[Openicc] Drop size calibration

[Robert Krawitz]

   From: "Hal V. Engel" <hvengel at astound.net>
   Date: Mon, 28 Jan 2008 15:12:41 -0800

   On Monday 28 January 2008 14:31:45 Gerhard Fuernkranz wrote:
   > Hal V. Engel wrote:
   > > I suspect that drop size calibration has a lot in common with calibrating
   > > the light/dark ink transitions in CcMmYKk type printers.
   >
   > Btw: This statement, which brings drop sizes and light/dark
   > transitions into relation, reminds me on another approach I have
   > seen somewhere in the literature. This approach does the
   > dark/light splitting in the dithering engine: It treats the
   > e.g. 16 dot-on-dot overprint combinations of the three dot sizes
   > of "C" and the three dot sizes of "c" as if they were 15 drop
   > sizes of a single virtual cyan channel. Then these 15 virtual
   > drop sizes are sorted by density, and dithering is done between
   > the resulting 16 virtual levels. A requirement for this approach
   > is of course that the hues of the dark and light inks must match
   > sufficiently well, i.e. the dark inks should be basically the
   > same colorant as the light ones, but at a higher concentration. I
   > have no idea though, how well this approach works in practice...
   >
   > > It is posted here:
   > > http://lprof.sourceforge.net/images/Channel-Error.jpg

   I could recreate these without too much trouble.  I improved the
   spreadsheet since I created that graphic and it now graphs the K
   channel.  I could simply print out a new target with the GutenPrint
   default settings and re-measure it and drop those measurements into
   the spread sheet.  I could post the spreadsheet where people could
   have a closer look since this would have the graphs and the
   measurement data.  In addition others could use the spread sheet to
   plot other things, for example, how the hue of a channel changes
   with density.  I could have this available tommow (to allow for
   drying before measurement) if anyone is interested.  It only takes
   about 5 minutes to measure a target and about one minute to get the
   measurements into the spreadsheet once the target is ready to
   measure.

I'd be very interested in this.

If you want to isolate issues of drop size vs. light ink, you can do
this:

1) Print normally at very high resolution (5760x1440 or higher), where
   only the smallest drop size is used.  2880x2880 or 5760x2880 may be
   even better due to the form factor, but on most printers those
   resolutions have problems near the bottom of the page.

2) Print at lower resolution using "4 color standard" (i. e. CMYK)
   ink.

You can also print using CMYK at high resolution if you want a
baseline.

I'm very interested in how Ordered New dither performs against
Ordered.  Ordered dither is just 2 levels, so it fills up one drop
size before starting to use the next.  Ordered New uses 3 levels, and
it starts a new drop size when the previous one is 50% in use.

-- 
Robert Krawitz                                     <rlk at alum.mit.edu>

Tall Clubs International  --  http://www.tall.org/ or 1-888-IM-TALL-2
Member of the League for Programming Freedom -- mail lpf at uunet.uu.net
Project lead for Gutenprint   --    http://gimp-print.sourceforge.net

"Linux doesn't dictate how I work, I dictate how Linux works."
--Eric Crampton