - just a way to conceptualize 
the process. What is really hard to predict is the actual density readings
of any combination of inks - and that's where some trial and error
comes into play. There is just too many variables (dot gain for one) to
make a generalized prediction.



> 
> 3. There seems to be more than one set of curves that will produce 
> good results (sometimes the light grey is used through almost the 
> entire range and sometimes not).  Are there standard approcaches to 
> designing curves?


See my first answer: it's about hiding dots without overloading the media.
As an exercise, do a whole curve with black only. Read it and
plot it in the linearization tab.
Then pull the black back to 50% and bring in the lightest gray. Make
a complete functional profile  by, again, ploting and linearizing etc.
With just 2 inks you can play with different starting positions for  the 
black and you will quickly see the effects.
When you find you get dots in the midtones, for example,
 bring in the 3rd gray etc.

There is one big deal in all this and it has to do with shadows:
When the black starts to rise to the 100% position (patch 26)
it wants the minimum amount of other ink to be printing along with
it. If you let the other grays come in at 100% they will not let the
black go down at full strength and you will loose dmax. It's
important  to hit that point just right: you want some gray so you don't
run the risk of seeing white dots if the paper  shows through 
the dithered black at the transition point -
but you want to make sure to keep it to a minimum as you get to 100%.
That appies to all pigment inks I know of and is why when you examine
a profile that comes with IJC, you will see those bell curves for all inks
except black.





> 
> 4. Roy seems to use some voodoo or other black magic to create the QTR 
> curves mathematically from the relative densities of the inks.  Are 
> there set approaches to this or is Roy just extra clever :)

You have to ask Roy that...   but I think he is extra clever anyway <BG>.




> 
> 5. In IJC the curve peak can be increased or decreased which I assume 
> is setting the amount of ink to a higher or lower level than the ink 
> limits that are set earlier in the process.  Is thre a general rule 
> why one would adjust the ink limit in preference to adjusting the 
> shape of the curve?


Ideally you do the first pass on a new ink/paper/printer to determine
how much ink you can lay down and still see a difference from patch
to patch. Most of the times, you'll find that all inks go out to patch 26
except black which hits its max around 14-17.  The limits only make
it so you don't  make a curve expecting that the more ink that's
laid down the higher the density will be on  paper. 

If you make a curve for the black ink for example, set its limits to 26
(= no limit) and pull it all the way up to its max, 
the last bunch of steps in the curve will be meanigless because
they will all produce the same density - since after step 14 (or whatever)
the paper doesn't get any darker. By putting the ink limit to 14 prior
to making the curve, it makes it so the highest point on the curve
never gets past where more ink = more density on paper.
 In effect it moves to patch 26 (the 100% mark) whatever 
density the black ink produced at patch 14 when no limits were set .


However, there is another use for the ink limits:
 Once you make your curves, it sometimes seems that an entire 
ink-curve needs to put out less ink.  Rather than pull down the curve
itself and potentially alter its shape in a way that will introduce errors, 
it's a lot easier to pull down the limits for that ink.
 In fact, when you gain some experience, you may find that, 
even if a gray ink can go to 26 and produce visible density
differerence patch-to-patch, you may choose to set the limit to lower 
values; perhaps 15 or 20. It will give you a little headroom down the line
if you need more ink after you finalize the shape of your curve.





> 
> 6. Is there any recommended reading material? 

I'm not aware of one, other than the pdf manual that ships with IJC.


Hope this helps -


Antonis