Dan Margulis Applied Color Theory
Image Correcting for a Desktop Printer
Date: Wed, 30 Nov 2005 11:43:14 -0800
From: Dennis Dunbar
Subject: Re: Re: Image correcting for a desktop printer
On 11/30/05, Mike Russell wrote:
The unwanted color is an unwieldy concept in RGB.
In part this is because
for most objects the unwanted color resides in two
channels, not one. So
RGB practitioners often seek detail by curving the
dominant color of an object.
In RGB saturated colors tend to blow out individual
channels, so it is easy
to associate detail and tonality with color gamut, when
in fact they are a
phenomenon of luminance.
This is why my preferred method of color correcting in
RGB is to use 2 separate curve layers, one set to "Color"
blending and the other to "Luminance". Not only does it give more
control, but it makes color correcting in general much easier.
Dennis Dunbar ;-)>
___________________________________________________________________________
Date: Thu, 1 Dec 2005 09:17:09 -0800 (PST)
From: John Opitz
Subject: Re: Image correcting for a desktop printer
So I'm up for learning to correct in RGB, but am
curious as to
whether the same principles are applicable and more or
less as easy
to apply? For example, Dan talks extensively about the
role of the
unwanted color. So if I were working with fleshtones
and wanted to
use the cyan channel in it's role as the unwanted color
to add depth
and form to the image, how would you do that in RGB?
Like Curly would say: Sointenly…It’s monkey
work! Use the green or *blue channel. Do this before using curves (color
correction/contrast enhancing). Using the apply image command. Use the
green or *blue channel and apply it to the composite (RGB master) channel
in darken mode. After giving it the oakey doakey! Do a fade to luminosity
and set your opacity. Or if you want to get more romantic about it. Playing
footsie. Apply the above method on a luminosity layer. If things are going
right for you, rounding second just don’t stop. Go for home. You can
do a second apply image, and use the Green channel and apply image it, just
to the red channel (activate the red channel in the channels pallet) in
normal mode. While the red channel is active (your sharpening the red
channel only) apply sharpening using conventional or high radius, low
amount method, then set your opacity on the luminosity layer. And
don’t be afraid of pre-sharpening. The luminosity layer is set to
100%. You’re going to adjust it after, to 40% to 60%. Making an
action, script of it is eaiser, faster than doing it by hand if your main
thing is skin tones. When things were looking like its an easy score, you
find, things might not have been as planned. You find that the skin tones
are in front of a 21c, 90m, 90y, 21k background. Ohhh, Nyuk-nyuk! Add a
layer mask. Reveal all. Use the apply image command; apply the red channel
to the layer mask in hard light mode. You can be more selective using
curves, forcing the mask towards b & w in selective areas.
*Be careful of channel blending with jpegs. Color as
well as luminosity artifacting may be trouble with jpegs. After checking
the channels, will show the green to be better. Smoothing the channels
(color and luminosity) helps a great deal.
John Opitz
___________________________________________________________________________
Date: Thu, 1 Dec 2005 12:39:32 -0600
From: RJay Hansen
Subject: Re: Image correcting for a desktop printer
John, I'd like add a couple of points to what has been
said on this matter.
There are many CM advocates that say you should never
convert to cmyk unless and until the final step of a cmyk workflow. They
will cite the irretrievable "data loss" or as Terry Wyse more
correctly pointed out "color loss". Terry also pointed out that,
especially w/an sRGB profile, many photos are very close to the cmyk gamut
even when still in rgb mode.
I work in prepress so my experience w/color started
with cmyk and I'm still much more comfortable working in cmyk than rgb,
although I've gotten much better with rgb over the past two or three years.
Still, I routinely convert photos from my (consumer level) digital camera
to cmyk to make corrections/adjustments if I think I can accomplish them
better than in rgb and then convert back to rgb for printing on my inkjet
or posting to my website. With a bit of experience you can begin to easily
distinguish colors that likely are out of the cmyk gamut and judge whether
you can live with the diminished saturation and possible hue shift they
exhibit after converting to cmyk. Of course, you can always go ahead and
make the conversion before deciding whether the color loss is acceptable or
not and undo it if it isn't. In my experience, with many photos you are
hard pressed to see any difference in color after making an rgb to cmyk
conversion.
For a quick visual example, you can look at Mike
Russells's curvemeister challenge and compare a couple of images done by
list members. If you go to the results of the challenge here:
http:
//www.curvemeister.com/Challenge/Challenge04/index.htm
then view picture number 12, done by Marco Ugolini and
picture number 21 done by RJay Hansen (yours truly), you can compare one
done completely in rgb by Marco and one done w/a combination of rgb and
cmyk done by myself.
Setting aside the difference in the way we individually
chose to "interpret" the photo (I went for a higher contrast
appearance than Marco did), and just focusing on color, you don't see a
significant difference in saturation or hue between the two photos.
I'm not saying there's anything at all wrong about
confining your corrections and adjustments to rgb. I do that myself in some
cases (it's photo dependant), but at the same time, I don't think you
should feel you must rule out a conversion from rgb to cmyk and then back
to rgb for printing if the photo and situation call for it.
RJay
--
http://www.rocketjam.com
___________________________________________________________________________
Date: Thu, 01 Dec 2005 13:29:50 -0800
From: Marco Ugolini
Subject: Re: Image correcting for a desktop printer
In a message dated Thursday, December 1, 2005 10:39 AM,
Rjay Hansen wrote:
There are many CM advocates that say you should never
convert to cmyk
unless and until the final step of a cmyk workflow.
That's a workflow option called
"late-binding", as opposed to "early-binding", in which
the conversion is done at the beginning or at an early stage in the
production phase.
They will cite the irretrievable "data loss"
or as Terry Wyse more correctly
pointed out "color loss".
In other words, a larger gamut sRGB is squeezed into an
overall smaller one (say US Web Coated SWOP v2). Once done, you cannot go
back. So the loss *is* irretrievable (though we can quibble whether it's
"data" or "color", since color lives in the file as
data too, after all).
Terry also pointed out that, especially w/an sRGB
profile, many photos are
very close to the cmyk gamut even when still in rgb
mode.
Yes, but that all depends on a careful use of profiles
and conversion techniques, that is, on the "color meaning"
assigned to the sRGB values within the workflow.
Though markedly smaller than AdobeRGB, sRGB is still
quite larger than US Web Coated SWOP v2 or US Sheetfed Coated v2 -- except
for an area in the Cyan-Greens where both these CMYK spaces exceed sRGB by
a considerable margin. (You can easily verify all this by graphing these
profiles in 3D Lab mode in Chromix ColorThink).
I work in prepress so my experience w/color started
with cmyk and I'm still
much more comfortable working in cmyk than rgb,
although I've gotten much
better with rgb over the past two or three years.
Still, I routinely convert
photos from my (consumer level) digital camera to cmyk
to make
corrections/adjustments if I think I can accomplish
them better than in rgb
and then convert back to rgb for printing on my inkjet
or posting to my
website.
If that works out for you, fine. In other instances,
gamut-chopping may not be the best choice if done just for the sake of the
operator's comfort level and at the expense of some achievable gamut in
possible output scenarios.
With a bit of experience you can begin to easily
distinguish colors that
likely are out of the cmyk gamut
Unfortunately, there is more than just one "CMYK
gamut". In today's world, one never knows whether a given job is going
to end up printed on a newspaper or a magazine, traditional halftone or
stochastic, CMYK or Hexachrome, glossy or matte, high-quality or
low-quality paper, in the US or Europe or Asia, etc.
These possible scenarios all differ substantially, and
one should not overestimate one's ability to keep them all straight in
their head. Speaking for myself, I certainly cannot do that. Caution is the
catchword.
and judge whether you can live with the diminished
saturation and possible hue
shift they exhibit after converting to cmyk.
The question I usually ask myself is not whether *I*
can live with it, but whether *the client* can. I err on the side of
caution and struggle to preserve the widest possible range of choices that
the client will then be able to choose from.
Of course, you can always go ahead and make the
conversion before deciding
whether the color loss is acceptable or not and undo it
if it isn't. In my
experience, with many photos you are hard pressed to
see any difference in
color after making an rgb to cmyk conversion.
If you start from sRGB, convert to a CMYK space, then
go back to sRGB, I can see that this would be the case. Also, one must
remember that there are many colors (even ones printable within a US Web
Coated SWOP v2 scenario) that even a well-calibrated and -profiled monitor
is unable to display adequately or at all. That's another reason why I
mentioned caution earlier.
For a quick visual example, you can look at Mike
Russells's curvemeister
challenge and compare a couple of images done by list
members. If you go to
the results of the challenge here:
http:
//www.curvemeister.com/Challenge/Challenge04/index.htm
then view picture number 12, done by Marco Ugolini and
picture number
21 done by RJay Hansen (yours truly), you can compare
one done
completely in rgb by Marco and one done w/a combination
of rgb and cmyk
done by myself.
I will leave it to others to express an opinion on the
merit, and also ask anyone interested in this to read the text accompanying
my submission. I would like to point out that depending on one's
browser to judge the color results of this Curvemeister challenge is
questionable at best. It's a pity that there seems to be no way (none that
I can find) to download the images submitted to this "contest",
so that one is able to view them more accurately within Photoshop.
Setting aside the difference in the way we individually
chose to
"interpret" the photo (I went for a higher
contrast appearance than
Marco did), and just focusing on color, you don't see a
significant
difference in saturation or hue between the two photos.
Again, I wish I could download the image to be able to
view it in Photoshop. Until I do that, I reserve my judgment.
I'm not saying there's anything at all wrong about
confining your
corrections and adjustments to rgb. I do that myself in
some cases
(it's photo dependant), but at the same time, I don't
think you should
feel you must rule out a conversion from rgb to cmyk
and then back to
rgb for printing if the photo and situation call for
it.
If that works for you, it's all good and fine. But
there are situations when that may not be enough.
Best regards.
--------------
Marco Ugolini
Mill Valley, CA
___________________________________________________________________________
Date: Thu, 1 Dec 2005 14:52:34 -0800
From: Mike Russell
Subject: Re: Image correcting for a desktop printer
From: "Marco Ugolini"
I would like to point out that depending on one's
browser to judge the color
results of this Curvemeister challenge is questionable
at best. It's a pity
that there seems to be no way (none that I can find) to
download the images
submitted to this "contest", so that one is
able to view them more
accurately within Photoshop.
I do convert challenge images, if necessary, to sRGB,
using Relative Colorimetric intent before putting them on the web page.
This is the first time I've been made aware of this as
a limitation, and I will consider making the original submitted image
available as well.
Mike Russell
www.curvemeister.com
___________________________________________________________________________
Date: Thu, 1 Dec 2005 14:58:28 -0800
From: Stuart Larson
Subject: Re: Image correcting for a desktop printer
I noticed that there is a link to the corrected image
by Mike DeSantis that says "Download Mike's CMYK tiff file", but
that was the only one I noticed that had such a link.
Stuart Larson
___________________________________________________________________________
Date: Thu, 01 Dec 2005 15:42:48 -0800
From: Marco Ugolini
Subject: Re: Re: Image correcting for a desktop printer
In my message of November 28, I pointed out how RGB and
CMY are the inverse of one another, and how R, G, and B can add or subtract
Cyan, Magenta or Yellow from an image, due to that inverse relation.
What I failed to mention is another side of this same
relationship:
As I wrote:
C is the
inverse of R
M is the
inverse of G
Y is the
inverse of B
But it's also true that:
C corresponds
to G + B
M corresponds
to R + B
Y corresponds
to R + G
Which is the inverse of:
R corresponds
to M + Y
G corresponds
to C + Y
B corresponds
to C + M
So, any one color in each of the two sets is equal to
the sum of the two colors in the other set that are not its inverse. At
least that is the theory (which in practice is far messier and more
complex).
So, to be complete in my explanation, I should have
said that if, for example, you are in RGB and wish to add Cyan to the
image, you can either *decrease Red* or *increase Green and Blue together*
(by roughly equal amounts, plus an adjustment in the composite RGB curve to
compensate for the increase or decrease in luminosity).
I was reminded of my omission when I read John Opitz's
remark (in his message of December 1) in which he correctly mentions that
you can use the Green and Blue channels as one possible way to adjust the
Cyan levels while working in RGB mode.
I hope this sounds clear enough.
Regards.
--------------
Marco Ugolini
Mill Valley, CA
___________________________________________________________________________
Date: Thu, 01 Dec 2005 16:44:55 -0800
From: Marco Ugolini
Subject: Re: TR001
In a message dated Thu, Dec 1, 2005 4:11 AM, Peter
Leyland wrote:
There seem to be two schools of thought with one side
leaning towards
standards, repeatability and other worthly aims and the
other prefering what
they see. Reality vs expectation.
I would respectfully submit that, if by saying
"one side leaning towards standards, repeatability and other worthy
aims" one wishes to characterize, among others, those who adopt
color-managed techniques, and by "the other preferring what they
see" one refers instead to those who work "by the numbers",
then this is a false dichotomy.
We work together, not against each other.
Color management techniques, when done with common
sense and respect of the results, are far from denying reality. They are
about helping manage it.
Though some individuals *may* prove to be dogmatic
proponents of these solutions, most people I know use CM as another tool (a
proven and very helpful one) to achieve final results that match
expectations.
No conflict with "reality" there, as far as I
know, nor between "reality" and "expectation".
Regards.
--------------
Marco Ugolini
Mill Valley, CA
___________________________________________________________________________
Date: Thu, 1 Dec 2005 16:26:19 -0800
From: Paul D. DeRocco
Subject: RE: Re: Image correcting for a desktop printer
From: Marco Ugolini
So, to be complete in my explanation, I should have
said that if, for
example, you are in RGB and wish to add Cyan to the
image, you can either
*decrease Red* or *increase Green and Blue together*
(by roughly equal
amounts, plus an adjustment in the composite RGB curve
to
compensate for the increase or decrease in luminosity).
If by "add cyan" you mean the equivalent of
adding cyan ink, then you have to turn down the red. People who think in
terms of ink think in terms of adding darkness. People who think in terms
of light think in terms of adding lightness.
--
Ciao,
Paul D. DeRocco
___________________________________________________________________________
Date: Thu, 01 Dec 2005 19:27:23 -0800
From: Marco Ugolini
Subject: Re: Re: Image correcting for a desktop printer
I was making a point in terms of abstract color theory.
In those terms, adding Cyan means increasing the
appearance of a Cyan-like component. In an RGB image, Cyan is defined as
either the absence of Red or the combination of Green and Blue.
Ciao.
--------------
Marco Ugolini
Mill Valley, CA
___________________________________________________________________________
Subj: Re: [colortheory] Re: Image correcting for a
desktop printer
Date: Friday, December 2, 2005 7:20:29 AM
From: J Walton
So far I have found this thread fascinating.
Everyone has attempted to make the connection between RGB and CMYK
easier to understand. What's fascinating to me is that everyone's
explanation is so complex! I keep thinking, "why don't they just
put it this way, the way that makes so much sense?"
Obviously there is no problem with the explanations, we
just have different ways of storing this information in our brains and
making sense of it. And that's just interesting.
So, Mike? When's the next Curvemeister challenge?
J Walton
___________________________________________________________________________
Subj: Re: [colortheory] Image correcting for a desktop
printer
Date: Friday, December 2, 2005 6:17:52 PM
From: RJay Hansen
On Dec 1, 2005, at 3:29 PM, Marco Ugolini wrote:
___RJay___
They will cite the irretrievable "data loss"
or as Terry Wyse more
correctly pointed out "color loss".
___Marco___
In other words, a larger gamut sRGB is squeezed into an
overall smaller
one (say US Web Coated SWOP v2). Once done, you cannot
go back. So the
loss *is* irretrievable (though we can quibble whether
it's "data" or
"color", since color lives in the file as
data too, after all).
---
No argument there.
___RJay___
Terry also pointed out that, especially w/an sRGB
profile, many photos
are very close to the cmyk gamut even when still in rgb
mode.
___Marco___
Yes, but that all depends on a careful use of profiles
and conversion
techniques, that is, on the "color meaning"
assigned to the sRGB values
within the workflow.
Though markedly smaller than AdobeRGB, sRGB is still
quite larger than US Web Coated SWOP v2 or US Sheetfed Coated v2 -- except
for an area in the Cyan-Greens where both these CMYK spaces exceed sRGB by
a considerable margin. (You can easily verify all this by graphing these
profiles in 3D Lab mode in Chromix ColorThink).
---
Yes, I'm very aware that there are many colors in an
rgb gamut that you can't hit with a cmyk gamut, just as a cmyk gamut
contains colors that rgb can't reproduce well.
___RJay___
I work in prepress so my experience w/color started
with cmyk and I'm
still much more comfortable working in cmyk than rgb,
although I've
gotten much better with rgb over the past two or three
years. Still, I
routinely convert photos from my (consumer level)
digital camera to
cmyk to make corrections/adjustments if I think I can
accomplish them
better than in rgb and then convert back to rgb for
printing on my
inkjet or posting to my website.
___Marco___
If that works out for you, fine. In other instances,
gamut-chopping may
not be the best choice if done just for the sake of the
operator's
comfort level and at the expense of some achievable
gamut in possible
output scenarios.
---
I'm referring to a specific, known output scenario.
Also, I'd point out that attempting to retain all achievable gamut is a
worthy goal if it is actually going to contribute to or enhance what you
are trying to accomplish with the final image. That may or may not be the
case, depending on the image and the purpose it is intended to serve.
___RJay___
With a bit of experience you can begin to easily
distinguish colors
that likely are out of the cmyk gamut
___Marco___
Unfortunately, there is more than just one "CMYK
gamut". In today's
world, one never knows whether a given job is going to
end up printed
on a newspaper or a magazine, traditional halftone or
stochastic, CMYK
or Hexachrome, glossy or matte, high-quality or
low-quality paper, in
the US or Europe or Asia, etc.
These possible scenarios all differ substantially, and
one should not overestimate one's ability to keep them all straight in
their head. Speaking for myself, I certainly cannot do that. Caution is the
catchword.
---
The original questioner was asking about printing his
photos on his desktop inkjet printer. A very specific scenario.
___RJay___
and judge whether you can live with the diminished
saturation and
possible hue shift they exhibit after converting to
cmyk.
___Marco___
The question I usually ask myself is not whether *I*
can live with it,
but whether *the client* can. I err on the side of
caution and struggle
to preserve the widest possible range of choices that
the client will
then be able to choose from.
---
Once again, the original questioner was referring to
*his work* on *his printer*. I took it that he only had to please himself.
I also said I do this with photos from my consumer level camera that I'm
working on for my own use.
As for clients, I know the presses in my shop and the
expectations of customers who request color correction from us. I haven't
received any complaints yet. Obviously there are many scenarios where your
points are very valid however, in the situations I was addressing they
don't necessarily apply.
___RJay___
For a quick visual example, you can look at Mike
Russells's
curvemeister challenge and compare a couple of images
done by list
members. If you go to the results of the challenge here:
http:
//www.curvemeister.com/Challenge/Challenge04/index.htm
then view picture number 12, done by Marco Ugolini and
picture number
21 done by RJay Hansen (yours truly), you can compare
one done
completely in rgb by Marco and one done w/a combination
of rgb and cmyk
done by myself.
___Marco___
I will leave it to others to express an opinion on the
merit, and also
ask anyone interested in this to read the text
accompanying my
submission.
I would like to point out that depending on one's
browser to judge the
color results of this Curvemeister challenge is
questionable at best.
It's a pity that there seems to be no way (none that I
can find) to
download the images submitted to this
"contest", so that one is able to
view them more accurately within Photoshop.
---
I can't argue with your point about browsers and color.
I've considered whether people viewing the challenge results are getting a
reasonably accurate look at what's been done to the entries. That's going
to be dependent on their hardware and setup. So then I wonder why people
like you and me bother to submit an entry to the challenge? ;-)
___RJay___
Setting aside the difference in the way we individually
chose to
"interpret" the photo (I went for a higher
contrast appearance than
Marco did), and just focusing on color, you don't see a
significant
difference in saturation or hue between the two photos.
___Marco___
Again, I wish I could download the image to be able to
view it in
Photoshop. Until I do that, I reserve my judgment.
---
If you really want, I'd be happy to email you the jpeg
I sent to Mike Russell. I even still have the large, unflattened photoshop
file for my entry.
___RJay___
I'm not saying there's anything at all wrong about
confining your
corrections and adjustments to rgb. I do that myself in
some cases
(it's photo dependant), but at the same time, I don't
think you should
feel you must rule out a conversion from rgb to cmyk
and then back to
rgb for printing if the photo and situation call for
it.
___Marco___
If that works for you, it's all good and fine. But
there are situations
when that may not be enough.
---
I hope I've made clear that I'm not trying to say he
shouldn't do all correction in rgb *if that works for him*. I'm just trying
to point out that he shouldn't feel he *has* to stay in rgb and that all
his photos will always print worse if he jumps into cmyk.
RJay
--
http://www.rocketjam.com
___________________________________________________________________________
Subj: Re: [colortheory] Image correcting for a desktop
printer
Date: Friday, December 2, 2005 6:32:27 PM
From: Rick Gordon
What you can do is view the source code for the page
showing the large images. Look for something like this (using Marco's as an
example):
<img name="Image22" src="http:
//www.curvemeister.com/Challenge/Challenge04/Results/12/Original-fullsize_Ugolini.jpg"
border="0" height="512" width="768">
You can then find the pathname of the modified graphic
you're looking for. (Or if you don't see the full pathname, just tag image
name onto the exiting directory path). You can then load that in your
brawser and download it.
Rick Gordon
___________________________________________________
RICK GORDON
EMERALD VALLEY GRAPHICS AND CONSULTING
___________________________________________________
PHONE: 415-663-8652
WWW: http://www.shelterpub.com
___________________________________________________________________________
Subj: Re: [colortheory] Image correcting for a desktop
printer
Date: Friday, December 2, 2005 9:33:23 PM
From: Stuart Larson
I noticed that there is a link to the corrected image
by Mike DeSantis that says "Download Mike's CMYK tiff file", but
that was the only one I noticed that had such a link.
Stuart Larson
___________________________________________________________________________
Date: Sat, 3 Dec 2005 16:12:51 -0000
From: "Peter Leyland"
Subject: Reality vs expectation
------------------------------------------------------------------------
-------
Marco:
I would respectfully submit that, if by saying
"one side leaning towards standards, repeatability and other worthy
aims" one wishes to characterize, among others, those who adopt
colour-managed techniques, and by "the other preferring what they
see" one refers instead to those who work "by the numbers",
then this is a false dichotomy.
We work together, not against each other.'
------------------------------------------------------------------------
--------
Peter - It wasn't really my intention to place the
creator and the
service provider on opposite sides. If indeed that is
what you thought I
meant? I have been reading, and at times (most times)
struggling to
follow, the discussions, arguments and merits of
various profiles
between the three colour authors with regards to CMYK
Workflow. I might
be naive here but amongst the other nuggets of
information gained I am
fairly clear in my own mind that given that no two
people see exactly
the same thing there remains the basis for
disagreement... No matter how
minor and arguably insignificant (reality vs
expectation). One could
then add the various output intents available on the
typical high end
digital printer/press: Photographic, Relative and
Absolute Perceptual.
Most Postscript RIPs can vary the dot shapes as well as
line screens and
no doubt all these variables affect the output (why
else have them
available)
Colour Management in a closed loop scenario is a fairly
simple concept
but once you expose those files (tagged of course) to
the wide world the
subject seems to grow arms and legs.
------------------------------------------------------------------------
---------
Color management techniques, when done with common
sense and
respect of the results, are far from denying reality.
They are about
helping manage it.
------------------------------------------------------------------------
---------
Totally agree. However life is rarely that simple and
'we' are some way
away from the point when it will be possible, perhaps I
should qualify
that and say certain, that the originator of a digital
image would be
able to give that file to anyone anywhere and have any
degree of
certainly of the result of the printed output. Bold
statement and whilst
I have no doubt that properly applied colour management
will reduce the
differences to arguably acceptable differences - that
to my mind is
Reality over Expectation.
------------------------------------------------------------------------
---------
Though some individuals *may* prove to be dogmatic
proponents of
these solutions, most people I know use CM as another
tool (a proven and
very helpful one) to achieve final results that match
expectations.
------------------------------------------------------------------------
--------
Peter - We're on the same side believe me but, there's
always a but, I
see lots of different people (customers) in a week and
whilst very few
seem to know the slightest thing about CM we have a
few, some even
encouraged by myself, who do. Expectation - an Inkjet
print on photo
gloss paper / Reality No: 1 - a laser print on quality
bond / Reality
No: 2 - a litho print on 130gsm Matt Art paper. Not
everyone
understands, some want to - others already know
everything they what to
know...
------------------------------------------------------------------------
---------
Now to get back to the real point of my original post -
Henry was on the
verge of talking about various dot screens and altering
the colour
curves on his presses to meet the expectations of his
blue chip
customers. I am interested to hear what he has to say
and no doubt
equally interested to hear what comments and issues or
thoughts this
produces from others on the list.
So please Henry by all means continue with your
thoughts before you
expose your soft side to us lot but please don't let
that put you off.
Peter
___________________________________________________________________________
Date: Tue, 06 Dec 2005 21:51:12 -0800
From: Marco Ugolini
Subject: Re: Image correcting for a desktop printer
In a message dated 12/2/05 12:32 PM, RJay Hansen wrote:
I'm referring to a specific, known output scenario.
Also, I'd point out
that attempting to retain all achievable gamut is a
worthy goal if it
is actually going to contribute to or enhance what you
are trying to
accomplish with the final image. That may or may not be
the case,
depending on the image and the purpose it is intended
to serve.
Right. So, in the prepress phase, it would seem that
preserving the widest possible gamut is the wise course when the final
destination is not known beforehand, given that different CMYK scenarios
inhabit different areas of visible color.
The original questioner was asking about printing his
photos on his
desktop inkjet printer. A very specific scenario.
Then that is all the more reason not to confine the
printed gamut of an inkjet printer to what fits in a CMYK space, which in
most cases is considerably smaller than what that device is capable of
producing.
Once again, the original questioner was referring to
*his work* on *his
printer*. I took it that he only had to please himself.
I also said I
do this with photos from my consumer level camera that
I'm working on
for my own use.
True. But you were mentioning your work in prepress.
That is what I was keeping in mind.
As for clients, I know the presses in my shop and the
expectations of
customers who request color correction from us. I
haven't received any
complaints yet. Obviously there are many scenarios
where your points
are very valid however, in the situations I was
addressing they don't
necessarily apply.
OK. We disagree there, which is perfectly acceptable.
I can't argue with your point about browsers and color.
I've considered
whether people viewing the challenge results are
getting a reasonably
accurate look at what's been done to the entries.
That's going to be
dependent on their hardware and setup. So then I wonder
why people like
you and me bother to submit an entry to the challenge?
;-)
Didn't read the fine print, I guess... No, wait, what
fine print?
If you really want, I'd be happy to email you the jpeg
I sent to Mike
Russell. I even still have the large, unflattened
photoshop file for my
entry.
Please do (offline). I will send you mine, if you would
like that.
I hope I've made clear that I'm not trying to say he
shouldn't do all
correction in rgb *if that works for him*. I'm just
trying to point out
that he shouldn't feel he *has* to stay in rgb and that
all his photos
will always print worse if he jumps into cmyk.
Hopefully, once his level of expertise grows, he will
be confident enough to determine for himself which images would benefit
from any one of these different methods, as opposed to the possible others.
Best regards.
--------------
Marco Ugolini
Mill Valley, CA
___________________________________________________________________________
Date: Thu, 8 Dec 2005 23:24:57 EST
From: Dan Margulis
Subject: Re: Reality vs expectation
Marco writes,
Sorry again if I am not interpreting you correctly, but
what is a
"closed-loop color-management scenario"?
Color management is a very open
architecture, where players can enter freely and
without restrictions. That
is its strength. Nothing proprietary about it. The
standards are
well-publicized.
This comes close to being a claim of a monopoly on
virtue. It seems to parrot the claims of some of the more rabid ICC-philes,
namely
1) Any form of color management that is not ICC color
management is not a form of color management.
2) Any profile that is not an ICC profile is not a
profile.
Both statements are quite analogous to saying that any
camera that is not manufactured by Canon is not a camera.
Obviously, effective color management is possible in a
closed-loop scenario--that's been the standard among the better printers
for nearly forty years. Effective color management is also possible in an
ICC scenario. But ineffective color management is possible in either.
The keys to effective color management are process
control and common sense. Without process control, any form of color
management fails. With it, almost any form works. Without common sense, we
fall into calibrationism, where we automatically believe whatever a machine
tells us. With it, we make judicious changes in procedures to guarantee
closer matches.
Dan Margulis
___________________________________________________________________________
Date: Thu, 08 Dec 2005 21:18:12 -0800
From: Marco Ugolini
Subject: Re: Reality vs expectation
In a message dated 12/8/05 8:24 PM, Dan Margulis wrote:
Oh, Dan, itching to argue again...
I did not create the terminology, but here is how it
goes: "color management" in this day and age has come to mean
something very precise in the field of imaging, namely the use of ICC-based
profiles and techniques to control and predict the color appearance of
digital image files in a variety of conditions and scenarios.
Naturally, in real-world situations, closed-loop or
not, these newer techniques will, and ought to, still mix with others from
a previous era of"manual color management" (which is what we may
choose to call this more traditional brand of appearance-control
techniques).
You may wish to say that this is confusing to some, who
would like "color management" to mean something other than what
it has come to mean today. Speaking for myself and those I know who agree
with me, "we color management bigots" certainly do not wish to
monopolize the virtues of color correction. More simply, "we"
believe we have something valuable to contribute, the same as you do (and I
will not attempt to challenge that, neither for the sake of trying to score
a cheap point in my favor nor for any other reason), so your charge of
righteousness in the "CM ranks" is slightly off the mark.
One thing seems certain: color management will not
change its name, and that's likely to be the way that cookie will keep
crumbling. We should try to make peace with that and with each other,
instead of tilting at windmills.
Best regards.
--------------
Marco Ugolini
Mill Valley, CA
___________________________________________________________________________
Date: Fri, 9 Dec 2005 00:15:52 -0500
From: Terry Wyse
Subject: Re: Reality vs expectation
Dan, Marco,
On Dec 8, 2005, at 11:24 PM, Dan Margulis wrote:
This comes close to being a claim of a monopoly on
virtue. It seems to parrot
the claims of some of the more rabid ICC-philes, namely
1) Any form of color management that is not ICC color
management
is not a form of color management.
I think it's GENERALLY assumed that "color
management" in today's world could be meant to imply ICC color
management but of course "we", as in us that have been
around a while, now realize that we were using a form of color
management since the days of camera separations and drum scanners
where we attempted to adjust unsharp mask filter ratios, camera
exposures and, later on, drum scanner analog knobs.
2) Any profile that is not an ICC profile is not a
profile.
Maybe ICC "purists" would say that but those
of us who might be a bit more "open" to other technologies
would include "proprietery" options such as device linking
technologies in products from Left Dakota, Alwan, GMG, CGS/ORIS and
others. For those that live where the "rubber meets the
road", we'll use whatever technology that works. But those
technologies still involve 1) a machine to measure with (spectro) and 2)
some kind of characterization data, concepts embodied in ICC color
management as well.
Obviously, effective color management is possible in a
closed-loop
scenario--that's been the standard among the better
printers for
nearly forty years.
Effective color management is also possible in an ICC
scenario. But ineffective
color management is possible in either.
I certainly agree in your interpretation of
"closed-loop" color management but I must confess that the
first thing that came to mind when that term was used in this thread
was truly close-loop color management at the press
where...ahem...machine measurements of a press sheet are taken and
that data is interpreted and fed directly back into the press's ink
control system. That's at least what is meant by closed-loop color
management in pressrooms today.
The keys to effective color management are process
control and common sense.
Without process control, any form of color management
fails. With it, almost
any form works. Without common sense, we fall into
calibrationism, where we
automatically believe whatever a machine tells us. With
it, we make judicious
changes in procedures to guarantee closer matches.
And process control is really only possible by
"employing" machines to do the grunt work. I'll certainly
tell the machine what I think is the "correct" color but I
could never rely on my eyeballs to keep things "correct"
consistently. I'm too moody and have to many good/ bad days to be trusted
to keep a color system in check. And I lost my calibration plaque years
ago.
:-)
Calibrationally Yours,
Terry Wyse
_____________________________
WyseConsul
Color Management Consulting
704.843.0858
http://www.wyseconsul.com
http://www.colormanagementgroup.com
___________________________________________________________________________
Date: Fri, 9 Dec 2005 10:43:39 -0000
From: "Peter Leyland"
Subject: Plumbers and CM
From: Marco Ugolini
Color management may not eventually become a
push-button deal after all,
but even in that case it would still be (a) very
useful, (b) of proven
value, and (c) not applicable without considerable
study and training.
After all we do not accuse plumbers of practicing the
black arts...or do
we?
Hi Marco
Interesting analogy that. I think though you miss the
point or at least mine which may be due to the way it was put? I guess I
advocate that CM will need to be made simple be integral and standardised
to the point that it is inbuilt with every application, operating system
and piece of hardware before CM will work as universally intended.
Whilst CM is relatively new and therefore: technically
challenging and therefore expensive to adopt, you will in the main only
find its adoption in the larger printers and by the more technically
aware/demanding user. I can fully understand that some will feel it
necessary, advantageous even, to employ the services of a plumber such as
yourself. The larger organisation will have both the resource and indeed
the mindset to outsource such technical services. Another factor being that
the new generation of image setters and digital presses often come with
workflow software and CM pre-installed. My experience shows the technical
support staff of the vendors to be especially weak on such subjects (most
have a strictly mechanical background) opportunity or what? I can accept
the need for a plumber and the service engineer come to that. I can even
understand the need to get up to speed on CM and there are obvious benefits
in employing a consultant BUT you won't figure in everyone's budget -
more's the pity...
However, for us smaller fish it is a different story.
We didn't get where we are: working seven days a week at all hours and all
for a pittance by relying on others. If we need to know something we jump
in right at the deep end splash about and unravel things - whatever it
takes. Me, I bought the books of which I put The Real World of Colour
Management right at the top. I joined lists such as this and to give credit
where it is due read most, if not all, of your comments and observations
both on this list and the Colour Sync users one. I think I have gained much
from the experience.
When we move down a further level in the pecking order
where most people and companies reside all you will find is 'expectation'
plain and simple expectation. They're not interested, not in the slightest
and why should they be - they have other demands made of their time. If the
image looks good on their computer and prints out OK on the $50 inkjet
where's the problem? Incase we forget, this is where the majority of the
customers live and one should never lose sight of the customer... At the
moment 'we' blame them for our problems simply because we know all about CM
and they don't. We have it they don't! There is a fundamental flaw in such
logic - it's bad for business.
As I write this and essentially think on my feet it
starts to become fairly obvious where this is going and who is going to
drive it. At the service provision level (the print houses and copy shops)
it will be the hardware manufacturers who have the ongoing issue of
service. They already provide workflow software and the colour management
tools but rarely provide the backup needed to make use of it. Now they need
to improve the interface (software) to dumb the whole thing down a level or
two. Many, perhaps most of us will already use products such as Enfocus
Pitstop and Server (again there is a whole industry of guru support
services available) but some of us you will have guessed manage all by
ourselves - my next job perhaps? Give it a few years and I am sure that
this end of the business will have been cracked with most service providers
being CM aware and practising. While I'm on this roll I also run Windows
Server 2003 with about 14 workstations a mix of Macs and PCs, host my own
and customer websites and for good measure can run the whole show from home
in my spare time.
Garbage in - garbage out still remains the primary
issue. When un-CM meets CM things really can go pear-shaped. My mate Murphy
is always at work here and too often proved right I suspect. It would
therefore seem logical, to me, that this will be the next big thing and CM
will indeed become both main stream and available to all. It will work away
in the background totally unnoticed but completely accepted as a normal
event by the millions. Most of us only challenge the simple things in life
we accept the complex without question. To use your plumber analogy. I am
sure most of us readily understand the mechanics of plumbing we say that we
just don't want to get our hands dirty. Same thing with the builder we all
did that in nursery school. Everyone can use the internet or switch on the
TV but I'd bet not many can explain how it works. One day CM will be seen
in exactly the same light (pun intended).
Perhaps it should be titled: Make hay while the sun
shines?
Best regards
Peter Leyland
PDQ Print Services
93 Commercial Street
Dundee DD1 2AF
Tel: 01382 201778
___________________________________________________________________________
Date: Fri, 9 Dec 2005 11:34:22 -0000
From: "Peter Leyland"
Subject: FW: Reality vs expectation
Hi Marco
Sorry again if I am not interpreting you correctly, but
what is
a "closed-loop color-management scenario"?
Color management is a very
open architecture, where players can enter freely and
without
restrictions. That is its strength. Nothing proprietary
about it. The
standards are well-publicized.
Closed loop (my interpretation) being where the input
and output devices are in the same location and therefore under the control
of one person. Apologies to our photographer friends if I am being too
general but many of them, I suspect, do operate in a closed loop scenario.
I am sure you have worked it out by now that I am a
'service provider' in fact we go a little further as I have graphic
designers together with both litho and digital output. CM within the
workplace is a qualified 'simple' and under control. However we also accept
files from a whole variety of sources and skill levels the majority (over
95%) with no evidence of any CM applied, or understood for that matter (you
will have a job for a while yet I'm sure, albeit I suspect, the cynic in
me, you generally preach to the converted?)
The 'problem' with CM is that too few use it and until
it is widely adopted by the main stream not much will change but at least
you will be gainfully employed. For that to happen (your retirement not
employment) it (CM) will need to be dumbed down and 'free'. Again for that
to be I suspect that the issue of standards and the adoption of same by all
the respective manufactures will need to be addressed e.g. until TWAIN came
along scanning from different applications was almost impossible. I know
absolutely nothing about whether or not this is even on the cards let alone
likely to happen. That said I installed a $50 dollar Epson scanner for a
neighbour last night and it came with basic ICC profiles and tagged scans
with either sRGB or Adobe RGB didn't notice anything about calibration
though... Perhaps universal CM is not so far away after all? Applying CM
standards to printers is the real issue...
Mmmm, do I really thinks that life is simple? I
couldn't make a
living if that were the case. There would be no need
for my services!
You and all those Management Guru's? Seriously though
as long as CM remains in the exclusive domain of the professional it will
cause as many problems as it solves in the real world of the service
provider. I am sure that I am not alone in applying best judgement to the
majority of customer files we receive as we convert whatever to suit our
imposed CM workflow.
Regards
Peter Leyland
PDQ Print Services
93 Commercial Street
Dundee DD1 2AF
Tel: 01382 201778
___________________________________________________________________________
Date: Thu, 08 Dec 2005 20:39:54 -0800
From: Marco Ugolini
Subject: Re: Reality vs expectation
Let me be slightly facetious and say that, though
plumbing skills are still rather mysterious to most (me included), the
value of plumbing is well-proven, and plumbers are well-paid and
widely-needed specialists who do a fairly competent job most of the time.
And if they tell us not to close those faucets too
tight, so as to preserve the washers, or give us other bits of advice, we
just try to do as we are being advised to do, without too many questions.
We trust the plumber and his knowledge. And if something else breaks, we
call for his services: we do not try to fix things by ourselves, unless we
either really know what we are doing, or are real macho and reckless fools.
Color management may not eventually become a
push-button deal after all, but even in that case it would still be (a)
very useful, (b) of proven value, and (c) not applicable without
considerable study and training.
After all we do not accuse plumbers of practicing the
black arts...or do we?
Regards.
--------------
Marco Ugolini
Mill Valley, CA
___________________________________________________________________________
Date: Sat, 10 Dec 2005 09:17:44 EST
From: Dan Margulis
Subject: Re: Reality vs expectation
Terry Wyse writes,
of course "we", as in us that have been
around a while, now realize that
we were using a form of color management since the days
of camera
separations and drum scanners where we attempted to
adjust unsharp
mask filter ratios, camera exposures and, later on,
drum scanner
analog knobs.
OK, but I was thinking more in terms of getting
predictable output, print-to-proof match, etc.
Maybe ICC "purists" would say that but those
of us who might be a bit
more "open" to other technologies would
include "proprietery" options
such as device linking technologies in products from
Left Dakota,
Alwan, GMG, CGS/ORIS and others. For those that live
where the
"rubber meets the road", we'll use whatever
technology that works.
Agreed.
But those technologies still involve 1) a machine to
measure with
(spectro) and 2) some kind of characterization data,
concepts
embodied in ICC color management as well.
Not to make a *profile*, not if you're interested in
quality. Would you play tennis with a hockey stick? Would you wear a
bathing suit to a formal wedding? That's what you're doing when you allow a
machine to make aesthetic decisions.
Granted, machines can make fairly satisfactory RGB
profiles, but CMYK profiling is orders of magnitude more difficult. And
granted, the machine CMYK profiles are better than they were seven years
ago, but they have a long way to go. Instead of being a total
laughingstock, as they were in 1998, they have reached a point where they
can be described, if we are being extremely charitable, as mediocre. But
they are by no means at the quality level of a human-generated profile, as
we just saw in the thread concerning SWOP v2, where the machine-generated
profile was badly beaten by two eyeball-generated profiles.
To say that a machine can generate a satisfactory CMYK
profile without human assistance is to say that a machine can judge how
well an image displayed on a monitor matches the same image in print.
Maybe in fifty years or so machines will reach that level of
sophistication.
I certainly agree in your interpretation of
"closed-loop" color
management but I must confess that the first thing that
came to mind
when that term was used in this thread was truly
close-loop color
management at the press where...ahem...machine
measurements of a
press sheet are taken and that data is interpreted and
fed directly
back into the press's ink control system. That's at
least what is
meant by closed-loop color management in pressrooms
today.
Ah. All of a sudden the hockey stick is being used to
play hockey, and the bathing suit is being donned when you want to go
swimming. Comparing the value of a printed swatch to a desired value is a
something that a machine does very well. Adjusting ink on the fly when it
detects an anomaly is another thing that a machine does well.
Quality-conscious printers had adopted these methods
long before "today's color management." It is interesting to note
that, if Chris Murphy's statement is correct, the TR 001 standards run, the
embodiment of "today's color management", was not run with the
inking that good traditional process control--including the use of
machines--would have achieved.
And process control is really only possible by
"employing" machines
to do the grunt work.
Exactly right. The difference between the successful
color manager and the calibrationist is that the successful color manager
treats the machine as his servant and not his master.
Dan Margulis
___________________________________________________________________________
Date: Fri, 09 Dec 2005 18:07:43 -0800
From: Marco Ugolini
Subject: Re: Plumbers and CM
In a message dated 12/9/05 2:43 AM, Peter Leyland wrote:
Interesting analogy that. I think though you miss the
point or at least
mine which may be due to the way it was put? I guess I
advocate that CM
will need to be made simple be integral and
standardised to the point
that it is inbuilt with every application, operating
system and piece of
hardware before CM will work as universally intended.
I hope too that CM will be integrated and standardized
one day across all applications that deal with color output. And CM
hardware may eventually come already built into printing hardware, and be
touted as a selling point.
Even then, though, it may still not be as simple as
pushing a button: I dare to predict that there will still be a need to make
informed decisions.
Whilst CM is relatively new and therefore: technically
challenging and
therefore expensive to adopt,
Well, not so expensive any longer. There are
alternatives that are relatively good and priced low enough for people
whose color needs are not too demanding.
you will in the main only find its adoption in the
larger printers and by the
more technically aware/demanding user. I can fully
understand that some will
feel it necessary, advantageous even, to employ the
services of a plumber such
as yourself. The larger organisation will have both the
resource and indeed
the mindset to outsource such technical services.
Another factor being that
the new generation of image setters and digital presses
often come with
workflow software and CM pre-installed.
Being that I do not in service bureaus, I have no
knowledge of that, but, if that is so, I bet that getting a handle of how
these CM mechanisms operate is still a challenge, and not at all child's
play.
My experience shows the technical support staff of the
vendors to be
especially weak on such subjects (most have a strictly
mechanical background)
Oftentimes, CM is still an afterthought in many
consumer-level devices, not yet an integral part of a well-thought-out
feature package.
opportunity or what?
For the manufacturers, yes, who could be smart enough
and do a better job at improving the way their products handle
color-managed workflows, and then publicize how good their products are at
producing predictable color.
Which is something that they don't do yet.
I can accept the need for a plumber and the service
engineer come to that. I
can even understand the need to get up to speed on CM
and there are obvious
benefits in employing a consultant BUT you won't figure
in everyone's budget -
more's the pity...
Oh, you put the finger where it hurts...
Unfortunately, if I were to base my livelihood on my
color management assignments, I would not even be able to pay the rent,
specially in the SF Bay Area. They represent a small fraction of my total
work, which still mainly consists of image retouching, color correction and
mechanical production.
You have no idea how resistant people in the design
world (which is where I operate) are to hiring the services of a color
management consultant. They would rather spend money in untold amounts of
wasted consumables trying to hit a certain elusive color result than spend
a little money now to save lots of money later. And those very few times
when they do say yes, they want to pay next to nothing for my services, and
pinch every penny. It's enough to make you want to give up, believe me.
However, for us smaller fish it is a different story.
We didn't get
where we are: working seven days a week at all hours
and all for a
pittance by relying on others. If we need to know
something we jump in
right at the deep end splash about and unravel things -
whatever it
takes. Me, I bought the books of which I put The Real
World of Colour
Management right at the top. I joined lists such as
this and to give
credit where it is due read most, if not all, of your
comments and
observations both on this list and the Colour Sync
users one. I think I
have gained much from the experience.
I am honored to hear that, honestly. I had no idea that
there are people who are actually interested in what I have to say.
When we move down a further level in the pecking order
where most people
and companies reside all you will find is 'expectation'
plain and simple
expectation. They're not interested, not in the
slightest and why should
they be - they have other demands made of their time.
If the image looks
good on their computer and prints out OK on the $50
inkjet where's the
problem? Incase we forget, this is where the majority
of the customers
live and one should never lose sight of the customer...
At the moment
'we' blame them for our problems simply because we know
all about CM and
they don't. We have it they don't! There is a
fundamental flaw in such
logic - it's bad for business.
Yes, it's silly to look down on the average Joe and
Jane consumers and poo-poo their "ignorance." Clearly, not
everyone who uses a camera or an inkjet printer needs to know about color
management. Most want prints that are just good enough, the same way most
people are satisfied when the 1-hour photo place returns photos that do not
look too crappy. The margins for customer satisfaction there are very wide.
Still, if and when consumers start realizing that color
can be controlled more closely and reliably, and that there are ways to
improve the look of their images at just a very small extra cost if any,
they will become educated very quickly, and start to demand improved
results in very short order.
As I write this and essentially think on my feet it
starts to become
fairly obvious where this is going and who is going to
drive it. At the
service provision level (the print houses and copy
shops) it will be the
hardware manufacturers who have the ongoing issue of
service. They
already provide workflow software and the colour
management tools but
rarely provide the backup needed to make use of it.
Yes, and often they seem neither to offer
well-thought-out tools, nor to have themselves a good grasp of how the
whole thing is supposed to work, judging from the way they talk about their
CM support features in some of their instruction manuals. There is great
room for improvement there.
Now they need to improve the interface (software) to
dumb the whole thing down
a level or two. Many, perhaps most of us will already
use products such as
Enfocus Pitstop and Server (again there is a whole
industry of guru support
services available) but some of us you will have
guessed manage all by
ourselves - my next job perhaps? Give it a few years
and I am sure that this
end of the business will have been cracked with most
service providers being
CM aware and practising.
I will become a sandal-wearing practicing Buddhist when
that happens...
It would therefore seem logical, to me, that this will
be the next big thing
and CM will indeed become both main stream and
available to all.
I eagerly await the day. What have I to lose? It's not
like I am making that much money from it anyway...
It will work away in the background totally unnoticed
but completely accepted
as a normal event by the millions.
Unnoticed except when needed, at which point one better
know what to do to get what is wanted. It's never going to be a brain-dead
kind of thing, I am quite certain of that.
Most of us only challenge the simple things in life we
accept the complex
without question. To use your plumber analogy. I am
sure most of us readily
understand the mechanics of plumbing we say that we
just don't want to get our
hands dirty.
But do we really understand the mechanics of plumbing,
though we would like to think we do?