Re: MIS quadtone inks on Epson paper

[Dan Culbertson <danculb@concentric.net>]

Regarding the question below on the sharp detail of the Piezogrphy BW
driver:
There are three issues here.  The first is that a 1200 has a much smaller
dot than a 3000 so the image detail has a better chance to come through the
standard Epson driver.  Detail with the 1200 Epson driver is to my eye very
close to detail  with the Piezography driver on a 3000.  Second, the 1200
places more ink in the highlights since it has three built-in light ink
channels (light c, light m, and yellow).  More light ink in the highlights
means the dots in the highlights don't have to be spread too far from the
printer's maximum true dot density of 720 dots per inch to get any specified
gray tone.  Since the dots aren't separated as far they can retain more of
the printers theoretical maximum true detail (and I won't speculate what
*that* maximum is - but it is relative to how close you can put the dots
together).  The Piezography driver mimics this virtuously by partitioning
the light ink channel into the highlights and the dark ink channels into the
darker zones so the Piezography driver retains more of the theoretical max
detail in the highlights.   You can do exactly the same thing on the 3000
with a RIP and a manual CMYK channel separation but that is a very complex
procedure.  Finally the third point - the Piezography driver is supposed to
actually *increase* the theoretical max dot density and thus the theoretical
max detail.  I don't know much about the mechanical methods used by the
driver so I can't say if this claim is accurate.  However I do know that the
Piezography driver only makes a single pass as opposed to the Epson drivers
four pass printing so there is a possibility that there is a small loss in
the Epson driver due to mechanical registration inaccuracies between passes.
So there is some possibility that on any one printer type the Piezography
driver would have less problem with retaining image detail (and more of a
problem with banding) as a result.

Again, in my first post, I was comparing a six color small dot printer (the
1200) to a coarse dot larger printer (the 3000) and I find that the smaller
dots and light ink channels of the 1200 offset much of the gain one would
have using a 3000 with special CMYK/RIP techniques or with the Piezography
driver (which essentially incorporates a special CMYK technique to achieve
some of its greater detail).   I *predict* that at least some of this will
hold true when using the 7000 and the 9000 six color printers with
multi-tone gray inks and the standard Epson driver.  At the very least what
should hold true is that the Epson driver with a printer using the light ink
channels in any six channel Epson printers for light gray inks will give as
smooth a highlight tone as a four channel Epson printer with a special
driver or CMYK technique.  So it may be true that a 7000, using the standard
Epson RGB driver, will give as good a grayscale print as a 3000 with the
Piezography driver.  But then the 3000 including the Piezography driver
costs a heck of a lot less than a 7000 - which is why I have three 3000s and
no 7000 :-)

Dan Culbertson

> 
> Hi, I supposed that one of the most important features of Piezo was it's
> ability for good image detail (shape detail I mean). And this was
> achieved by using the specialized driver. But from your comment below it
> sounds like a pure ink based system (plus native driver) can challenge
> the whole Piezo approach.
> 
> Did I misunderstood something?
> 
> Rodrigo
> 
> 
> "F. Neil Simms" wrote:
>> 
>> date. But
>> it's beginning to sound like John N.'s won't work too well on 4-color
>> printers (compared to 6-color ones), so I'm in a state of dither... I
>> may ditch the 1160 and buy a 1200/CIS to use with John N.'s inks if that
>> combination is shown to produce prints of quality comparable to
>> Piezography, and iff they make them available in bulk.

-
Turn off HTML mail features. Keep quoted material short. Use accurate
subject lines. http://www.leben.com/lists for list instructions.