[argyllcms] Re: RGB Printer Profiling / Issues with Blacks and Shadows
- From: "Alan Goldhammer" <agoldhammer@xxxxxxxxxxx>
- To: <argyllcms@xxxxxxxxxxxxx>
- Date: Tue, 02 Aug 2011 13:23:21 -0400
Andreas,
Your Argyll settings are pretty much the same ones that I use. One thing I
did notice, I don't think you can use the '-f' command with colprof without
going through and measuring the spectral data (and you need an i1 pro
without the UV cut filter to do this). I don't know if this is the issue or
not. I have not done this with any of the papers I print on and thus have
not used the '-f' command. I also use the default chartread commands (no
'-H' or '-T' but again don't see that these would make a difference)
Here are some other thoughts:
How old is the matte cartridge in your printer? Some of the pigments may
have settled and you might try taking it out and shaking it back and forth
to distribute the pigments (if they did settle you would not have as deep a
black but I'm not even sure this is the issue if you got a deep black using
another software program but it could be if the iMatch profile was done some
time ago). You can confirm this by one of two ways: 1) print out the test
print using NO color management at all and measure the black patch; it
should read in the neighborhood of 1.6 since Dmax should not be affected; or
2) put your chartread data into an Excel spreadsheet and see where the black
patch is on your test charts and read that patch to get the Dmax value; I'm
not sure you can get it directly from the T3 data since these are in
colorspace values and I don't know how to convert them Bruce Lindbloom has a
good sight with color calculators on it: http://www.brucelindbloom.com/ ).
You can use the '-L' command to also print out LAB values next time you read
a chart since there is a direct correlation between L* and density. I went
back and looked at my Hahnemuhle Photorag Ultrasmooth profile data and I got
a Dmax of 1.7 using Argyll profile tools.
If you colors look OK, there has to be something wrong with the black ink
and it could not be printing on the wrong side of the paper since you would
also see washed out colors. The only other thing to double check is that
you used the correct Epson paper selection in the driver (which is why I
always set up presets in the Epson driver to make sure I automatically have
the right paper selection; I've had more than enough instances of problems
without doing this and things print wrong; remember the default is to Luster
with PK ink I beleive).
Alan
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx]
On Behalf Of Andreas Hebeisen
Sent: Tuesday, August 02, 2011 11:11 AM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: RGB Printer Profiling / Issues with Blacks and
Shadows
Hello again
Thank you, Phil and Alan, for sharing your thoughts about this.
First, an important correction to my first message.
Meanwhile I realized that I've soft-profed/printed sample 4) accidentally
with the wrong profile, namely the profile created with Eye-One iMatch
software instead of the mentioned ArgyllCMS-Profile.
Therefore I've got the same black level as with sample 1) and it has nothing
to do with the render intent but only with the differences between the two
profiles.
So the issue with the lost shadow details is expected behavior when using
the "Relative Colorimetric without BPC" ... nice demonstration that BPC is
quite useful with this render intent. ;)
But this also means that the issue with "black not being real black" with
ArgyllCMS-generated profiles remains.
Thanks to Phils and Alans answers, I realized that I also had the problem
with the silently added sRGB-profile, which has also been described here:
http://www.luminous-landscape.com/tutorials/solving.shtml
Because I've always used EyeOne iMatch to print the targets, I never faced
this problem in the past. But now, printing the TIFF-files generated by
ArgyllCMS "by hand", it also got me caught.
So I did a rerun of the complete process (which I describe below) but now
printing the targets with the Adobe Color Printer Utility mentioned by Phil.
Unfortunately, the issue with the blacks remains.
When I look at my test prints made with the ArgyllCMS-Profile, the colors
look nice, but as a whole, everything looks like the printer wasn't throwing
all the ink out he usually does with every other profile/paper/settings
combination I used to work with.
It looks dull like a print done on a plain A4 laser printer paper without
special ink jet coating. The potential of the EPSON Enhanced matte paper
isn't really played out. It's like driving while forgetting to release the
handbrake.
In numbers; here are the Dmax values of the square black field in the top
right corner of the mentioned evaluation image, printed on EPSON Enhanced
Matte Paper:
EyeOne iMatch Profile: 1.6 or more (as expected with matte papers an K3 ink)
ArgyllCMS Profile: 1.33 or less
Here are the steps I used to create the new profile with ArgylCMS:
1. targen -v -d2 -G -f1764 -g128 2011_V6-0_3800mk_Epson-EMP_EMP_1440_0_8bit
2. printtarg -v -ii1 -t360 -pA4 2011_V6-0_3800mk_Epson-EMP_EMP_1440_0_8bit
3. Four target sheets printed with "Adobe Color Printer Utility".
Color Settings to "Off (No Color Adjusment)" in the "Print Settings"
part of the printing dialog.
4. Letting the papers dry for something between 1 and 2 hours.
5. chartread -v -H -T0.4 2011_V6-0_3800mk_Epson-EMP_EMP_1440_0_8bit
6. colprof -v -qu -i D50 -o 1931_2 \
-S /Library/Application\
Support/Adobe/Color/Profiles/Recommended/ProPhoto.icm \
-cmt -dpp -f 2011_V6-0_3800mk_Epson-EMP_EMP_1440_0_8bit
7. Printing the evaluation image from Apple Aperture with render intent set
to Perceptual and
another print with render intent set to Relative Colorimetric with
activated BPC.
The AgryllCMS settings used above were inspired by:
http://www.luminous-landscape.com/forum/index.php?topic=53408.msg436392#msg4
36392
My issue also appears when I use much simpler settings as used in this
guide:
http://www.argyllcms.com/doc/Scenarios.html#PP1
And again, for those who like to take a deeper look, all the files (except
the TIFF-images) generated with the commands listed above, downloadable from
here:
http://home.ggaweb.ch/files/mail/2011/2011_V6-0_3800mk_Epson-EMP_EMP_1440_0_
8bit.zip
@Alan: Is it possible that you can send me the settings of your
profiling-workflow and printer-settings you're using with the Hahnemuehle
Matte Papers? This way I can make another try with settings known as working
and delivering a Dmax I also like to see.
Thanks again.
Andreas
On 01.08.2011, at 13:28, Andreas Hebeisen wrote:
>
> Hello there.
>
> On my quest to find a affordable solution which lets me use my
1ipro-Instrument also with OS X Lion, i've stumbled upon ArgyllCms.
> Which will hopefully also be supported on OS X Lion some time in the
future; something Xrite obviously isn't willing/able to do for my existing
software Eye-One Match 3. At least not for an reasonable amount of money.
>
> Ironically, I'm facing the same problems which I already had to deal with
when I started working with Eye-One Match 3 some years back.
> Luckily, i found this article
> http://people.csail.mit.edu/ericchan/dp/i1GamutMapping/index.html
> which unveiled a hidden feature within the software allowing me to get the
results I intended.
>
>
> First some details about my working environment:
>
> OS: Apple Mac OS X Snow Leopard (10.6.8)
> Instrument: Gretag/Xrite eye-one-Pro, Rev. D
> Printer: Epson Stylus Pro 3800
> Paper/Media type: Epson Enhanced Matte Paper
> Color Matching Settings in printer driver while printing test charts:
Epson Color controls
> Color Settings in Epson printer driver while printing test charts: Off (No
Color Adjustment)
> ArgyllCMS: latest version, 1.3.3
>
> For all my testing I use the A4 printer evaluation image from Uwe
Steinmüllers site.
> It can be downloaded here (scroll down to the bottom of the page):
> http://www.outbackprint.com/printinginsights/pi048/essay.html
> My version is converted to my working color space, which is ProPhotoRGB,
but besides that, i didn't change anything in the file. I'm doing all the
soft proofing and printing from Apple Aperture, using the profiles generated
with Eye-One Match and ArgyllCms. Aperture offers the render intents
Perceptual and Relative Colorimetric, with and without Black Point
Compensation.
>
>
> For creating my very first ArgyllCms profiles, I used the steps described
in this post:
>
http://www.luminous-landscape.com/forum/index.php?topic=53408.msg436392#msg4
36392
> Except that I went with the ProPhotoRGB source profile instead of
AdobeRGB.
> Since the author of this post created a profile for the Hahnemuehle Bamboo
paper, which I also use from time to time, I figured this would be a good
starting point.
> Except that I'm using Epson Enhanced Matte paper at the moment. Mainly for
going easy on the budget while tampering with a new software and cranking
out tons of test prints. ;)
>
>
> And finally the problem:
>
> When I print the evaluation image mentioned above, I don't get the blacks
and shadow details I'm looking for on this matte paper. The problem already
shows up in the softproof view of Aperture and remains when the (matte
black) ink hits the paper.
> In the attached image, I've tried to visualize the problem. It contains a
crop from the bottom left corner of the evaluation image with different
profiles and render intent settings.
>
>
> Image explanations and comments from top to bottom:
>
> 1) Profile created with Eye-One Match 3, set to "NewGamutMapping L0.3"
using the trick in the article I've mentioned at the beginning.
> The render intent was set to "Perceptual" which always seemed to give me
the results I've wanted in the past years using profiles created this way.
>
> Blacks: With this profile/setting combination the maximum possible Dmax
with the used ink/paper combination can be reached.
> A crucial criteria, as everyone printing on matte Fine Art Papers probably
figured out.
>
> Shadow separation: pretty good; on a printed sample, I'am able to
distinguish between patch no. 10 from the surrounding black with ease.
>
>
> 2) Profile created with ArgyllCMS. Render intent: Perceptual.
>
> Blacks: lousy, as one can see in the image, the black is more of a very
dark grey, than really black.
> I didn't bother measuring the Dmax since this is obviously pretty far away
from what this print/ink/paper combination is able to deliver.
>
> Shadow separation: slightly better than with 1) but with black being dark
grey not really a surprise.
>
>
> 3) Profile created with ArgyllCMS. Render intent: Relative Colorimetric,
Black Point Compensation ON
>
> Blacks: as lousy as with 2), on paper even worse
> I didn't bother measuring the Dmax since this is obviously pretty far away
from what this print/ink/paper combination is usually able to deliver.
>
> Shadow separation: slightly worse than with 2)
>
>
> 4) Profile created with ArgyllCMS. Render intent: Relative Colorimetric,
Black Point Compensation OFF
>
> Blacks: great, that's the way I like it. But only on paper, in the soft
proof view, I see little difference while turning BPC on and off.
>
> Shadow separation: not really existent, on paper, even patch no. 24 isn't
really separated from the surrounding black!
>
>
> Besides those findings, I have to mention, that everything else is as
expected.
> Good color match to (eye-One Match calibrated) screens, nice saturation
etc.
> For my eyes, the color looks best with Relative Colorimetric and BPC
turned off.
> But this has also to do with the darker blacks and therefore more contrast
in the print.
>
> A little side note: When I use Apples Color Sync Utility to compare the
two profiles mentioned here, the ArgyllCMS profile never reaches the bottom
of the given space, as the eye-One profile does. Seems there is a connection
between what i'm struggling with and what ColorSync shows. Or ist this just
coincidence? Am I confusing two different things here?
>
> Whatever.
> At the moment i'm _really_ looking for deep, really dark blacks the way I
get them with sample 4) but with the shadow details on a level close (or
even better) to the ones in 1).
> I've played around for hours with different settings, searching the
internet, performing all kinds of CMS voodoo rituals I'm aware of; but non
luck.
> Still stuck to "not-really-black" blacks and lousy shadow separation.
>
> Any help out there?
> Someone else must have seen this before or am I going mad?
> Thanks for _any_ help in advance!
>
> Cheers
> Andreas Hebeisen
>
>
>
> P.S. For those who want to analyze my issue, you can download the ti3-File
and the created ICC-Profile from this URL:
> http://home.ggaweb.ch/files/mail/2011/2011-08-01_profiles.zip
> I've also included the mentioned ICC-Profile I created with Eye-One Match
3 for the same printer/paper/ink combination.
>
> The attached images are also downloadable from there in the original
resolution:
>
http://home.ggaweb.ch/files/mail/2011/Profile-Comparisons-by-Print+Softproof
-Preview-in-Aperture.jpg
> http://home.ggaweb.ch/files/mail/2011/ColorSync_Profile_Comparison.jpg
>
>
<Profile-Comparisons-by-Print+Softproof-Preview-in-Aperture.jpeg><ColorSync_
Profile_Comparison.jpeg>
Other related posts:
