Yves,
Have you ever considered building and using a ‘synthetic’ profile that would
match your wide gamut display, instead of using ProPhoto which has, as we all
knows, imaginary primaries?
Also, I find your 80 Cd/m2 on little on the “low” side. It won’t solve your
“’out-of-gamut” problem but, for color accurate work, why not bump that up a
bit?
Canon’s “Canned” profiles may not give you the best for your money.
I can see you are not complaining about the “color match” between screen to
proof but you are using a D65 white point on your display which, in my very
humble experience, is quite different from a D50-ish calibration.
Still, I think your problem of “managing out of gamut colors” does not seem
impossible to resolve… You might consider doing your own colorimetric analysis
that don’t use Photoshop or ColorThink. And from there, see which one is
“telling the truth”. Adobe does things, here and there, that bothers me,
sometimes, and I get frustrated at ColorThink often unstable “behavior”…
/ Roger
From: argyllcms-bounce@xxxxxxxxxxxxx <argyllcms-bounce@xxxxxxxxxxxxx> On Behalf
Of Yves Gauvreau
Sent: Tuesday, November 26, 2019 3:11 PM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: Rendering intent?
Samuel,
I've setup my little workplace as close as possible to some standard I don't
recall the number off but when I think of it all of this is futile as very few
real world situation comes close to this standard.
My image are 99.99% raw files from a D610 and if needed I always convert them
to tiff 16 bits + ProPhoto RGB and for internet and only on final image I use
jpeg and sRGB.
10 bit LCD monitor + 10 bit video card + wide gamut monitor calibrated D65 +
Gamma 2.2 + Brightness of 80 cd/m2 ambient at +- 5000k and 50 lux.
Characterization done with DisplayCal + I1 Display Pro 2 or I1 Studio
Print viewing, 5000k Nominal CRI > 80 , 500-2000 lux.
Work done mostly in LR and PS and soft proofing in PS using generic profile
from paper manufacturer at this time. I intent to create my own in the near
future.
I'm printing on a Canon PRO-1000 with Lucia pigment inks, I always use the 16
bit driver and more and more I prefer using relatively wide gamut smooth and
textured museum grade matte papers but of course I also print on more glossy
paper sometimes but I try to stay away from papers with OBA. I call them my
blue papers.
Below you mention this "Is the profile for output properly constructed?" I
don't know since I haven't build one myself yet. Argylllcms does recommend a
fairly large number of patches to make a profile and the I1 Studio software is
doing this with a fraction of this, max 2 sheet of 8.5" x 11". Obviously the
results must differ but until I try I wont know.
I made a print (Red River Palo Duro Etching) the other day of a very colorful
image and I'm pretty satisfied of the result but the contrast could still be
improved a little I guess, PS gamut warning showed very little out of gamut
colors but this morning I use ColorThink to show off the same image and paper
combination and the result greatly differ, much more out of gamut colors. I
don't know which to trust now. It could explain why my prints don't look much
like the PS softproof version. This image protrude mainly above and below the
paper gamut, changing saturation would be useless, no out of gamut color in the
a or b direction , the only change that would work is a change in L*.
This is one off me biggest concern, who and which software can I trust to do
the work properly. I already know PS does many thing in our backs that are
questionable, others that are inaccurate, etc. So I would be surprised if I
shouldn't trust some or all their color handling stuff as well.
Until I workout for myself a pretty good workflow to realize a pretty good
screen matching print consistently, something like you said above. I'll do both
a soft proof and small hard proof(s) until I get it right.
Some suggestions to get me on the right track as soon as possible?
Thanks,
Yves
On 11/26/2019 8:54 AM, Samuel Chia wrote:
Hi Yves,
I'm glad you enjoyed the lengthy treatise about gamut mapping. Yes,
unfortunately, we still do not have a proper solution to this problem after all
this time. Beautiful, top-notch gamut mapping is the last great hurdle of
digital imaging yet to be breached for the reproduction of general photographs
of fine quality. Many still don't understand just how important this is for the
visual quality of their prints so it is rarely discussed, understood, and thus
there is little pressure on profiling software makers to make radical changes
to their mapping. A chicken and egg problem. Most folks just prefer to use
Relative Colorimetric with BPC, without even questioning why the out of gamut
colours map the way they do. Let alone (what everyone calls) the mystery meat -
Perceptual.
If you are printing images that are within the gamut of your output medium but
you are getting unpredictable results with significant loss of contrast,
something is wrong with your workflow. Too dark output could be many things. Is
your print viewing light bright enough? Is your monitor too bright relative to
your print viewing environment? Do you use a surround of white pixels at least
an inch all around for your image on the monitor? Is the profile for output
properly constructed? Are all the printer settings correct for the chosen
paper? Etc. Etc. It would be difficult to give more specific advice without
knowing more details about your situation.
I'm finding I enjoy printing on smooth matte fine art papers a lot more than
glossy baryta papers these days. The measured Dmax and gamut difference is
significant, and without proper luminance-preserving gamut mapping I couldn't
love the medium as much. In real-world situations, glossy media always, I mean
always, has some degree of glare which reduces its contrast and gamut. If it
catches a reflection, it can reduce the viewer's ability to see the printed
image almost entirely, and suddenly we are faced with the reality that matte
media, on the contrary, can have significantly better contrast and gamut in the
real world. Modern pigment inksets are finally beginning to give us just barely
enough Dmax to make matte prints sing when you want punchy blacks, though the
Lucia inks are still somewhat lacking in that regard. Overall gamut is pretty
good if one does not over-saturate their images, though by then, even glossy
media may not look good enough compared to the eye-popping colours on OLED
displays.
Regards,
Samuel
On Tue, 26 Nov 2019 at 20:08, Yves Gauvreau <gauvreau-yves@xxxxxxxxxxx
<mailto:gauvreau-yves@xxxxxxxxxxx> > wrote:
Samuel,
I find your "Full story" most interesting (sorry I'm not good with words, I
think your paper deserve more but, I don't know...) and from below I see the
problem isn't fully resolve yet, that is, if I understand correctly.
As you said yourself,
"The ideal result is one that comes visually closest to the
original image, what we saw after spending many hours editing the image on
our monitors to exactly what we like"
I would believe most if not all people would agree with you and this goal,
especially in the case of "photographic" image. Beside trying to find a paper
that as enough gamut to include our image colors or trying to eliminate out of
gamut colors from our image for this or that paper, still, such images (no
oog), very often, don't print as the soft proof image looks like and don't look
like the original either, lack of contrast flat looking, to dark, etc.
What would you suggest we do to get as close to this goal as possible most if
not all the time?
Thanks,
Yves
On 11/25/2019 5:27 PM, Samuel Chia wrote:
Hello Yves and all interested in this topic,
The mapping for Colorimetric is well-defined for in-gamut colours, while there
is no so-called proper way to map out of gamut colours, and that is where
different profiling software can still give different Colorimetric table
results. Usually, they forgo way too much lightness to preserve some
saturation, in an attempt to make the shortest path into the output gamut.
For Perceptual, we have been misled to believe that the only possible mapping
is a non-linear compression type, and so the mapping is "unpredictable" and
thus completely different secret recipes from one software to the next. In
fact, there is nothing preventing the Perceptual intent from being treated the
same way as Colorimetric for in-gamut colours, and when we open ourselves to
this way of thinking, we can consider something else that is much more
important.
I would like to correct the concept that the shortest movement in 3D space is
necessarily the best conversion when it comes to regular photographs. As
Florian has pointed out, lightness shifts are most noticeable while chroma
shifts are the least.
When many people do gamut mapping comparisons, they like to use synthetic
gradients to evaluate the mapping of profiles for both Colorimeteric and
Perceptual tables, which gives a completely false impression of how well they
work for ordinary photographs, which normally vary significantly in hue and
lightness in a small region, not smoothly graduating from one tone to another,
one colour to another, and usually only have small regions of out of gamut
colours. Thus, lightness-preserving gamut mapping is significantly better at
maintaining the appearance of the original image, especially when out of gamut
colours are concerned, to avoid loss of detail due to excessive contrast
reduction. I've explained this to some length previously in an earlier email to
the list, here is the link to that. Lightness-preserving gamut mapping is the
key to getting successful optimal results on relatively low contrast matte
media as opposed to glossy media, however, despite my many attempts at
encouragement to Graeme to implement this properly, including paying him an
hourly fee and travelling to Australia to work with him full time to get this
correct and eliminate communication misunderstandings over email, I have not
been successful at convincing him. Graeme did implement a luminance-preserving
gamut mapping function for Colprof, however, it is not working as well as it
should. Despite mathematically being a 100% luminance preserving operation
according to the code, it operates in CIECAM space, and some additional
movements are happening, such as the Helmholtz-Kohlrosch compensation, which is
still way too strong and still resulting in significant contrast loss with out
of gamut colours of dark colours
Full story here:
https://www.freelists.org/post/argyllcms/Reddish-cast-in-shadows-for-Perceptual-rendering-intent-and-commentary-on-gamut-mapping
Allow me to provide two visual examples. These are for printing on HP Indigo on
coated media, which has a gamut virtually exactly the same as Fogra39, so let's
compare to the way the Fogra39 profile is doing gamut mapping in Relative
Colorimetric with BPC to Argyll's Lightness-Preserving mapping, which works
well here because the out of gamut colours are light colours rather than dark
colours. The sunrise hues are out of gamut and you will see the textural
differences preserved much better with luminance-preserving gamut mapping than
with Fogra39, which is preserving more saturation (and thus shorter movement in
3D space) but looks worse:
https://www.dropbox.com/s/qaj99v3o234giva/Luminance-preserving-example-1.jpg?dl=0
In this second example, the sky and clouds in the top right corner are out of
gamut in the output media, and need to be mapped into printer space. Normal
gamut mapping sacrifices some luminance contrast (not all of it of course, that
would be terrible and no software does this), to the point where the clouds
have lost definition against the sky. Argyll's luminance mapping is preserving
this contrast difference, despite the much greater saturation loss. However, if
you look only at the rightmost image, you would still consider it to look fine
despite the saturation decrease, while in the middle image, you would not even
know that clouds were were defined in the upper right sky in the original:
https://www.dropbox.com/s/jixuv0x12770btq/Luminance-preserving-example-2.jpg?dl=0
Best regards,
Samuel
On Tue, 26 Nov 2019 at 05:58, Yves Gauvreau <gauvreau-yves@xxxxxxxxxxx
<mailto:gauvreau-yves@xxxxxxxxxxx> > wrote:
My only "questioning" with using only saturation to bring a color into
the gamut of the destination colorspace is that its a 1d operation and
it may not be the shortest path to the gamut boundary (destination) in
other words, a possibly less obvious conversion may exist. The answers
of Florian suggest other approach would difficult to implement with the
current CLut structure.
Not much to do with the term itself.
Thanks Roger,
Yves
On 11/25/2019 12:15 PM, graxx@xxxxxxxxxxxx <mailto:graxx@xxxxxxxxxxxx> wrote:
Yves,
Look into *any* Adobe profiles, they have all three mandated ICC intents.
I can see you don't like the term "Saturation", consider the term "Chroma"
instead.
/ Roger
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx <mailto:argyllcms-bounce@xxxxxxxxxxxxx>
<argyllcms-bounce@xxxxxxxxxxxxx <mailto:argyllcms-bounce@xxxxxxxxxxxxx> > On
Behalf Of Yves Gauvreau
Sent: Monday, November 25, 2019 11:50 AM
To: argyllcms@xxxxxxxxxxxxx <mailto:argyllcms@xxxxxxxxxxxxx>
Subject: [argyllcms] Re: Rendering intent?
Roger,
I don't like the idea of just decreasing the saturation at least from a
numerical perspective. From what I've read we are not that good at
discerning colors, it takes serious effort on our part and practically
laboratory conditions to discern colors with less than 2-5 DE?? of
difference. If this is right, just changing saturation would probably not
result in optimal result especially at low L* for example. It maybe a
"shorter" path to increase L* then to vary a* and b* such as to vary only
the saturation. This should be especially true if the shortest path also
provide a less discernible color difference, doesn't it?
As for PS you could be right I don't know but it could be also because the
printer profile doesn't provide the 2 rendering intent tables?
Yves
On 11/25/2019 10:17 AM, graxx@xxxxxxxxxxxx <mailto:graxx@xxxxxxxxxxxx> wrote:
Yves,documentation?
How does Argyll do Perceptual Intent?
I have not read anything about that. It's probably in Argyll
But, in general, it's a matter of decreasing saturation to move the out ofgamut colors inside the Destination gamut, presumably preserving the
original hue angle, but there's a whole science in how to do this, and not
all profilers do it the same. Some move greens further in, others move reds
further in and so on. Many will use a decreasing saturation function in
that, only out of gamut colors will be progressively desaturated at the
gamut's edge. Others will apply a wholesale desaturation, decreasing all
Source colors so that they fit inside the Destination gamut. There's a lot
of good papers out there on how this is done. But there's nothing like
finding it out for yourself empirically. It's not hard to analyze but it's
time consuming, without some kinds of analytical tools.
Take Photoshop. Photoshop is easy because there is no difference betweentheir Perceptual and Colorimetric intents -- to make their lives easier and
to avoid users calling them on the telephone to complain about the
"difference".
/ Rogerwhich tool you use, right? How Argyllcms does this?
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx <mailto:argyllcms-bounce@xxxxxxxxxxxxx>
<argyllcms-bounce@xxxxxxxxxxxxx <mailto:argyllcms-bounce@xxxxxxxxxxxxx> >
On Behalf Of Yves Gauvreau
Sent: Monday, November 25, 2019 9:17 AM
To: argyllcms@xxxxxxxxxxxxx <mailto:argyllcms@xxxxxxxxxxxxx>
Subject: [argyllcms] Re: Rendering intent?
Roger,
this would mean that "perceptual" as a different meaning depending on
Is the same true for Relative colorimetric? Is only a change in saturationconsidered with the relative intent or something else?
My understanding of a color is that it is defined by a point in some 3dimensional colorspace and if you have to move it around and you limit
yourself to only changing saturation your giving up 2 other potential
options. I would have thought the most recent "distance" metrics or even
the simple euclidean distance to be more appropriate then just shifting the
saturation as seems to be the case in the various docs I've seen. I know I
may not have seen the "right" one. That's why I ask.
this.
On 11/25/2019 8:23 AM, graxx@xxxxxxxxxxxx <mailto:graxx@xxxxxxxxxxxx> wrote:
Every profiler I know does the Perceptual intent *differently*.
It's considered their "secret sauce".
/ Roger
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx
<mailto:argyllcms-bounce@xxxxxxxxxxxxx> <argyllcms-bounce@xxxxxxxxxxxxx
<mailto:argyllcms-bounce@xxxxxxxxxxxxx> >
On Behalf Of Yves Gauvreau
Sent: Monday, November 25, 2019 8:21 AM
To: argyllcms@xxxxxxxxxxxxx <mailto:argyllcms@xxxxxxxxxxxxx>
Subject: [argyllcms] Rendering intent?
Hi,
I'm curious, every time I read some doc on how relative colorimetric
and perceptual intent works, it seems as if it only a saturation (2D)
change is considered, why not the shortest distance (3D) to the gamut
boundary of the destination space?
Can you direct me to something that would explain how Argyllcms does
Thanks,
Yves
