Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Steve Kale <stevekale@=
b...> 
wrote:
> Hi Roy
> 
> I am hoping to get back to ³making curves² this weekend for my 2100 (and
> from there to the soft proofing profiles).  With regard to ink limits, wh=
ere
> do you suggest I start?  Below are the ink limits that Carl used in his c=
ool
> and warm EEM descriptor files, respectively:
> 
> N_OF_INKS=7
> DEFAULT_INK_LIMIT=75
> 
> LIMIT_K=
> BOOST_K=90
> LIMIT_C=0
> LIMIT_M=0
> LIMIT_Y=0
> LIMIT_LC=27
> LIMIT_LM=27
> LIMIT_LK=56
> 
> N_OF_INKS=7
> DEFAULT_INK_LIMIT=75
> 
> LIMIT_K=
> BOOST_K=90
> LIMIT_C=0
> LIMIT_M=0
> LIMIT_Y=0
> LIMIT_LC=0
> LIMIT_LM=0
> LIMIT_LK=60
> 
> As I noted earlier, if I adopt these ink limits I get quite different
> partitioning values than those used by Carl.  Maybe this is something I a=
m
> doing wrong but I intend to start from scratch again in the morning.  I h=
ope
> to get curves done for EEM and HPR ­ adjusting for the effects of sprayin=
g
> with Lyson Print Guard ­ over the weekend.

I'm curious: Have you used Carl's curve?  Are you looking for something
particularly different?

> 
> >The other plus is that UC_NEUTRALIZER feeds the exact same curve shape i=
nto
> >LK LC and LM so no matter how you change curve shapes and ink limits
> >the 3 inks are always used in the same proportion which avoids color
> >crossover problem.
> 
> So effectively in the above cool scenario LK is limited to 27 anyway?  So=
rry
> I am confused by this and your next sentence:
Maybe its a typo, but LK=56 and LC=LM=27

> 
> >So the ink limit values i.e. ratios for LK,LC,LM determine
> >the color tone of the output.
> 
> Do you mean they are used increasingly together but always in the proport=
ion
> of 27/110, 27/110, 56/110?  So this ink ³mix² (an equal dose of LC and LM=

> applied to LK ­ a virtual mix) is that which is deemed to take the LK fro=
m
> warm to cold?  Ie if I could print a 100% patch of this combo it would lo=
ok
> cold? 

Yes, its basically like you took the 3 inks and mixed them into one bottle =
--
56 parts LK, 27 parts LC and 27 parts LM. 

> 
> And this combo of limits is deemed to give good coverage for a 100% patch=
 in
> the same way that 75 was deemed enough for K?   I am surprised that both
> 27/27/56 and 0/0/60 are deemed to each give good coverage. Wouldn¹t one h=
ave
> a lot more ink laid down than the other?

I think you're right, and if you look at the linearize densities you can se=
e that
the midtones of the cool curve are denser than the warm curve.  But the
linearization correction curve evens it all out in the end, so its not a pr=
oblem.

>   
> If this is the case then how is K cooled, if at all (I suspect it is not)=
?

There's no specific cooling of K but the LK,LC,LM inks do run out all the
way to 100% so there is some cool ink.  I think the main effect is that
once you get pretty dense the tone converges to the black ink no matter
what you do.  Plus, since you choose the blending of a warm and cool
curve at the end, you'll be choosing that based on all the ink together.

> I ask because, although when I look at the Ink Pattern Page printed using=

> calibration mode it seems that K is quite neutral, at least vs LK.  Howev=
er,
> I have been puzzled by the comments in the BO discussion which note (quit=
e
> clearly from the examples given) that BO prints with Epson UC K  are very=

> warm vs their BO Eboni counterparts.  Am I right in understanding that wi=
th
> BO printing ONLY K is used (ie no LK is used) or is the warmth of BO UC d=
ue
> in part to LK?  

I think BO printing shows the K ink color a lot more than when you have all=

the inks together.  This is especially true in the lighter areas with fewer=
 dots.

Roy

> 
> Thanks again for your help and generosity
> 
> Steve
> 
> From: "Roy Harrington" <roy@h...>
> Reply-To: DigitalBlackandWhiteThePrint@yahoogroups.com
> Date: Wed, 14 Jan 2004 20:08:23 -0000
> To: DigitalBlackandWhiteThePrint@yahoogroups.com
> Subject: Re: [Digital BW] QTR and Making Curves
> 
> 
> Hi Steve,
> 
> You're right that you ought to come up with the ink limits before decidin=
g
> on
> the partitioning values.  I'm not positive exactly how Carl did his curve=
s
> --
> it's somewhat difficult since ideally you should be partitioning the
> LK,LC,LM
> combination rather that just the LK.   Fortunately, it's not a big proble=
m
> because the final linearization smoothes out the step wedge very well.  T=
he
> other plus is that UC_NEUTRALIZER feeds the exact same curve shape into
> LK LC and LM so no matter how you change curve shapes and ink limits
> the 3 inks are always used in the same proportion which avoids color
> crossover problem.  So the ink limit values i.e. ratios for LK,LC,LM
> determine 
> the color tone of the output.
> 
> The linearization or "final smoothing" is really not that complicated in
> concept.  Smoothness is actually built into the partitioning and mixing o=
f
> the inks -- there are only smooth changes to all the inkjets.  The main
> necessity is that it's always increasing as you go from light to dark.
> Once that is true, it's a matter of a correction curve that puts the
> densities
> in the "right place".   You can think of it as first plotting densities
> versus
> input step values -- a bunch of (x,y) or (input,output) points.  The
> correction
> curve is just the mirror image: x and y swapped.  If you go into Photosho=
p
> and design a custom dot gain curve you are doing exactly the same thing.
> There are some hairy looking formulas for converting between density,
> luminosity and other units.  If you want more reading, Adobe has some
> info buried deep in their website.   Here's a website that has some
> pretty technical info, but your eyes may start to glaze over.
> http://www.brucelindbloom.com/
> 
> Roy
> 
> > The 2200 ink limits were done a little more ad hoc rather than strictly=
 by
> > my original documents.  The idea for cool 2200 prints is to have a
> > combination of LK,LC,LM play the role of one light black ink. (Basicall=
y,
> > instead of mixing the inks manually to get a neutral gray, the idea is
> > to mix them using software curves).  However the difference is that you=

> > now have 3 inkjets delivering ink so it made sense to lower the individ=
ual
> > limits to make sure we don't flood the paper.
> 
> 
> 
> 
> [Non-text portions of this message have been removed]