Digital BW, The Print

Tyler wrote:

>...
>...this idea is one of
>the central issues in Allen's approach with the Spectratones. Reverse
>engineering what the Epson driver is doing with respect to what
>densities are being delivered by which colors by the driver, then
>having mono ink densities in the right order to best advantage it
>seems to work well for them. Less so for 4 color printers, where more
>severe manipulations are required to eliminate dots.

The idea definitely has appeal for the newest 6-ink systems.  However, even
on the 7000, critical people appear to be able to see dots with the
light/photo density inks in the highlights.

If one is using the stock MIS densities, part of the problem would be that
the MIS "75" is very dark.  (The MIS percentage labels do not relate to the
percentage of black either with respect to visual density relative to black
or mixing ratios.  They are arbitrary and correlate with density order --
and not much else.)  This too-dark "75" not only darkens the image, but it
also makes the light/"photo" position inks, which should be about 60% of the
full strength visual density, too dark.  The appearance of dots could be
lessened if these inks were lightened.  (Note that Grady has to lighten his
image for printing.)

Additionally, the MIS light inks are probably unnecessarily light for the
modern 6-ink printers.  (The 15% is like water.)  The MIS "25%" is just a
little lighter than Piezo yellow (about 24% v. 27%).  These, no doubt, have
value for a 3000, but for any modern printer, the Piezo magenta (about 38%
relative to the Piezo black) is all the lighter you need for visually
dotless highlights.  The 1280 with the MIS VM inks use this density of ink
as the lightest gray, and at 2880 they are smoother than Piezo in my 1160.

Although the 6-ink printers do not all appear to be uniform, the 1280 and
1270 apparently use the same inks -- and they are good enough that the dots
should be irrelevant. (I still do partitioning in my variable-tone curves,
however.)  So, a system like this might be best aimed at the 1280.

On the other hand, Epson will probably release a next-generation printer
that uses different ink densities, so all the ink mixing would have to be
re-done.  The software-controlled systems may have a more difficult
workflow, but at least software/curves are quicker and easier to modify.

As one who likes to try different ink mixes, I still prefer the 4-inkers
just because they are much easier to experiment with -- fewer inks to mix.

Paul
http://www.PaulRoark.com

Paul wrote:
 
> The MIS percentage labels do not relate to the
> percentage of black either with respect to visual density relative to
black
> or mixing ratios.  They are arbitrary and correlate with density order --
> and not much else.
 
Paul,
do you know what are the actual densities of these inks (relative to the new
MIS black, that I suppose is used in the Full spectrum set)?
 
Thanks, Alex


[Non-text portions of this message have been removed]