QTR-Quadtone RIP

Richard,


I'm not yet scanning my 21-step targets with anything, but my eyes. Sorry if I didn't make this clear in the first place.
But I do have a scanner and could make PS readings.
I find a 16-bit limitation in the Lab eyedropper settings (first or second color readout) when 8-bit makes 0-100 Lab values (that don't exactly match grayscale inverted values).
So I'm confused about your 32-bit recommandation.
I use CS4.

Some precisions:

I'm carbon printing only with multiple layers in registration, doing mostly duotone images.
The first layer is very often a color pigmented layer, so not only am I not used to density measurements and linearization, but I wouldn't know what ideal values aiming for that color layer.
These days, I'm using a light orange/braun pigment for the first layer.
Let's take the 50% stepwedge patch: what orange/braun density is seen by humans as a 50% gray equivalent?
And how should I measure color targets and get grayscale values? Scan in RGB and convert in PS?

Still taking the 50% stepwedge patch exemple, it will actually be on the final print the addition of the first orange/braun layer and the following darker layer, not necessarily a black pigment layer, eventually a dark red/braun one.
The lightest second layer values may start about the 20% final print values. But 30% is also possible if I make a darker first layer! This is my decision, a subjective aesthetic choice…
What references do I have but ones I have to build myself? That means trial and error, defining my own ideal values.

Scanning a final print wouldn't tell me about each of the three negatives required. I should measure on each layer, before they are unified on paper.
Each layer is translucent pigmented gelatin on a transparent PET support.
Again, how can I avoid building my own ideal values when scanning transparent media, not a print on paper?

Right now, fine tuning the 0-5% values of my first carbon layer is my #1 concern.
QTR curves seem the best option so far, since one can write curve values between 95 and 100%, including decimal places, which can't be done in PS.
I will need more time to learn and find out if I can use your tools somehow, replacing the ideal values with mine, if you find that key back!

Many thanks.


Stefan