Digital BW, The Print

Barry,

I talked to Bob Ziess of MIS yesterday, and he told me that the c and 
m inks in FS-E were 1/3 dilutions of the normal FS inks. I don't know 
how this dilution translates to density, because it is not a linear 
function. In the normal FS inks, C=c (dark), and M=m (medium). So the 
FS-E inks tone order is CcMmY (inks position) = 
dark/medium1/medium2/light1/light2. I don't what the density 
relationship will be between medium1 and medium2, or light1 and 
light2. No matter, an RGB curve will need to be used to control the 
tones to bring each of the 5% print patches in-line with the expected 
density. Bob has not posted a curve yet, but he told me that he has 
received one from John Wolfe, and will be posting it soon. I am not 
sure which printer model it is for. However, I expect the curve 
generated on one hextone printer should be fairly close to universal 
for all 6 color printers, provided you maintain the same image space 
and other printer settings. Minor tweaks should be all that is 
necessary. Are you using an adjustment curve?

From my own personal experience, a proper match (but not necessarily 
exact match) between perceived gray density of the OEM color and true 
gray (FS) inks is all one needs for a seamless ramp, and little issue 
with posterization.  My judgement tells me that the most important 
thing is that the progression from dark to light be maintained in the 
correct order. The perceived gray density (lightness value?) of OEM 
color inks from dark to light is approximately CMcmY.  I use the FS 
inks, but with modified positions, in the following manner: CcMmY 
(ink position) = CMCMY (FS inks) = dark/medium/dark/medium. Then I 
created a global RGB curve to get proper separation of 5% patches. 
Because of the tone order I use, the curve does not have steep 
gradients. Steep gradients in the curve actually can lead to 
posterization (or more accurately solarization). This work-flow on my 
1280 gives me the best b/w prints I have ever made. The RGB curve can 
be found at the inkjetgoodies.com web-site (I have no interest in the 
company). This work-flow will, however, require you to use a CIS or 
fill your own cartridges. One thing I have found is that it helps to 
have a little toe in the dark end of the spectrum, i.e. the 
relationship is not linear at the dark end. I prefer my 5% RGB to be 
in the 2-3%, and the 10% be around 8%. This gives an overall 
impression of darker blacks in low-key images. For high-key images, 
it is better to have a straight linear relationship.

I have actually found that for the 1280 class of printer, you 
actually need only 3 tones (black, dark, and medium) to get very good 
prints. The folks at lincolninks.com have been advocating this 
approach for quite a while.

Good luck.

Shilesh


--- In DigitalBlackandWhiteThePrint@y..., "Barry Kelsall" 
<bktimes@y...> wrote:
> ----- Original Message -----
> From: "Paul Roark" <paul.roark@v...>
> 
> > With the hextone machines, the Epson driver does not
> > work at all because the FS and Piezo cyan & photo (light) cyan
> > are the same density -- likewise for the magenta
> > & photo magenta.  MIS now has what they call the FS-"E"
> > series that is geared to the Epson driver, but I'm
> > not sure anyone has developed workflows for the inks.
> 
> I am using the FS-N "E" in my 1200 and can say it works great for
> high-contrast, busy images.
> 
> But on portraits I see mild posterization.
> 
> I am no expert, but I am thinking that the density of the 
replacement ink
> must match the Epson OEM color ink for this to work with the native 
Epson
> driver. For example, say the hex-black cyan is a tad less dense 
than the
> Epson cyan, and the hex-black photo-cyan is a tad more dense than 
its OEM
> counterpart. Then the transition in the Epson driver between these 
tones
> will not work, producing a flat spot where the inks mix that no 
amount of
> RGB curve tweaking can cure. And only one of the inks needs be off 
to
> produce this problem (this is my fuzzy logic here, though I state 
it like an
> accepted pardigm).
> 
> Also disturbing are the reports I have seen of users seeing 
posterization
> with the Piezography driver. Even a dedicated partitioning solution 
has
> problems?
> 
> -BK
> who is wondering what to try next