Paul and Roy,
Thanks for this.
I created a 1 pt type file (whether it is accurate, I'm not sure).
Had trouble getting it right in CS3, so I did it in a drawing program
and converted it to a TIFF (but at only 360 dpi) in CS3.
I printed it on a 4000 with QTR and K7 inks, and on a 4800 with Epson
K3 inks with photo black (using color setting not ABW) -- both at 2880
dpi. Looking at the print with a 10x loupe, both were totally legible.
This was done on a EEMatte type paper. The K7 printed type might have
been a bit smoother, but it also seemed a bit lighter. Interestingly,
on the K3 printed type no color inks seemed to be used (unlike in the
promotional example). I then tried printing it as gray, rather than
black text with the 4800/K3, as that might make more use of the color
inks. It still looked monochromatic -- at least under the loupe.
Needless to say, it doesn't look like the Epson print sample in the
promotional example. I'll try printing it in ABW on glossy paper (ABW
does not appear to be available for use on matte papers, at least when
photo black is loaded) tomorrow.
I'll also scan the print samples in tomorrow -- it's a 3200 dpi Epson
scanner, so it should produce a comparable scan to your own for
comparison.
--- In QuadtoneRIP@yahoogroups.com, "Roy Harrington" <roy@...> wrote:
>
> This is an interesting experiment. I regularly use a 3800 with K3
inks and QTR.
> So I tried a few prints. These are Harman paper with photo black.
>
> Here are three scans -- scanned at 3200, sharped like Paul's and
> reduced to 1600dpi.
>
> 1) My regular setup: 1440super, bi-directional, mixture of curves.
> 2) 2880dpi, uni-directional, same curves, these have a little too
much ink so:
> 3) 2880dpi, uni-directional, same curves, but the image density was
reduced.
>
> See: http://www.harrington.com/restest1.jpg
>
> It looks to me that the biggest issue is using 2880dpi for printing.
>
> Roy
>
> On Sat, Aug 30, 2008 at 3:31 PM, Paul Roark <paul.roark@...> wrote:
> >
> >>If this now famous 1 pt image file would be made available
> >>we could make our own opinion.
> >
> > I've used resolution test files for some time that I simply make in
> > Photoshop. I believe the top resolution of the Epson driver is
720 ppi, so
> > that is what the file is. It can be downloaded at
> > http://www.paulroark.com/BW-Info/Resolution-chart-720ppi.jpg
> > (There does not appear to be any difference between this Jpeg and
the Tiff
> > original.) The file is 1" x 0.3".
> >
> > I've added a "one point" text sample to this. There are 72 points
per inch
> > in the current U.S. method, according to Wikipedia
> > (http://en.wikipedia.org/wiki/Point_(typography)) So one point
text is 10
> > pixels high in a 720 ppi file. This includes the top of the
tallest capital
> > letter and the lowest tail, in addition to a 1 pixel top margin.
This is
> > really small. I used the PS text function to make 10 point text
in a file
> > that was 7200 ppi and then resized to 720 ppi. I made one of the text
> > sections with no anti-aliasing, and one with "crisp"
anti-aliasing. Given
> > the few pixels involved, neither is particularly good even on the
monitor.
> >
> > Evaluating the printed images is tough if one does not have a very
good
> > scanner. Seeing the details is a bit beyond the limits of my 1600
dpi Epson
> > flatbed, but after sharpening 125% PS, one can start to evaluate the
> > relative quality of the printing.
> > See http://www.paulroark.com/BW-Info/Resolution-comparison.jpg
> >
> > I've included a screen grab of the Piezo test in the above image.
> >
> > A drum scanner is really needed for this task. On the other hand,
with a
> > good loupe, one can see more than my scanner, and the text is actually
> > readable -- with some effort.
> >
> > The bottom line I saw was that with the 2200 and 1400, with
Carbon-7 and
> > Eboni-6, respectively, and using the Epson driver and QTR, the one
point
> > type is reasonably clear with all of the above combinations.
There is no
> > significant difference between the Epson driver and QTR. The 1400
is a bit
> > sharper than the 2200, but in normal viewing, one would not see any
> > difference. The Piezo printing from the test on the web appears
to be in
> > the same league with what I see with the Epson driver or QTR and
the inksets
> > I use.
> >
> > I do not have a K3 ABW setup to test. It would be interesting to
see a good
Show quoted textHide quoted text
> > scan of a print of the above test chart with K3 ABW.
> >
> > Paul
> > www.PaulRoark.com