[argyllcms] Re: Use of DTP-22 digital swatchbook with Argyll CMS

  • From: "Mario Becroft (mailing list)" <list@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Mon, 05 Feb 2007 10:00:32 +1300

Hi Mike,

A little while ago I wrote on the Argyll CMS mailing list with questions
about using the DTP-22 and you mentioned that you had one, but you had
not yet been able to achieve acceptable results with it when profiling
your inkjet and dye-sub printers.

I have now got a DTP-22 and I have just tried it out for the first time
this weekend. I have not used with Argyll CMS yet (although I intend to
try that out sooner or later) but I have been using it with the Onyx RIP
I use on my HP large format inkjet printers.

So far I have calibrated my Designjet 2500CP machine using a 31 point
per colour linearisation pattern and a 500 point ICC profile pattern,
and although I have encountered a few problems, which may go away once I
become more familiar with the profiling method, the results are
initially extremely promising. The output is far better than anything I
have ever seen on my Designjet printers, and that was just my first try
with only a 500-point ICC profiling pattern (the Onyx media profiling
software recommends using 1000 or 2000 points).

I had a bit of trouble setting the ink limits correctly and I think the
profile may be inaccurate in the very dark colours, but I suspect that
this might improve once I redo the ICC profile with more points and
corrected ink limit settings. I also had some trouble with the colour of
the paper, which for the media I have been testing with so far, came out
as noticably cyan, even though to the eye the media looks white.
Therefore it was compensating by making the output more red, which
looked wrong; first instance neutral greys looked slightly red, even
though the DTP-22 read them as exactly grey. But by changing the
parameters for the ICC profile generation I was able to prevent it from
applying this compensation, and the result then looks correct to the eye
even though it reads wrong. I am not sure exactly why this is, but
presumaby the eye compensates for the cyan hugh of the paper and makes
it look white--either that or my DTP-22 is reading wrong. But I tried
reading some other paper with it, and while most of the papers I tried
seemed to have a similar cyan bias according to the DTP-22, there were
several which were exactly white--so I guess the papers really do vary
and you just can't see it.

So at this stage, this is just to say that it definitely is possible to
make useful profiles with the DTP-22, but clearly it requires some
knowledge of how best to use the tool. Once I try it out with Argyll
(which I want to use for profiling my Epson printer which does not use
the Onyx RIP) I will let you know how I get on.

In the Onyx media profiling software, there is a test tool which lets
you take readings with the DTP-22 and display them either as CMYK
densitometric or LAB colorimetric values on-screen. This is quite useful
for checking that the device basically works and reads colours
correctly. I imagine the Argyll system (or the colorport software you
are using) no doubt has a similar test tool.

I presume you have the calibration target for your DTP-22 and that you
have calibrated it? Using Onyx, after calibration it then reads the
calibration target density as 0.04 in all the channels. Before
calibration, my DTP-22 was completely wrong.

I think that to test your DTP-22 and find out whether it is working
properly, the best way would be to get an IT8 target, which should come
with a sheet showing the actual colorimetric values for each point. You
can then measure these with your DTP-22 and check whether the readings
are as expected--this would clarify whether there could be a fault in
your DTP-22 or whether it is just the profiling process that is going
wrong. I intend to try this out myself just for curiousity's sake.



Other related posts: