Digital BW, The Print

> From: Paul Roark <paul.roark@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Fri, 3 Dec 2004 14:19:29 -0800
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Subject: RE: [Digital BW] Tonal range and linearization
> 
> 
> Steve,
> 
> OK, then that is consistent with the average comment being that the prints are
> a bit lighter than the monitor image.
> 

Yes.  Once you understand the relationships between what is in the file, on
screen and on the print (and actually how they relate to their darkroom
counterparts if you started out that way) then this becomes useful.  Take a
look at the other steps as well.  When you ask your question do you simply
show a big 50% step on screen and print (no other steps)?  Remember the
other steps will be off to a greater extent.  If someone is looking at the
whole step wedge - even if you ask them about the 50% step only - they may
also be swayed by the other steps due to "brightness induction".  I recently
stumbled across this website which has a couple of interesting tests:

http://www.nebulus.org/tutorials/2d/photoshop/color/color3.html

> Most of the curves require that a grayscale file be converted to Adobe RGB
> first.  Then the RGB curves map to the CMY inks, black being generated by a
> combination of the curves.  So, the RGB curves provide input-output
> relationships that allow me to specify the density of the image at each point.
> It's just using Photoshop curves and the driver controls or built-in
> characteristics (including cross-overs) to do most of what RIPs do.

Yep.  I was wondering how mid grey (50% or 128) was modified.  I haven't
seen any of your curves. How 50% maps to your print result is a function of
(a) your curves and (b) the transfer function buried in the Epson driver and
even more complicated by the fact that the Epson driver is expecting totally
different inks.  So not transparent like a RIP but you do get amazing
results.  I love The Narrows photo on your site.

> 
> My Kodak standard gray, which I believe is 18% reflectance, reads as Lab L=50
> with my X-Rite.  That is more like 0.73 density.
> 
> So, bottom line, what I'm hearing is that 0.66 might be closer to the mark as
> far as a standard midpoint density for the final print.

If you have a gamma of 2.2 for the monitor yes.  It will, I believe, be
different for each gamma.

> 
> I generally like the idea of having a standard midpoint as opposed to cutting
> the total range in half.

Not sure what you mean by cutting the tonal range in half. I would never
propose doing so either. But if I had a system whereby the in-gamut tonal
range was reproduced faithfully then I could on screen look at my image in
that tonal range space and compress (or clip) in the manner I choose. In
this environment, the only transfer function at work is my one PS "print"
curve.

> I think the standard midpoint makes it easier to
> print the same image on different papers without having to re-do the file.

You would never need to redo the file - just apply the "print curve" that
fits the image into the printer/paper/ink tonal space.  Then take a last
look at the image on screen knowing that because it is all in-gamut it will
be faithfully reproduced.  New ink/paper combo?  No problem.  For file
preparation (as opposed to RIP preparation) all you need to know is the
tonal range of the combo - paper white and dMax - and dial them into the
"print curve" -  by simply changing the two end points of an existing print
curve (although you might like to tweak how the curve approaches the "knees"
of the tonal response of the ink/paper combo).

So the workflow becomes: 1. Get the image looking nice on screen using the
full gamut available as usual - save as file copy.  2. Apply a print curve
which specifically takes into account the tonal range of the ink/paper combo
you intend to print to (these could simply be stored somewhere) - you can
physically see the end points in the tonal range, the midpoint and all in
between. Now the image on screen is all in-gamut and the tonal relationships
are exactly how they will print. 3.  Make any last minute, image specific
adjustments - these would be simple tweaks to the print curve. 4. Print.
> 
> Interesting stuff, but it's too easy to get hung up in the numbers and ignore
> the aesthetics of the image.
> 

I agree.  But it would be easier if we got what we were looking for when we
push a button or shifted a curve.  That is, full logical transparency.

Cheers

Steve