Digital BW, The Print

Hi Keith,

> I have now ordered a new printer (a 3800) 

Nice.  I think it and similar machines are the future.

>... dedicating my 2100 (2200 in thw US) to B&W. 


>(yes I know the 3800 can do quite good B&W, but why compromise).

With a rip, I'm not sure how much of a compromise it would be.  The basic
factors for good B&W include, first, to use the most carbon and least color.
This helps with respect to both metamerism and fading.  With a rip you can
do that with the 3800.  Second, there need to be enough levels (dilutions)
of carbon ink to make a smooth gradient.  As the printers have gotten
better, the number of dilutions needed has decreased.  I think in a modern
printer 3 Ks are enough and a good compromise.  I, for example, dropped the
super light grays because they fade faster and print warmer, requiring more
faster-fading color inks to offset their warmth.   

>I can either use QTR, or Photoshop curves etc., 

I'm now using a non-blended inkset approach in my 2200 that largely follows
the Epson K3 model.  See http://www.paulroark.com/BW-Info/4K+.pdf  

While I also use the UT-3D inkset
(http://www.paulroark.com/BW-Info/UT-3D_Readme.pdf) in another printer, I
must admit that the non-blended approach with QTR is easier to profile than
the 3D with Photoshop curves (and QTR's Create ICC).  

I think Roy and I have pretty much settled on a k3 type approach for the
Golden Trout photo workshop this summer, in part because it can be used on
virtually any hextone or better printer and will provide a rather
un-interrupted learning curve or growth path to the large format printers.
That is, not only is the workflow the same, but the actual curves one makes
for a 220 with a k3 analog are easily transferred to larger printers.  Even
the bulk inks one buys will continue to be useable as new printers are
purchased.

So, for those who don't want anyone to see color dots under a
high-resolution scan, the UT-3D is what I'd favor.  However, for those who
don't care about that (and we no longer have to hide that fact that we use
inkjet printers), the k3-type approach (what I have also called 4K+ or 5K+)
provides the most consistent and continuous growth path, not to mention
easier profiling and far greater stability on large format printers --
including the 3800.

Since you have a 3800, you'd be able to use the same basic workflow for both
to make B&W with the minimum color.  

The one thing I would add on the 2100, however, would be the PK in a color
spot.  This ink between the LK and MK does allow one to use a little less
color ink in the shadows (denser carbon inks print more neutral) and it also
avoids the need to change black inks -- which is built into the 3800, of
course.

The practical disadvantages of the k3 analogs relative to the 3D approach
with the Epson driver include that QTR (and IJC) are slow, in Windows QTR
does not print from PS, QTR does not recognize alpha channels (IJC does on
Windows), and QTR and IJC are not "color managed" such that a standard Gray
Gamma 2.2 will match the print (I use a standard layer to correct for this
in printing, but always now edit my files in GG 2.2 space).

> but I do want to go for refillable carts to keep the costs sensible.

I prefer the MIS carbon inks.  They are warmer and not greenish like the
Epson LK and LLK.  They are also have lower bronzing.  The MIS base appears
to have a light UV blocker in it.  As such, I get better lightfastness than
with the Epson inks.  The exception here is that the Epson magenta pigment
appears to have the edge in my tests.  One major advantage to a k3 analog
and non-blended inks is that one can mix inksets.  So, I use MIS inks for
the most part, but Epson LM.

It'll be interesting to hear how the 3800 compares to whatever approach you
use in the 2100.  If I had to live with only one printer, it'd be the 3800,
and I'd then farm out my larger prints to a service bureau.  I think that is
the future.

Paul
www.PaulRoark.com