Digital BW, The Print

Joost,

<paul.roark@...> wrote:

>> ... approaches to QTR profiling ...
>>... what gives the best results with the least work. 
>>.... my latest thoughts on procedure, for what it's worth. 
>> See http://home1.gte.net/res09aij/QTR_2200_4K+_MBW-Carb-w7.pdf 


>... Looking forward to see where this route will lead to.

Making the toning easier and reducing the need for full Photoshop and a
spectro are on the list of things to do.  I think (hope) B&W printing with
full tone control can be cheap and easy.

>... what would be your expectation on the complexity to 
> adapt the Cyan and Magenta toner curves for each and every paper?

Once the toner curve shape is appropriate, it's very easy to change overall
toning via the ink limit numbers -- move both the same direction
proportionately to change Lab b, make offsetting proportional moves of the
limits to change Lab a.

The problem is the toner "curve" shape. (I'm inclined to use the table
instead of an external curve so that all the information about the toner
"curve" is easily accessible by users.)  I hope I can get the number of
points on the table low enough and placed appropriately that we can just
increase or decrease a few numbers to change the shape of the curve. This
effort will probably take more experience with more papers. 

>... I'm working with the UT3D set and find it actually quite
> simple to create a new curve, including getting the tonality right.

Good.  I actually think the non-blended inksets are even easier than what I
was able to do with the 3D approach.

>... do you expect that, once the basic approach has 
> crystallized, this approach is suitable for ordinary human beings ;-)

That's where I'm trying to go.

>... why don't you convert the file prior to 
> printing to the Gray-LAB space (=QTR space) instead 
> of correcting the Gamma 2.2 profile with a PS curve?

I don't think PS Elements (or other non-full Photoshop) users can convert to
Gray-Lab space.  Gray Gamma 2.2 is the standard default for almost
everything, and I think we are better off staying with industry standards.
Today I'll post a method for Elements users to use the GG 2.2 to QTR curve
view a layer.  Most non-Adobe image editors have some sort of curves such
that they can copy the points I've published and match the monitor to print.
I don't want the cost of Photoshop to be a barrier.  (I'm a former antitrust
law enforcer.  So fighting monopolies is in my blood.)

Frankly, I think QTR should have an option to linearize to GG 2.2.  (And I
have a potential "hacker" method to do that.)  In fact, I wish QTR had an
option to be "color managed" the way Create ICC is.

> I tend to edit leave the image in 
>RGB (with a color channel mixer layer as a non-destructive RGB2gray 
>conversion). For printing purposes I flatten the file and convert to 
>Gray-LAB and save as separate print file. Using the gray-lab profile 
>is what I understood to be the preferred workflow, but you choose 
>differently apparently. I'm curious to learn the background.

We all tend to develop our own approaches to B&W.  However, from a novice's
perspective, look at the complexity this might involve.  Until I started
using GG 2.2 as my standard working space, I used a custom dot gain curve.
But it occurred to me then that it was a step that should not have to be
there.  The machines are built to industry wide color management standards,
and I think we're best off when we utilize those standards as opposed to
fighting them.  But, there's no "best" approach for everyone.

Paul
www.PaulRoark.com