Digital BW, The Print

Paul: 

I wouldn't expect the density (= log(1/reflectance) to be linear or 
that the 50% grayscale patch to have a density midway between approx. 
0 and 1.6 using the MIS inksets -- it should perceptually be midway 
between the two extreams.   The often used 18% gray card has a 
reflected density of about 0.74 (log(1/0.18)).  However the two 
questions I haven't found real answers to are: 18% is the middle of 
what range and is perceived under what conditions?  If 18% 
reflectance is the perceptual middle of the 0 to approx. 2.0 range 
shown in the Kodak Q-13 card, then the middle of the range for 
quad/hex inkjet prints with a range of 0 to approx. 1.6 range should 
be a little lighter (nearer to your approx. 0.6). Or is 18% the 
middle of a broader range like 0 to 2.5 or......

Other than the middle of what range, the other variable, is the 
viewing conditions, at least the intensity of the viewing lightsource 
and maybe even the spectral composition.  The *uncontrolled* 
experiments that I have conducted of viewing the Q-13 wedge (and 
wedges I've printed using the MIS FS and FSE inksets) under indirect 
sunlight, full sun light, Ott-light, and incandesant light all 
perceptually give me slighly different mid points.  Also varying the 
background (light or dark) also changes my perception.

The short of it is that I am now convinced that there is no single 
*correct* answer to what density should the 50% grayscale patch be. 
The eye is just too adaptive.  There is some discussion of the 18% 
gray card in the Accurate Image Manipulation website, www.aim-dtp.net.

BTW, now that I have a Swatchbook (thanks for convincing me) and can 
accurately measure densities, I find that the 1160 curves I developed 
yield a 50% grayscale = approx. 0.6.

Jeff Randall 

--- In DigitalBlackandWhiteThePrint@y..., "Paul Roark" 
<paul.roark@v...> wrote:
> Cleavis,
> 
> >[Is] there a reliable relationship between the
> >info box (Photoshop) and the output tone once the curve is applied?
> 
> >If you have a K only sampler, will that indicate a percentage tone
> >laid down by the ink?
> 
> I hope so.
> 
> The simple answer is that when the eyedropper in Photoshop reads a 
patch as
> 0% -- pure white -- the printer is supposed to put zero ink there, 
and if
> the eyedropper reads a patch as "100%" -- "pure black" -- the 
printer should
> be laying down as much black ink as possible.
> 
> However, there is a question as to how one measures the print 
densities and
> what a "linear" output should be.  That is, between those obvious 
end
> points, what should the "gamma" or slope of the curve be?  How 
dense should
> the 50% patch should be?
> 
> If you print the grayscale ramp with many printers, the 50% patch 
might be
> lighter or darker than the 50% patch that results from one of the 
vm curves
> or from the Piezo driver.  And, what the "50%" patch on a test 
strip will
> read on a desktop scanner is a function of how the scanner gamma is 
set.
> 
> I called Epson to see if there was any way to get my scanner to 
be "linear."
> There isn't.  They don't claim that any particular setting will 
give any
> particular slope to the curve.  It's all just relative.
> 
> The densitometers are a more reliable way to see what the densities 
are.
> The average 50% patch on the old PiezoBW test strips I've made and 
seen is
> about 0.61 - 0.62, and 0.61 is what I use for the vm curves.  This 
is
> clearly not half way between the paper white -- about 0.04 -- and 
the VM-K
> 100% patch density of about 1.65 (depending on printer and paper 
type
> setting).
> 
> So, the only simple answer relates to the end points.  After that, 
one has
> an open question as to what the "50%" patch should read.  The three 
test
> strips that are in my fader had 50% patch readings as follows: 
2200 - 0.57,
> MIS Archival color - 0.60, and PiezoTone-S - 0.69.
> 
> Paul
> http://www.PaulRoark.com