QTR-Quadtone RIP

ferdinand_paris wrote:
> I'm still having terrible trouble reading the test strip printout in
> strip mode and would appreciate any tips on how to edit it so as to
> make it easier to read.
> 
> Do people have any views about whether the i1 is more accurate in test
> or patch mode?  You can make the i1 fairly fast in patch mode by
> putting it on an auto-delay.  I'd have thought that when the i1 is
> stationary it may get a better read, provided you position it
> correctly.  (However I'd hate to do colour profiling this way!)
> 
> F_P


I'm still using a SpectroCam for measuring and I do not have 
any trouble when I use the random 21x4 or 51 patches 
targets. For profiling I use two random 51 patches now and 
make one profile with a spreadsheet step in between to get 
the patch numbers and measured results in line.

With the other targets it is often the lines in between that 
may not be observed by the meter. Either you have to set 
other spectrometer settings that it switches correctly or go 
slower or faster. For the Z3100 that I have I made the 
targets 300 PPI without resampling and print them 1:1 
through Qimage with extrapolation and sharpening off. They 
are larger that way but the lines are not too wide I have 
experienced. Anyway I use the random ones mainly.

More off-topic:

As I'm using this on a Z3100 the method is slightly 
different. I first do an internal calibration of the Z3100, 
it measures about 48 grey patches in total then of the 4 K 
inks (and more of the hues). That's more or less the 
linearisation. Then I have to select a HP Z3100 ABW driver 
mode that delivers a near linear output before I can make 
the profile. Right now I can see on the profiling target 
whether I had the driver mode correct for near linear output 
On a 51 patches target reading there always is a step or two 
not precisely in line. With the internal calibration of the 
Z3100 the base is kept consistent and the profile gives me 
the softproof and the perceptual adaption of the image file 
to the paper.

I write more or less the linearisation as it calibrates the 
4 K ink tone ranges separately and not the resulting tone 
range of the four or three partitioned K inks together like 
QTR linearisation does. I think it works however, no bumps 
or dips visible where the 3 or 2 partition transfer points 
must be. My best guess is that HP shifts the partitioning to 
the calibration numbers otherwise it would show. I faced the 
same issue when I made a quadtone driver on my Wasatch 
SoftRip before QTR could drive an Epson 9000 quad.

When I use another mode with the Z3100 driver, feed the ABW 
driver R=G=B monochrome/color files (AdobeRGB assigned) with 
driver CM on (+AdobeRGB space) but greyscale printing (all 
according HP's manual for B&W) then I get what I think 
almost the same output as with the QTR profile. So HP did it 
well for normal B&W use. I still have to test that on the 
numbers but it doesn't give me the softproof or a choice of 
rendering in the conversion so I probably wouldn't use it 
other than for a case where I have no profile yet. Lots of 
sliders in the ABW mode to do it another way but that is 
different and the preview is on a constructed image not of 
the one you print.


-- 
Met vriendelijke groeten,  Ernst


|  Dinkla Grafische Techniek  |
|     www.pigment-print.com    |
|             ( unvollendet )            |