[argyllcms] Re: Question regarding White Point and spotread

  • From: <robert@xxxxxxxxxxxxxxxxxx>
  • To: <argyllcms@xxxxxxxxxxxxx>
  • Date: Tue, 4 Nov 2014 10:47:46 -0000

Actually, Graham, my original question (going back to October 20th), which
hasn't been answered, was based on these commands here:

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 using no color management.
chartread iPFTest
Pause The test results will be in iPFValidate.txt
colverify -v2 -N -k -s -w -x -L iPF6400_HP_ID_Satin.icm iPFRef.ti3
iPFTest.ti3 >iPFValidate.txt

This seems to work OK, but chartread (like spotread, I assume, although your
documentation doesn't state that either use absolute for reflective
measurements) does absolute readings (probably obvious to you, but not to
me).  I assume, again, that the -N in colverify compensates for this ... but
this is just a wild guess on my part. So, is this a correct way to validate
that the print system is working correctly (profile, printer etc?).  I have
cctiff with -r because I'm not interested in validating an absolute
colorimetric print since I never print using Abscol.

Following on from that, my question was:
"The reason I'm interested in going a bit further is that I would like to be
able to generate a target from an image (for example a single-color
gradient) in order to check out how well the profile is performing". I
received no suggestions to that question (probably because there's no way of
doing it using Argyll?), so I then asked a question regarding manually
constructing a .ti1 file (you have XYZ values in there that seem to do
nothing as far as I can see, but it took me ages to figure that out). As I
was getting nowhere with this I tried spotread and got errors of around dE94
of 1.6 which seemed too high.  I asked whether or not this had to do with a
white point mismatch and Brad has done his best to help me, but neither of
us really understands what's going on.

Well, from what you say it DOES have to do with a white point mismatch since
the difference between RelCol and AbsCol is effectively the white point (you
will probably correct me ... but in doing so you will no doubt confuse me
even more).

What you don't understand, because you are so deep into all of this and know
so much about it, is that having a list of commands with a bunch of
parameters that do a whole lot of different things is fine for someone who
really understands everything about color management, profiles ... AND
ArgylCMS. For someone like me (even though I am a software engineer) half of
it is gobbledygook.  To get things to do what I want I have to do a lot of
head-scratching, trial-and-error ... and then keep my fingers crossed that
what I'm doing is right. (In fact, if it was not for DispCalGUI I would
NEVER have even looked at Argyll).

Of course I do have the choice of not using Argyll - after all I have
purchased i1Profiler Publish ... but this seems a pity considering the
months of investment I've already put into Argyll and the fact that it is
undoubtedly a very flexible software suite, and that it can do some things
that i1Profiler probably can't do (or at least not directly).

For example, you say:
"Convert the absolute target color to the device values using absolute
colorimetric intent."

How? In Photoshop? Using Argyll? What command? What flags?

And then:

"Or if you have a target color that is relative,
convert to device space using relative intent, then
convert that back to L*a*b* using absolute intent,
and use that as your spotread verification target."

How? What commands? What flags?

A lot of the problems we have could be answered using a full set of usage
scenarios with a clear explanation of what the commands are doing; or a
fairly simple wizard-based system could guide us through the steps.  A
proper GUI would be wonderful, but that's a major development of course.

Robert




-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx]
On Behalf Of Graeme Gill
Sent: 04 November 2014 02:50
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: Question regarding White Point and spotread

robert@xxxxxxxxxxxxxxxxxx wrote:

> Actually, I think the way that AbsCol is described (that is that the white
> point is not shifted) is very confusing to common mortals like me.

I'm not sure where you got that description from. Absolute means
that the measurement is with reference to the perfect diffuse reflector.

Relative means that it is with reference to the actual paper color.

> I still think that it is probably necessary to compensate the spotread
value
> for the paper white.

I think you have got it the wrong way around. Convert the absolute
target color to the device values using absolute colorimetric
intent. Or if you have a target color that is relative,
convert to device space using relative intent, then
convert that back to L*a*b* using absolute intent,
and use that as your spotread verification target.

(And please edit your posts. No-one wants to see the same
previous emails over and over.)

Graeme Gill.



Other related posts: