[argyllcms] Re: Print Validation
- From: <robert@xxxxxxxxxxxxxxxxxx>
- To: <argyllcms@xxxxxxxxxxxxx>
- Date: Sat, 18 Oct 2014 21:47:01 +0100
Hi Jan-Peter,
Thanks for your advice ... and definitely converting to AdobeRGB was the
wrong thing to do.
I modified the commands to these:
=====================================================================
targen -v -d2 -G -f100 iPFTest
copy iPFTest.ti1 iPFRef.ti1
printtarg -v -r -ii1 -a1.0 -T300 -M6 -pA4 iPFTest
cctiff -v -ir -e iPF6400_HP_ID_Satin.icm iPFTest.tif iPFTestO.tif
move /Y iPFTestO.tif iPFTest.tif
fakeread -v -Ir iPF6400_HP_ID_Satin.icm iPFRef
Pause Print iPFTest.tif with no colour management
chartread iPFTest
Pause The test results will be in iPFValidate.txt
colverify -v2 -N -k -s -w -x iPFRef.ti3 iPFTest.ti3 >iPFValidate.txt
======================================================================
and I'm now getting good results:
Worst 10% errors (CIEDE2000) peak = 1.414888, avg
= 1.371666
Best 90% errors (CIEDE2000) peak = 1.101138, avg
= 0.708107
avg err X 0.006700, Y 0.006789, Z
0.005514
avg err L* 0.810990, a* 0.468245, b*
0.481245
I didn't leave much time for the ink to dry, so perhaps the results could be
even better.
I've set the rendering intent to be relative colorimetric in both cctiff and
fakeread ... is this correct if I want to do a RelCol check? I'm really not
clear how cctiff in particular should be used for this test.
At any rate this seems to be a reasonable way to check how well the print
environment currently is (that is, the profile/calibration/paper). It could
be run, for example, after changing paper roll to see if recalibration or
reprofiling was needed, or periodically to check for drift.
One thing I'm wondering is about the normalisation. I set the colverify
flag to -N which means normalise to XYZ, so I'm assuming that this corrects
for paper white, and is the correct setting for this test. Is that correct?
Robert
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx]
On Behalf Of Jan-Peter Homann
Sent: 18 October 2014 14:23
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: Print Validation
Hello Robert,
The proofing workflow Graeme is describing adresses normally a source
gamut which is smaller than the target gamut and an absolut colorimetric
match from source to target.
You are using AdobeRGB as a source, which has several colors, which are
out of gamut of the target colorspace (printer)
you also using the relative colorimetric intent.
...
If you want to test, how good matches a printer the printer profile:
create a a print wedge in the printer colorspace
(same colorspace you have the printer profile with)
create a text file with the same colors as the print wedge
convert the text file with fakeread to L*a*b*
print the print wedge in device mode mode (colormanagement off)
measure the print wedge
compare the measurement results with the fakeread output
Regards
Jan-Peter
Am 17.10.14 18:39, schrieb robert@xxxxxxxxxxxxxxxxxx:
> Hi Graeme,
>
> I've tried the workflow you've suggested (below) to check my print
accuracy
> and my results are well off. Here are the first few (worst values):
>
> No of test patches = 100
> File 0 Chose patch 1 as white, xyz 0.964203 1.000000 0.824905
> File 1 Chose patch 1 as white, xyz 0.837679 0.872949 0.741557
> L*a*b* white reference = XYZ 0.964200 1.000000 0.824900
> 21: 18.038560 25.951478 22.489887 <=> 39.803954 11.831945 -35.079664 de
> 37.269236
> 28: 34.185591 -67.085881 -6.455341 <=> 62.944388 -90.026713 44.587042 de
> 34.884374
> 41: 40.524868 -33.803806 -52.438105 <=> 65.223756 -60.678084 -21.994298
de
> 29.777970
> 26: 52.908776 -78.403017 -17.947748 <=> 71.957265 -79.305174 41.137194 de
> 29.656735
> 12: 57.830887 -42.597937 -57.500006 <=> 76.080635 -62.957608 -8.243975 de
> 27.776304
>
> So I'm clearly doing something drastically wrong!
>
> I would really appreciate it if you would have a look at these commands to
> see where the mistakes:
>
>
> targen -v -d2 -G -f100 iPFTest
> copy iPFTest.ti1 iPFRef.ti1
>
> fakeread -v -Ir iPF6400_HP_ID_Satin.icm iPFRef
>
> printtarg -v -r -ii1 -a1.0 -T300 -M6 -pA4 iPFTest
> cctiff -v -e AdobeRGB1998.icc iPFTest.tif iPFTestO.tif
> move /Y iPFTestO.tif iPFTest.tif
>
>
> Pause Print iPFRef.tif using iPF6400_HP_ID_Satin.icm profile and Rel. Col.
> chartread iPFTest
>
> Pause The test results will be in iPFValidate.txt
> colverify -v2 -N -k -s -w -x iPFRef.ti3 iPFTest.ti3 >iPFValidate.txt
>
> I'm printing iPFTest.tif using Photoshop, RelCol to
iPF6400_HP_ID_Satin.icm.
>
>
> Many thanks!
>
> Robert
>
>
> -----Original Message-----
> From: argyllcms-bounce@xxxxxxxxxxxxx
[mailto:argyllcms-bounce@xxxxxxxxxxxxx]
> On Behalf Of Graeme Gill
> Sent: 15 October 2014 14:26
> To: argyllcms@xxxxxxxxxxxxx
> Subject: [argyllcms] Re: Print Validation
>
> robert@xxxxxxxxxxxxxxxxxx wrote:
>
>> I've tried your first method and it's just fine. I'll try the variation:
> is
>> that effectively equivalent to using profcheck?
>
> Yes, it's going to be similar.
>
>> I don't really understand your last suggestion.
>
> In a proofing type workflow, the aim is to get a printer (or other
> display device) to emulate some particular target colorspace. You do
> this by transforming the colors you want to reproduce through a source
> profile (which defines the target you are trying to emulate) and the
> device profile.
>
> If you want to check how accurate your proofing system is,
> then the type of thing you might do is generate a test
> set in the target device colorspace, and run it through the
> source profile with fakeread to create the reference .ti3 file.
>
> Then you run the same colors through your proofing workflow
> (e.g. create a .tif using printtarg, then apply the source
> to printer device link to the .tiff using cctiff and print it,
> or run the .tif through whatever ICC based workflow you have
> setup for proofing), measure the result and compare
> the .ti3 to the reference using colverify.
>
> Graeme Gill.
>
>
>
>
>
--
homann colormanagement ----------- fon +49 30 611 075 18
Jan-Peter Homann -------------- mobile +49 171 54 70 358
Herzbergstr. 55 FG 3.01 -- http://www.colormanagement.de
10365 Berlin ---------- mailto:homann@xxxxxxxxxxxxxxxxxx
Other related posts:
