Out of Gamut Colors.
Robert E. Williams
are "out of gamut". I understand that to mean that those colors cannot
be reproduced accurately in the CMYK color space used in printing.
However, some upscale inkjet printers use a 6 color, CcMmYK color space.
Is the CcMmYK color space any wider than the CMYK space.
Do the 6 color inks make any difference on which colors are out of
gamut?
If it did, there would be no way PS could know about it, because PS
doesn't know how you are going to print your image.
Just curious.
Bob Williams
Matti Vuori
news:3F598BB7...@cox.net:
> Is the CcMmYK color space any wider than the CMYK space.
Why else would they use those extra inks?
> Do the 6 color inks make any difference on which colors are out of
> gamut?
Of course.
> If it did, there would be no way PS could know about it, because PS
> doesn't know how you are going to print your image.
Of course it does, when you set up your color management and tell what
printer (what profile) you use.
--
Matti Vuori, <http://sivut.koti.soon.fi/mvuori/index-e.htm>
Johan W. Elzenga
> "Robert E. Williams" <myt...@cox.net> wrote in
> news:3F598BB7...@cox.net:
>
> > Is the CcMmYK color space any wider than the CMYK space.
>
> Why else would they use those extra inks?
Because there can be other reasons and there are. 'm' is just diluted
magenta, so the gamut doesn't neccessarily change. It probably does NOT
change. It's just the printers ability to print smooth light colors
rather than using some kind of rendering which changes. Inkjet printers
are not 'continuous tone' printers and the extra inks make them a little
better in that respect.
Compare this to the grey ink sets you can now buy for B&W printing. With
only black ink, you can print all 256 shades of grey, that's not the
problem. But it will be done by printing a raster and that makes the
prints 'grainy'. With extra grey inks you still need a raster (because
you still do not have all 256 shades of grey ink, just three or four
extra shades), but the result is a lot smoother.
Conclusion: CcMmYK ink does NOT neccessarily have a wider gamut than
CMYK. It MAY have, but more likely than not it doesn't. That doesn't
mean that inkjet printers can't have a wider gamut than a printing
press, but that is another matter.
--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/
Gernot Hoffmann
Yes, thatæ„€ my opinion too.
http://www.fho-emden.de/~hoffmann/gamuts08072002.pdf
Files for inkjets are normally in RGB. A simulation of the
CMYK result requires an accurate ICC profile in PhS.
Made for the machine, the ink, the dithering mode and the media.
Canned profiles by manufacturers are useless.
Best regards --Gernot Hoffmann
Tacit
No.
The m is just lighter magenta. The c is just lighter cyan.
The purpose of CcMmYK is to reduce the appearance of dithering in light colors.
If you print very light, pastel colors on a 4-color consumer inkjet printer,
you'll sometimes notice visually objectionable dithering; the 6-color printers
print very light colors much more smoothly.
--
Rude T-shirts for a rude age: http://www.villaintees.com
Art, literature, shareware, polyamory, kink, and more:
http://www.xeromag.com/franklin.html
Bill Hilton
>Is the CcMmYK color space any wider than the CMYK space.
Depends on the inks, but generally yes.
>Do the 6 color inks make any difference on which colors are out of
>gamut?
Yes.
>If it did, there would be no way PS could know about it, because PS
>doesn't know how you are going to print your image.
Photoshop knows when you tell it to use something other than the default CMYK
gamut.
For an inkjet printer the inks don't have a gamut in and of themselves, only
when printed on paper, and different papers have different gamuts for the same
inks.
You can easily check for this in Photoshop 6 and 7 with "soft proofing" if you
have an ICC profile for the printer/paper/ink combo by going to View > Proof
Setup > Custom and under "Profile" pick the profile for your paper/ink/printer.
Once it's loaded click View > Gamut Warning to see if you have out of gamut
colors for THIS profile.
You can also do Window > Documents > New Window to create a dupe view of the
file and assign a different profile to this view so you can see both on the
sceen side by side for comparison, which is very handy.
To do this right you need accurate ICM profiles for the printer/paper/ink.
Epson has a composite ICM file for most of its printers that has all the
profiles bundled together and you can't soft proof to this, but the later
models like the 2200 include the profiles separately on the install disk if you
want to load them, and the 1280 class printers have the ICM files available if
you load the PIM module.
You can also load ICM profiles from commercial printers to see how they
compare, for example I have profiles from Calypso Labs' LightJet 5000 loaded to
compare to the Epson profiles. You can download these for free from Calypso.
It's eye-opening to get profiles from different places for the same printer and
compare them ... I have Epson 2200 profiles from Epson and some much larger
(presumably more accurate) ones from George Lepp which give very different
results and have different gamuts.
The boys at Adobe did a nice job with soft-proofing, I feel. Very useful *IF*
you have accurate printer profiles and your monitor is calibrated.
Bill
pro...@att.net
>
> >From: "Robert E. Williams" myt...@cox.net
>
> >Is the CcMmYK color space any wider than the CMYK space.
>
> Depends on the inks, but generally yes.
>
> >Do the 6 color inks make any difference on which colors are out of
> >gamut?
>
> Yes.
> >If it did, there would be no way PS could know about it, because PS
> >doesn't know how you are going to print your image.
>
> Photoshop knows when you tell it to use something other than the default CMYK
> gamut.
>
> For an inkjet printer the inks don't have a gamut in and of themselves, only
> when printed on paper, and different papers have different gamuts for the same
> inks.
>
> You can easily check for this in Photoshop 6 and 7 with "soft proofing" if you
> have an ICC profile for the printer/paper/ink combo by going to View > Proof
> Setup > Custom and under "Profile" pick the profile for your paper/ink/printer.
> Once it's loaded click View > Gamut Warning to see if you have out of gamut
> colors for THIS profile.
Didn't realize that Gamut Warning is profile specific. As long as we on
Soft Proofing, here are a few other questions:
What does it mean when a color changes drastically when Soft Proofed,
and yet that color has no Gamut Warning?
What is the meaning of the Proof Color values in the Info Palette, and
how can one make use of them?
It would be great if there is a way to know what the acceptable color
values (rgb and cmyk) are for a particular profile. With that, one could
avoid them during editing and before getting to the Soft Proofing stage.
Also, if one only intends to print with a certain profile and media in
mind, is there any reason not to set the working space to that profile
all the way through the workflow?
> You can also do Window > Documents > New Window to create a dupe view of the
> file and assign a different profile to this view so you can see both on the
> sceen side by side for comparison, which is very handy.
Agreed. Do that all the time as the last step before printing.
> To do this right you need accurate ICM profiles for the printer/paper/ink.
> Epson has a composite ICM file for most of its printers that has all the
> profiles bundled together and you can't soft proof to this, but the later
> models like the 2200 include the profiles separately on the install disk if you
> want to load them, and the 1280 class printers have the ICM files available if
> you load the PIM module.
Soft Proof would allow you to use the Epson composite ICM file, but the
proof would not look any different. I think that's what you meant.
> The boys at Adobe did a nice job with soft-proofing, I feel. Very useful *IF*
> you have accurate printer profiles and your monitor is calibrated.
Soft Proofing is indeed very useful, but I wish PS could have provided
some answers to the above questions.
Bill Hilton
>Didn't realize that Gamut Warning is profile specific. As long as we on
>Soft Proofing, here are a few other questions:
>
>What does it mean when a color changes drastically when Soft Proofed,
>and yet that color has no Gamut Warning?
Dunno ... I don't really see this myself, if the device cannot duplicate a
color then by definition that color should be out of gamut.
Typically when I turn on 'simulate: paper white' the screen image gets a lot
duller, reflecting the reduced dynamic range of paper compared to your screen.
That's the biggest change I normally see ... if you're getting "drastic"
changes in colors then maybe your profile for that device is off? Or your
monitor isn't accurately profiled? Not sure what you're seeing here ... I
don't see this for printer profiles myself.
>What is the meaning of the Proof Color values in the Info Palette, and
>how can one make use of them?
The RGB triplet values are shifted for the new profile to give you the closest
possible match in color on the profiled device. I guess this feature tells you
what the numbers will get changed to. Not sure how you'd use this info though.
If this isn't clear, check this article by alpha color geek Bruce Fraser which
shows what RGB value 247/160/91 looks like to a scanner, monitor and printer
(all slightly different), and then explains how the device ICM profiles have to
change the scanner RGB numbers to 250/175/100 for the monitor and 244/192/148
for the printer to get the best color match.
http://www.creativepro.com/story/feature/13605.html
Fraser and two co-authors wrote an excellent book, "Real World Color
Management", which is worth reading if you're really interested in this topic.
>It would be great if there is a way to know what the acceptable color
>values (rgb and cmyk) are for a particular profile. With that, one could
>avoid them during editing and before getting to the Soft Proofing stage.
It's pretty easy to just turn a profile on and off to check as you work,
especially if you save it as a custom preset. Many pro users feel most out of
gamut colors occur in darker saturated colors and they don't worry about them
as much as you and I might. If highlights or skin tones are out of gamut
that's different, but if it's mostly in the darker tones or shadows maybe it's
not a big deal?
You can also buy one of the profile editors or viewers that plots profiles on
2D or 3D graphs. From these you quickly get an idea of gamut width for
different profiles and which colors are likely to be out of gamut first. This
is a very handy tool.
>Also, if one only intends to print with a certain profile and media in
>mind, is there any reason not to set the working space to that profile
>all the way through the workflow?
This is a good way to start a holy war :) A lot of the pre-press guys working
with a scan targeted for a certain CMYK press prefer to work this way. This is
often done for a magazine or book or catalog where the image will be used one
time for a known set of print conditions.
On the other hand, guys like me who will be printing to different RGB printers
like an Epson (typically to several different papers) or a LightJet 5000, and
maybe making web images from the same files at some point, would be better off
editing in one of the abstract RGB working spaces. I use three of them (sRGB,
AdobeRGB, Joe Holmes' Ektaspace), depending on the source of the image and its
destination.
>> Epson has a composite ICM file for most of its printers that has all the
>> profiles bundled together and you can't soft proof to this
>Soft Proof would allow you to use the Epson composite ICM file, but the
>proof would not look any different. I think that's what you meant.
You can soft proof (sorta) to it, but it's not accurate so there's no point in
doing it.
There are two problems with trying to soft proof with the Epson composite ICM
file. To take the 2200 as an example, the first problem is that there are 6
Epson papers for the printer, with different black inks recommended for glossy
vs flat finishes, and the papers each needs a different profile. If you select
the generic "Stylus Photo 2200" profile you don't know which paper is getting
profiled. I just dinked with this and it looks like the generic profile uses
PGPP, which means it's giving an inaccurate view for the other 5 papers. You
can check this by opening up multiple windows and comparing the composite ICM
profile to something like the Watercolor profile, for example, which is
radically different.
The other problem is that with the composite profile you can't turn on
'simulate: paper white' because the whites are different between papers and
Photoshop doesn't know which one to use, so you get an overly optimistic proof
of the brightness range, even for PGPP. You can *print* to the generic Epson
profile since the Epson driver figures out which paper you're using from your
paper settings, but you cannot accurately soft proof using it. This isn't a
big deal with the newer Epsons since you can download the 1280 profiles off
their site or the 2200 profiles are included on the install disk, though you
have to load them separately when you install.
>Soft Proofing is indeed very useful, but I wish PS could have provided
>some answers to the above questions.
Try a program like Corel Painter 8 to really appreciate how well Photoshop
handles Color Management (grin). It would be nice if they include an ICM
viewer and maybe an editor in PS 8.
Bill
Robert E. Williams
Bill Hilton wrote:
Thanks for the interesting and informative answers.
I'm still slugging it out with PS 5.0 and don't have such niceties as Soft Proof
:-(
Probably should bring myself into the 21st century with PS 7.0 or wait a bit longer
for v 8.0.
Only thing is, I get such good results with V 5.0, it's hard for me to see how my
prints could be significantly better.
My Epson 780 prints match the quality of any online printer I've used.
Bob
Flycaster
> Didn't realize that Gamut Warning is profile specific. As long as we on
> Soft Proofing, here are a few other questions:
>
> What does it mean when a color changes drastically when Soft Proofed,
> and yet that color has no Gamut Warning?
The color change simply reflects how the profiled device will "see" the
colors. If the device is calibrated accurately, and the profile is good, a
zero gamut warning indicates there will be no color substitution. It rarely
happens, however...if you zoom in, you will almost always see something out
of gamut. Whether or not it is important, however, is another issue.
> What is the meaning of the Proof Color values in the Info Palette, and
> how can one make use of them?
If you are referring to the Info options, second color read out numbers, it
is the new numerical value of the adjusted color, based on the profile
interpretation. Frankly, I have no idea how this can be put to good use
since it is device dependant.
> It would be great if there is a way to know what the acceptable color
> values (rgb and cmyk) are for a particular profile.
That is *exactly* what the soft-proof does!
> Also, if one only intends to print with a certain profile and media in
> mind, is there any reason not to set the working space to that profile
> all the way through the workflow?
If you never, ever will change printers, papers or inks, probably not. It
would be a bummer, though, if you have a great image, end up buying an 11
color printer someday (who knows?), and you come to realize that you've
unwittingly clipped a bunch of brilliant colors that you could now print.
It seems like this stuff improves almost weekly.
[snip
> Soft Proof would allow you to use the Epson composite ICM file, but the
> proof would not look any different. I think that's what you meant.
No. Without the media and ink *specific* profiles, soft-proofing doesn't
work.
> Soft Proofing is indeed very useful, but I wish PS could have provided
> some answers to the above questions.
Color management is complex. There are, however, a few great books that
explain (in layman's terms) a lot about Photoshop that Adobe does not
include in the manuals.
Flycaster
didn't mean to jump in on top of the conversation.
pro...@att.net
> >Soft Proofing, here are a few other questions:
> >
> >What does it mean when a color changes drastically when Soft Proofed,
> >and yet that color has no Gamut Warning?
>
> Dunno ... I don't really see this myself, if the device cannot duplicate a
> color then by definition that color should be out of gamut.
That's my assumption as well, but is not what I'm seeing.
> Typically when I turn on 'simulate: paper white' the screen image gets a lot
> duller, reflecting the reduced dynamic range of paper compared to your screen.
> That's the biggest change I normally see ... if you're getting "drastic"
> changes in colors then maybe your profile for that device is off? Or your
> monitor isn't accurately profiled? Not sure what you're seeing here ... I
> don't see this for printer profiles myself.
Enabling and disabling 'simulate: paper white' does make a pronounced
difference, and I typically disable it. But the changes I'm seeing are
much much more drastic than this. The problem is indeed the "darker
saturated colors" you referred to later. I can ignore them if they are
small or unimportant, but they are the focal points in some images.
BTW, what should be the default cmyk working space for editing in cmyk
and for an Epson inkjet printer output? Fraser's Real World PS book went
into great lengths about each color space, but I wish he could be more
specific for the home desktop printer users.
> >What is the meaning of the Proof Color values in the Info Palette, and
> >how can one make use of them?
>
> The RGB triplet values are shifted for the new profile to give you the closest
> possible match in color on the profiled device. I guess this feature tells you
> what the numbers will get changed to. Not sure how you'd use this info though.
A Proof Color that differs from an Actual Color does not necessarily
mean out of gamut?
> If this isn't clear, check this article by alpha color geek Bruce Fraser which
> shows what RGB value 247/160/91 looks like to a scanner, monitor and printer
> (all slightly different), and then explains how the device ICM profiles have to
> change the scanner RGB numbers to 250/175/100 for the monitor and 244/192/148
> for the printer to get the best color match.
>
> http://www.creativepro.com/story/feature/13605.html
>
> Fraser and two co-authors wrote an excellent book, "Real World Color
> Management", which is worth reading if you're really interested in this topic.
I'll check out that article.
pro...@att.net
Bill Hilton
>>>and yet that color has no Gamut Warning?
>> Dunno ... I don't really see this myself, if the device cannot duplicate a
>> color then by definition that color should be out of gamut.
>From: pro...@att.net
>
>That's my assumption as well, but is not what I'm seeing.
My next guess is that the profile is not accurate or the device has changed
since it was profiles. What specific profile are you talking about?
>what should be the default cmyk working space for editing in cmyk
>and for an Epson inkjet printer output?
The Epson inkjets expect a RGB file from you, not CMYK, and the driver
translates the RGB file to CMYK (or CcMmYKk or whatever that model uses). If
you give it a CMYK file then it translates to RGB and translates back again to
the printer values. So you should be editing in an RGB space like AdobeRGB or
similar, and *not* CMYK if targeting inkjets or LightJet or Chromira printers.
>Fraser's Real World PS book went
>into great lengths about each color space, but I wish he could be more
>specific for the home desktop printer users.
I have his "Real World Photoshop 6" book (not 7) and he only has a sentence or
so on Epson's so you're right, it's light on this topic. A very good book for
photographers is "Photoshop Artistry" by Barry Haynes, et al. He talks at
length about specific scanners and Epson inkjet printers and how to get the
most out of them ... his web site is http://www.maxart.com/ and if you click on
"Latest tips" there's some info from the book and his recent magazine articles
that's specific to the more recent Epsons like the 2200 and 1270/1280. This
book has a lot of specific practical info for photographers, I feel.
>A Proof Color that differs from an Actual Color does not necessarily
>mean out of gamut?
In theory you are right. In practice there are a lot of inaccurate profiles
out there. Bill Atkinson said he spent around 1,200 hours and took around
500,000 samples while trying to create extremely accurate profiles for the
Epson 7600/9600 printers so he could use them instead of a LightJet. He said
none of the commercial software he tried was up to the task and he ended up
creating a new test patch that GretagMacbeth now uses. These profiles were
better than anything Epson could create, even with the $7,000 profiling package
Epson uses ... makes you wonder just how accurate the profiles are for the
consumer grade printers with a high delta-E between units.
You can make your own printer profiles but it's tough going unless you have a
spectrophotometer. I have printer profiling packages by Monaco and Colorvision
but both rely on making the print measurements with a desktop scanner and it's
tough getting accurate profiles from this setup, for example. The scanner is
the weak link.
Color management using this ICC flow is still relatively new, I think the
consortium was formed in 1993 or so and many vendors still don't implement it
well or at all (I'm fighting this with Painter 8 right now, it's horrible on
CM). Like many new technologies it was oversold early on, with promises of
perfect matches between devices if you just hired the right consultant and/or
bought the right tools. Then there was a bit of a backlash by people like Dan
Margulis, who felt Color Management "failed" to deliver, especially during the
Photoshop 5 era. Now the consensus among the experienced like Fraser, John
Paul Caponigro, etc is that CM gets you about 95% of the way but you still need
to print a proof and make tweaks to the original instead of relying totally on
profiles and soft-proofing.
Bill
Gernot Hoffmann
> Thanks for the interesting and informative answers.
> I'm still slugging it out with PS 5.0 and don't have such niceties
> Bob
Bob,
I am running PhS 5,6 and 7 on different computers. PhS 5 has already
the gamut warning.
Best regards --Gernot Hoffmann
Bill Hilton
>I am running PhS 5,6 and 7 on different computers. PhS 5 has already
>the gamut warning.
I don't think Photoshop 5 lets you load different profiles and run gamut check
on them, I believe this was new with version 6. Could be wrong as it's been 3+
years since I used 5 ...
Bill
Tacit
>and for an Epson inkjet printer output?
Inkjet printers expect RGB data, not CMYK data. if you send CMYK data to a
consumer-grade inkjet printer, the printer driver software will do a (crude)
conversion to RGB, then back to the printer's own peculiar value of CMYK.
pro...@att.net
Bill Hilton wrote:
>
> >>>What does it mean when a color changes drastically when Soft Proofed,
> >>>and yet that color has no Gamut Warning?
>
> >> Dunno ... I don't really see this myself, if the device cannot duplicate a
> >> color then by definition that color should be out of gamut.
>
> >From: pro...@att.net
> >
> >That's my assumption as well, but is not what I'm seeing.
>
> My next guess is that the profile is not accurate or the device has changed
> since it was profiles. What specific profile are you talking about?
Epson's 1270 hwm profile.
> >what should be the default cmyk working space for editing in cmyk
> >and for an Epson inkjet printer output?
>
> The Epson inkjets expect a RGB file from you, not CMYK, and the driver
> translates the RGB file to CMYK (or CcMmYKk or whatever that model uses). If
> you give it a CMYK file then it translates to RGB and translates back again to
> the printer values. So you should be editing in an RGB space like AdobeRGB or
> similar, and *not* CMYK if targeting inkjets or LightJet or Chromira printers.
I should be clearer with my question: another poster responded the same
way. I understand that the Epson driver expects a rgb file. (BTW, I
thought that PS would covert a non-rgb file to rgb first before sending
it to Epson. If I try to Soft Proof a non-rgb file, I believe that PS
would convert it to rgb first.)
I normally work in the rgb1998 work space, but I'm beginning to
experiment with working in a cmyk space per Dan Margulis' book. My
question is which cmyk work space should be used for editing in the cmyk
mode.
>
> >Fraser's Real World PS book went
> >into great lengths about each color space, but I wish he could be more
> >specific for the home desktop printer users.
>
> I have his "Real World Photoshop 6" book (not 7) and he only has a sentence or
> so on Epson's so you're right, it's light on this topic. A very good book for
> photographers is "Photoshop Artistry" by Barry Haynes, et al. He talks at
> length about specific scanners and Epson inkjet printers and how to get the
> most out of them ... his web site is http://www.maxart.com/ and if you click on
> "Latest tips" there's some info from the book and his recent magazine articles
> that's specific to the more recent Epsons like the 2200 and 1270/1280. This
> book has a lot of specific practical info for photographers, I feel.
One problem with books like the Real World PS and Dan Margulis' is that
they do not provide enough specific help for the home digital darkroom
amateurs. While these books are intended for the pros, there must be a
fair number of struggling amateurs who can use their expertise as well.
The Artistry book is indeed more down to earth in this sense, and very
much tailored for the photographers. Haynes seems to have a problem
releasing his long waited for book on digital printing though.
PS7 came a long way since PS5, which was prematurely released for color
management. At this stage, I'm not looking for anything that is "95% of
the way", but definitely something that would not blow out a color in
Soft Proof.
Thanks.
Flycaster
> I should be clearer with my question: another poster responded the same
> way. I understand that the Epson driver expects a rgb file. (BTW, I
> thought that PS would covert a non-rgb file to rgb first before sending
The double conversion takes place in the Epson driver, which does a poor,
and inconsistent job of the task.
>...If I try to Soft Proof a non-rgb file, I believe that PS
> would convert it to rgb first.)
Technically speaking, Photshop actually uses L*A*B* color space in the
background for all color conversions, but that is irrelevant. The soft
proof is a screen renditon of what the file will look like if actually
converted to the profile color space. In fact, if you toggle off paper
white and do a side-by-side comparison of the soft-proof and the
profile-profile preview, you'll see they are, perceptually speaking,
identical. The whole point of soft-proofing is to show the difference
between the working color space and output color space, with the additional
benefit of showing the perceptual impact of the tonal compression resulting
from the ink black and paper white starting points.
> I normally work in the rgb1998 work space, but I'm beginning to
> experiment with working in a cmyk space per Dan Margulis' book. My
> question is which cmyk work space should be used for editing in the cmyk
> mode.
Mike Russell (a regular poster here) is a strong proponent of this. You
might want to take a look at his site
(http://geigy.2y.net/DigPhoto/CMYK-RGB/CMYK-RGB.htm), and see if this
methodology works for you. When it comes to photographic output, many of my
images suffer noticeable gamut loss in the double conversion. IMO, they are
simply different color spaces designed for different output.
[snip]
> At this stage, I'm not looking for anything that is "95% of
> the way", but definitely something that would not blow out a color in
> Soft Proof.
Even with the best profiles there *should* be slight, but noticeable
changes, especially in the extreme edges of the color gamut, and there
*should* be a noticeable "flattening" of the image as well if you toggle
paper white. There is no way any spray painting device can exactly match
the color and intensity of a transmissive device like a CRT. Nonetheless,
it sounds like you have other problems?
Let's get to the bottom of this: 1, where did you get the profile you are
using?, 2, is it a 1440 or 2880 profile (yes, there *is* a difference)?, 3,
how are you calibrating and profiling your monitor?, 4, just how much is the
profile off?, 5, what rendering intent are you using?, 6, what do you mean
by "blow out a color?", and 7, what test image are you using to proof?
Bill Hilton
>Let's get to the bottom of this: 1, where did you get the profile you are
>using?, 2, is it a 1440 or 2880 profile (yes, there *is* a difference)?, 3,
>how are you calibrating and profiling your monitor?, 4, just how much is the
>profile off?, 5, what rendering intent are you using?, 6, what do you mean
>by "blow out a color?", and 7, what test image are you using to proof?
These are excellent questions ... I'd add two more -- exactly what flow are you
using to print, ie, do you turn on the soft proof or not and what Epson driver
options are you setting. And which colors are "blowing out" vs which colors
print accurately to the proof?
pro...@att.net
> > I normally work in the rgb1998 work space, but I'm beginning to
> > experiment with working in a cmyk space per Dan Margulis' book. My
> > question is which cmyk work space should be used for editing in the cmyk
> > mode.
>
> Mike Russell (a regular poster here) is a strong proponent of this. You
> might want to take a look at his site
> (http://geigy.2y.net/DigPhoto/CMYK-RGB/CMYK-RGB.htm), and see if this
> methodology works for you. When it comes to photographic output, many of my
> images suffer noticeable gamut loss in the double conversion. IMO, they are
> simply different color spaces designed for different output.
My few experiments of working in cmyk mode have produced some very
positive results. Some color corrections are much easier to achieve than
in rgb mode. Conversion between the two modes do worry me some, but I
have not noticed any gamut loss so far. I'll check out Mike's site.
> > At this stage, I'm not looking for anything that is "95% of
> > the way", but definitely something that would not blow out a color in
> > Soft Proof.
>
> Even with the best profiles there *should* be slight, but noticeable
> changes, especially in the extreme edges of the color gamut, and there
> *should* be a noticeable "flattening" of the image as well if you toggle
> paper white. There is no way any spray painting device can exactly match
> the color and intensity of a transmissive device like a CRT. Nonetheless,
> it sounds like you have other problems?
As stated earlier I did notice the difference between turning on/off
paper white in Soft Proof. In 90+% cases, with paper white off, the
proofs do match closely to the monitor images, and so do the prints.
(Close enough for me at this stage.) However, once in a while, some very
saturated deep shadow colors would be completely off in Soft Proof and
the prints. Yet there would not be any gamut warning. (More on this
below.)
>
> Let's get to the bottom of this: 1, where did you get the profile you are
> using?
It is an Epson 1270 hwm profile obtained from Lyon's site a while back.
BTW, I use genuine Epson ink and hwm.
>2, is it a 1440 or 2880 profile (yes, there *is* a difference)?
As far as I can tell, it is 1440 since the 1270 is not 2880.
>3, how are you calibrating and profiling your monitor?, 4, just how much is the profile off?
This is the usual and strong suspect. I use PhotoCal on a Dell monitor
which does not support adjusting the three guns separately. I was unable
to set up the blackpoint according to PhotoCal's target, but their tech
support said to generate the profile anyway. I don't know whether the
profile is off or not, but it won't surprise me if it is.
>5, what rendering intent are you using?
Perceptual.
>6, what do you mean by "blow out a color?"
I'll look up and post the color values.
>7, what test image are you using to proof?
When I first got started, I tried the photodisc, digitaldog, and the
Artistry targets. The soft proofs and prints are all slightly off, but
are acceptable for me at this stage. None have that blown out color
problems I see in my own images.
Great questions. Thanks.
pro...@att.net
>
> >From: "Flycaster" nos...@noyb.com
>
> >Let's get to the bottom of this: 1, where did you get the profile you are
> >using?, 2, is it a 1440 or 2880 profile (yes, there *is* a difference)?, 3,
> >how are you calibrating and profiling your monitor?, 4, just how much is the
> >profile off?, 5, what rendering intent are you using?, 6, what do you mean
> >by "blow out a color?", and 7, what test image are you using to proof?
I've responded to these question in another post, except 6.
> These are excellent questions ... I'd add two more -- exactly what flow are you
> using to print, ie, do you turn on the soft proof or not and what Epson driver
> options are you setting.
After trying out different printing methods per Haynes and Lyon, I
settled with the following, which gets me good prints most of the time:
Soft proofing:
- Duplicate image in separate window.
- Soft proof in the second window
- Add adjustment layer(s) as needed in soft proof window until the two
windows look about identical (typically reducing slight magenta cast)
- Crop and print test strip from the soft proof window. Repeat test
strip and adjustment layer(s) until satisfied.
Printing:
In the PS Print dialog box:
- Check Show More Options
- Select Color Management
- Source Space: check Proof and select hwm profile [Sometimes instead of
printing from the soft proof window, I would print from the monitor
image window. In that case, I would flatten the image and convert to hwm
profile first. Source Space would be Document and hwm profile selected.]
- Print Space: Same as Source.
In the Epson Driver:
- Use Custom mode and select a previously created hwm settings.
- Hwm settings: No Color Management, media is hwm, Print Quality is
1440, High Quality half tone, High Speed off.
> And which colors are "blowing out" vs which colors
> print accurately to the proof?
This is pretty much the same question 6 from Flycaster. Here goes. On an
image dominated by different shades of magenta, one blown out color is a
saturated and dark magenta. The actual color value is 66-8-28 rgb or
48-100-81-55 cymk. After soft proofing, the color turns much less
saturated and lighter, with a proof color value of 100-0-37 rgb. A
lighter and less saturated magenta color that stays pretty much the same
after soft proofing has an actual color value of 113-49-56 rgb and a
proof color value of 152-29-79 rgb. The print tracks both of these
colors pretty closely as seen in the soft proof window. There is no
gamut warning.
Hope all these answers can shed some light on this problem. Thanks.
Flycaster
looking at your prints with? (you are aware that if either this is off, or
you have poor monitor calibration, the majority of any help we can offer is
pretty well moot...no?)
And if you've answered this, I can't find it in the thread either.
Bill Hilton
>After trying out different printing methods per Haynes and Lyon, I
>settled with the following ... <snip>
The method you described sounds fine to me, the only tweak I'd suggest is maybe
creating a second custom preset with 'relative colormetric' rendering intent
and checking the soft proof with both this and perceptual. Sometimes relcol
looks better, depending on how the profile was made and on whether or not you
have anything out of gamut. But that's a minor point.
Since your flow is fine per generally accepted guidelines that leaves the
profile, which you mentioned you got from Ian Lyons' site. This is an
excellent site but I tried Ian's profiles for the 1280 (1290 to him) about 3
years ago and felt I got very poor results using them compared to the Epson
profiles and some custom profiles I acquired. Basically I saw a similar
problem as you, with magenta. I *think* Ian uses a flatbed scanner to generate
these profiles, which isn't the most accurate way (may be wrong about what he
uses now, but several years ago when the 1270 profile would have been generated
I think that's what he used).
The 1270 and 1280 use the exact same ink carts and at some point early on these
inks were redesigned by Epson shortly after the 'orange fade' fiasco. I think
one of the Adobe guys also said the ink production was moved to China. So
basically there were enough changes that old original profiles from the first
inks might not be as accurate as hoped. Several people have reported a magenta
tinge with the later inks and original profiles, for example. Also, as a
consumer grade printer there's a fairly big delta-E between individual units,
which could account for some units giving different results (and generating
different profiles) than other units of the same model.
FWIW, when I took a class from Bill Atkinson shortly after he finished
generating the Epson 7600/9600 profiles (which were better than anything Epson
could generate) he said he found it impossible to accurately profile the 2000p
because of the metamerism issues, and almost impossible to accurately profile
the 1280 (to his admittedly ultra-high standards) because he saw a shift every
time he changed ink lots and often a shift in colors after running just a
cleaning cycle. So maybe the degree of accuracy you are getting is about all
we can expect from these models. I have a 1280 and a 2200 and get a better
match with the 2200 using the Epson ICC files, especially in neutral tones. If
I print a black/white gradient with the 1280 I just can't keep it neutral
across the entire gradient, for example.
Hope this helps. You are probably doing as well with the 1270 as the rest of
us :)
Bill
Brian
Probably not on an inkjet printer, no.
> Why else would they use those extra inks?
Because in many cases the double cyan and magenta are there to boost
those colors, which will not increase the gamut of the device to any
significant degree (they're there simply to make sure the inkjet ink is
more capable of printing normal CMYK to a fair degree of accuracy).
If you're talking about a true six-color method such as hexachrome
(CMYKOG) then that's a different story.