Digital BW, The Print

Paul Roark wrote:

>
>
>
>The question as to what is "consistent and true to proper B&W printing" is
>more complex.  There perhaps is no such thing.  However, here are some of my
>thoughts on the subject.
>
>Recently and for newer printers, I've moved to inksets that can be
>controlled via the Epson driver sliders as well as with Photoshop curves and
>rips like QTR and IJC.  The driver sliders don't give a lot of control, but
>if the B&W inkset is made for them, the print quality with minimal slider
>controls is often, on average, the best there is.  
>
>As one moves away from the Epson driver and all the jets firing, the chances
>of problems start to increase.  In particular, many printers appear to need
>all the jets or "channels" firing to hide what we'd consider printer defects
>that cause microbanding.  The more one pays for a printer, often the more
>likely it'll be built well enough to use radical curves (or Black Only
>printing) that cut out inks.  However, even with 2200s if curves cut out
>some "colors" or jets, a certain percentage of them will show some
>microbanding or other artifacts.  So, for those who get a printer that is
>marginal and can't handle the curves or other options we often talk about,
>being able to print with the Epson driver at near default settings is a good
>fallback that may well save their investment. 
>
>Also, being able to just print with the Epson driver at near default
>settings and get a good B&W print is much needed positive reinforcement for
>many.  The frequent posters on this forum are technically savvy and much
>more tolerant of the problems that might be encountered as one moves away
>from the Epson driver than is the typical B&W photographer who simply wants
>to output his or her digital files.  Part of my "mission" is to get top
>notch output to these more typical B&W photographers (including those with
>limited budgets).
> 
>At any rate, Epson doesn't make its systems for people like most of the
>posters here who use non-Epson inks and/or specialized workflows.  So, we're
>often asking our printers to do things they were never designed to do.
>
>So, what is "consistent and true to proper B&W printing"?  There is probably
>no such thing, but Epson would say buy a 2400.  In this printer they've
>integrated "B&W" printing into the driver, where they also have all the ink
>jets firing -- probably for the same reason I use that position. 
>
>On the other hand, a C86 will make more lightfast B&W for a fraction the
>cost of a 2400.  And with the R200, I hope to have not only more lightfast
>but also smoother B&W output at the $69 printer level.  
>
>And you can be sure that the approach I use for the 2400 I now have will
>have a mode that takes advantage of the Epson driver and has all the jets
>firing.  On top of this will be curves and rip controls that get more out of
>the system, but the starting point is the "rip" that was an integral part of
>the printer design -- the Epson driver at near default settings.
>
>I'm not sure the 2200 is being made any more.  If there are good ones
>available for cheap, it may still be a good way to go, but if one has to pay
>a price close to the 2400, I'd recommend stepping up to the 2400.
>
>Hope this rambling post helps.
>
>Paul
>www.PaulRoark.com 
>
>  
>
Paul,

I had written a related rambling for another thread but kept it in 
draft.  Now that I see that your view on this is close to my thoughts I 
like to add it here.

In general: When I think of the comments in the B&W list about what 
flaws people find in their setup I often wonder whether we keep enough 
"buffer" in the concepts. Given the complaints about banding, the 
visibility of dots, the linearity of tones, the detail in shadows one 
also wonders whether there's a crowd here that's 10x more quality 
conscious than in color printing or we are doing things wrong. Let's 
assume both is happening. We almost forgot that the printers are 
originally CMY(K) mixing color printers. The task of quad printers 
should be much easier in essence: black squirted in a nice way. We could 
split that black in 3 dilutions but  we already use 4,6,7 and the same 
number of heads. I'm aware that the K head in a CMYK printer doesn't 
have to do the full range from 0-100% grey steps but usually it kicks in 
at 50-60% depending on the black generation setting and does that part 
up to 100%, usually with some CMY composite grey assistance. That's 40% 
of the range. In a K7 setup one head does roughly 14% of the range but 
it has a sole responsibility for that small part of the range. If it 
changes in time, if it was a bit sub standard from the start, it still 
has to do that part of the ramp on its own. At the same time it isn't 
really equiped for the task as we didn't change the weaving-dithering 
patterns it has to lay down, it lays down 1/3-1/4 of a color printing 
pattern that will never get the other parts. In color printing that 
happens as well but not from 0-100% everywhere and we are less sensitive 
to hue changes in a color print than to tone changes. Much can be 
compensated by nozzle quantity, higher printer resolutions (= more 
passes which in a way resembles more nozzles but not entirely), 
linearisation, wider overlaps in partioning. Less problems occur with 
pro models as their droplet size and overall construction is more 
directed to consistency. The last means that linearisation also keeps 
longer in time. For the desktop models with their lower nozzle numbers, 
very fine but less consistent droplet size, overall more "plastic" 
construction including the small cart feeding I think it may be wise to 
add even more buffer in the quad setups. Part of that buffer was I think 
available in Paul's early PS curves as it relied more on mixing than the 
typical RIP partioning. It lacks the linearisation but with prebaked 
curves that aren't adjusted to the user's printer there is no place for 
that either. (QTR's profile creation could be used though to get that 
right for the PS curves + the Epson driver). A step from the curves to 
RIP partioning by a desktop model user could be a disappointment when 
his printer shows its flaws faster with what looks like a superior 
software setup. I do not keep a record of complaints but my gut feeling 
is that there's a tendency to stretch the capacity of the hard- and 
software while it doesn't give us nice B&W prints all the time. Using 
two quad ranges, longer overlaps in partioning, suitable monotone 
dithering/weaving, less toners,  this all could create some buffer.

Ernst