D
David Chien
Cross-posting here FYI so nobody spends 6+ months trying to find a
'solution', only to realize what I found.
=======================================================
Subject: Matching print colors to Monitor screen
Ive acquired an Epson C82 which I plan to use primarily for printing
of Duotone images and some color. When I print in color, the color of
the print varies considerably from the monitor image. I've run Adobe
Gamma, as well as being careful to match the paper choice in the
driver with the paper I'm using, but not much help. After printing,
I can match the print fairly close to the monitor by adjusting it
with "Color Balance", cyan slider to -10, magenta and yellow each to -
50. The same problem and "correction" is evident with duotone
printing. Using the uncorrected monitor image, I can get the print I
want by adjusting those colors in the printer driver just prior to
printing. Is there a more "global" solution to this problem, rather
than my tedious procedure? I'm aware of the ColorVision/Spyder, etc.
approach, but that's a little tough on my budget. I'd appreciate any
help. Thanks
==========================
1) Realize that the color gamut of the monitor will =never= match the
color gamut of the printer!!!
Sorry!
This is the most crucial mistake most people make when they start
looking into color matching, and most don't realize until after spending
$$$ that the two will never match. You can stop now if you don't
understand this point.
The reason for this is that the two technologies produce mostly the
same colors, but at the outer edge of their color gamuts, they differ.
2) Realize that viewing the prints under various lights of different
color temperatures will make them look different!!!
This is another big mistake people make. If you print everything as
good as can be, but view the prints under, let's say, standard indoor
nightime tungsten lighting (soft white bulbs), you'll see something
very, very different than if you view the exact same print outdoors in
daytime, under a fluorescent bulb, etc.
The only way to ensure you're getting the 'best' color matching is
to view the prints under a standard color temperature light source (in
most cases, 5000k/5500k are common, graphics arts viewing temperatures).
You will need to get a light bulb that'll produce this light
temperature, install it, and only use this light source to view your
prints. (yes, there's what they call viewing booths used in print
presses and graphic arts that are super-duper calibrated, stable and
accurate, but well outside your budget. Besides, only the light really
needs to be fixed to a standard color temperature - so swapping out
light bulbs is the cheapest way to get almost there.)
Here, it's easiest to buy a fluorescent bulb from bulbs.com and
install it into a desktop fluorescent lamp.
Remember, == you must == turn off all other light sources (block off
daylight if you're viewing prints in daytime) and only use this standard
color temperature light source to view prints for the best evaluation of
prints.
3) Keep in mind that print colors will look different under different
light sources!!! Just because the prints look stunning under that 5000k
light source doesn't mean they'll look as good under tungsten lighting!!!
That said, you have to make a choice here - do I make prints that
look great under a standard color light temperature (just about all of
the books, magazines, prints published today are done so) but not
perfect under other lights? Or, let's say I only view and hang my
prints to view indoors under tungsten lighting, do I make them look
great under the tungsten lighting, but terrible under a standard color
light temperature or even outdoors in daylight?
This is a ====huge==== point that people fail to realize!
Just because you can print something that's 'correct and accurate'
according to 'standard practices and lighting temperatures' doesn't mean
you can't throw that out the door and print something that looks great
=under your specific display lighting conditions=!
Almost always, people will try and try and try to print an image so
that it looks perfect under 5000k lamps, then stick it on their hallway
walls that are only lit indoors by tungsten. Well, then, if you know a
bit about colors, you know that those pictures will have a strong yellow
cast over them due to the tungsten lights (you'll see this in 35mm
photographs at night indoors w/o flash; your eyes automatically adjust
after a few minutes to correct for the excessive yellow and make it seem
white). Does this make for a perfect print?!? No! Even after your
eyes have adjusted for the tungsten yellow cast, the prints still won't
look as good vs. under 5000k lamps because the print colors aren't
optimized for that lighting.
Here, you'll have to hand-tweak prints to get the very best colors
as nobody's done any real work on making prints that look great under
these conditions.
4) A locked down sRGB workflow is one easy step to perfection, and a
good one to try first.
If you have a CRT monitor, you may just have to use Adobe Gamma (if
you're using Photoshop, etc.) or a color calibration tool just to get
the monitor to look right - CRTs are terrible at giving you accurate
colors right out of the box.
Let's assume you have a LCD monitor - you'll likely be able to get
90%+ match to sRGB easily. Why? Most LCD gamuts practically sit on top
of the sRGB gamut - if you get an Apple with their Apple LCD Displays,
the color match is almost perfect.
Most of these monitors can be set to sRGB mode. If that's not
present, usually you can set it to factory default, then press an AUTO
button to get it mostly there matching wise. Look for color temperature
and sRGB mode...it helps to buy a LCD panel that already has these.
(That said, my Winbook 15" was that easy - press factory default,
press auto, and I'm 95% matched to sRGB right away according to my
ColorVision Spyder.)
* Next, you'll want to set your color workspace (if you're using a
program that handles this - most windows programs simply assume sRGB;
Adobe products should be set for sRGB), and your printer mode to sRGB
(eg. most Epson printers have this option; other printers, well, suck
because they often don't have this option), and your scanner as well
(eg. Epson, you can set in advanced mode the scanning target color space
to sRGB).
There, now that everything's locked to sRGB, what you see on screen
will generally match what you scan in, and what you print.
PLEASE!!! Keep in mind still that what you see =will never
perfectly match= what you print (point #1 above). So even now that
everything's working just peachy, prints still won't look just 100% like
the screen. Sorry!
5) So what to do? Take a test target image like the Photodisc Test
Target (free, search for it), open it up on screen, print a copy in your
locked down sRGB workflow, and compare the print to monitor -- this is,
for the most part w/o tweaking, about as good as you'll get even after
some heavy duty color calibration/management tools.
Now, what you want on print will have to be adjusted for on screen
based on what you see, experience, and what you've seen printed. But,
because you have this standard test target on screen and on paper, you
will know that when your own picture looks good vs. the test target,
it'll look good in print - you will use the test target as a standard
reference point for rapid image adjustments and tweaks ot make it look
right.
That's right! If you adjust =your= images while viewing a =standard=
test target image that prints great, you can make great prints faster
because you have something to adjust your own images to! If you only
tweak an image by itself, you'll soon find yourself wondering is that
blue really this blue or so so blue? It's because when your eyes look
at certain colors too long, they lose any objectivity and reference as
to what's neutral, good, and correct -- and it's very difficult to
adjust correctly without some reference image (like driving at night
without any points of reference - just pure darkness).
6) You'll quickly find that with the above, you can easily reduce the
process down to one or at least two tweak then print steps before
achieveing a perfect print by using an onscreen reference image and a
locked down sRGB workflow with a little practice and experience.
What you'll do is simply adjust your image to look perfect next to
that test target, print and move on to the next image.
==================================
That said, what about the color management/calibration/correction tools
that cost a bundle ala ColorVision Spyder, MonacoEZ, etc?
They do have their place -- usually, the reason is for consistency --
ie. no matter which monitor you drop in, or printer, or scanner, they'll
all give you as close of a representation of color to what can be
expected of them.
It =does not= mean they won't look different after being
managed/calibrated/corrected!!!! Only that they'll produce as true of a
representation of color as possible.
Why is this useful? If you're matching your red to CocaCola's special
red, you'd naturally want the most exact match possible. But when
printing images, that's totally different - you usually don' t care if
that color or this is perfectly matched, but rather, only that they look
stunning in combination with the rest of the picture. Here, you're an
artist, and you can choose to make the colors a bit more dramatic to
achieve the effect you want in a final print. Yes, if you're doing
catalogs for clothing, you'd want the colors of that dress to match as
close as possible to the real thing - good for the $$$ tools, but most
of the time, you can do without these tools in a sRGB workflow.
---
Q: But isn't color m/c/c supposed to give you stunning prints?!?
A: NO!!!!
Color m/c/c/ only gives you the most accurate production of colors
possible, not the best image possible -- this is why we still have
humans running all of the print machines, to double check what's coming
out and to make 'subjective' adjustments to make a nice print into a
stunning print.
What does this mean?!?
=No matter how much you spend= on a color m/c/c/ setup, your
'average' images will still print 'dull and flat', your 'poor' images
will print 'poorly', and rarely will your 'great' images print
'stunningly' well.
( You can run right over to any press shop today and ask them, can
you print my book of 'stunningly' perfect images I've taken w/o doing a
test run, trial print, or tweak? They'll all laugh at you when they
ship you some poor looking books!
Every press on the planet has to adjust and tweak even though
they're all color m/c/c to produce the best looking prints! )
---
Of course, you don't want the monitor to change colors from one day to
the next -- makes it very difficult to know what colors look good, so
that's why you'd want to lock it down to an sRGB 5000k/5500k state, and
just don't touch it (here, a LCD panel will do this 100% better than a CRT).
======================
That said, is the Colorvision Spyder useful? I though so, bought one,
used it on my PC, and took it right off after weeks of use and never
looked back.
Why?
My LCD panel, like most sRGB gamut LCDs, was already 90%+ there matched
to the sRGB gamut. What little difference I saw on screen was like the
choice between cool white and warm white - something I can easily do
with my head, and a good test target on screen. Otherwise, what I saw
was basically just like I saw before, only a touch warmer overall.
(Don't even need to use Adobe Gamma here - a very subjective and 'bad'
tool IMO on a sRGB LCD monitor that's already accurate enough.)
Did the colorvision make better prints? No! I still had to tweak each
image for the best 'subjective' print - it didn't make better prints all
by itself.
I also have Epson everything - Epson 1200/925 printer in sRGB mode,
Epson 1200S scanner in sRGB mode, and the whole input to output pipeline
simply gave me pretty good base colors from one end to the other w/o
having to touch anything.
That said, I can firmly state today that these silly color m/c/c devices
perhaps are of use to a CRT monitor user (where most CRTs vary greatly
and aren't locked to sRGB gamuts), and to professionals wanting exacting
color matching, but for the most part, most consumers today wanting a
great print will 1) still have to hand-tweak 2) can use a sRGB workflow
to produce generally accurate colors 3) not have to spend $$$ on these
color m/c/c/ tools at all.
You can also easily reduce the cycle of tweak & print down to one or two
prints by simply having a standard test target visible on screen as a
reference to adjust your image to in a locked down sRGB workflow.
Whew!
=)
ps. that said, I've happily wasted 6+ months researching this whole
silly color m/c/c/ thing, spent $$$ more than I should, and found that
I'm back to where I began - hand-tweaking images for the best prints,
but now, in only one or two tweak & print cycles - and thrown the
Colorvision Spyder and Adobe Gamma out the door for a locked down, free
sRGB workflow.
honestly! Never once produced a better looking print with those color
m/c/c/ tools vs. those made without. As good, yes. Better, never.
'solution', only to realize what I found.
=======================================================
Subject: Matching print colors to Monitor screen
Ive acquired an Epson C82 which I plan to use primarily for printing
of Duotone images and some color. When I print in color, the color of
the print varies considerably from the monitor image. I've run Adobe
Gamma, as well as being careful to match the paper choice in the
driver with the paper I'm using, but not much help. After printing,
I can match the print fairly close to the monitor by adjusting it
with "Color Balance", cyan slider to -10, magenta and yellow each to -
50. The same problem and "correction" is evident with duotone
printing. Using the uncorrected monitor image, I can get the print I
want by adjusting those colors in the printer driver just prior to
printing. Is there a more "global" solution to this problem, rather
than my tedious procedure? I'm aware of the ColorVision/Spyder, etc.
approach, but that's a little tough on my budget. I'd appreciate any
help. Thanks
==========================
1) Realize that the color gamut of the monitor will =never= match the
color gamut of the printer!!!
Sorry!
This is the most crucial mistake most people make when they start
looking into color matching, and most don't realize until after spending
$$$ that the two will never match. You can stop now if you don't
understand this point.
The reason for this is that the two technologies produce mostly the
same colors, but at the outer edge of their color gamuts, they differ.
2) Realize that viewing the prints under various lights of different
color temperatures will make them look different!!!
This is another big mistake people make. If you print everything as
good as can be, but view the prints under, let's say, standard indoor
nightime tungsten lighting (soft white bulbs), you'll see something
very, very different than if you view the exact same print outdoors in
daytime, under a fluorescent bulb, etc.
The only way to ensure you're getting the 'best' color matching is
to view the prints under a standard color temperature light source (in
most cases, 5000k/5500k are common, graphics arts viewing temperatures).
You will need to get a light bulb that'll produce this light
temperature, install it, and only use this light source to view your
prints. (yes, there's what they call viewing booths used in print
presses and graphic arts that are super-duper calibrated, stable and
accurate, but well outside your budget. Besides, only the light really
needs to be fixed to a standard color temperature - so swapping out
light bulbs is the cheapest way to get almost there.)
Here, it's easiest to buy a fluorescent bulb from bulbs.com and
install it into a desktop fluorescent lamp.
Remember, == you must == turn off all other light sources (block off
daylight if you're viewing prints in daytime) and only use this standard
color temperature light source to view prints for the best evaluation of
prints.
3) Keep in mind that print colors will look different under different
light sources!!! Just because the prints look stunning under that 5000k
light source doesn't mean they'll look as good under tungsten lighting!!!
That said, you have to make a choice here - do I make prints that
look great under a standard color light temperature (just about all of
the books, magazines, prints published today are done so) but not
perfect under other lights? Or, let's say I only view and hang my
prints to view indoors under tungsten lighting, do I make them look
great under the tungsten lighting, but terrible under a standard color
light temperature or even outdoors in daylight?
This is a ====huge==== point that people fail to realize!
Just because you can print something that's 'correct and accurate'
according to 'standard practices and lighting temperatures' doesn't mean
you can't throw that out the door and print something that looks great
=under your specific display lighting conditions=!
Almost always, people will try and try and try to print an image so
that it looks perfect under 5000k lamps, then stick it on their hallway
walls that are only lit indoors by tungsten. Well, then, if you know a
bit about colors, you know that those pictures will have a strong yellow
cast over them due to the tungsten lights (you'll see this in 35mm
photographs at night indoors w/o flash; your eyes automatically adjust
after a few minutes to correct for the excessive yellow and make it seem
white). Does this make for a perfect print?!? No! Even after your
eyes have adjusted for the tungsten yellow cast, the prints still won't
look as good vs. under 5000k lamps because the print colors aren't
optimized for that lighting.
Here, you'll have to hand-tweak prints to get the very best colors
as nobody's done any real work on making prints that look great under
these conditions.
4) A locked down sRGB workflow is one easy step to perfection, and a
good one to try first.
If you have a CRT monitor, you may just have to use Adobe Gamma (if
you're using Photoshop, etc.) or a color calibration tool just to get
the monitor to look right - CRTs are terrible at giving you accurate
colors right out of the box.
Let's assume you have a LCD monitor - you'll likely be able to get
90%+ match to sRGB easily. Why? Most LCD gamuts practically sit on top
of the sRGB gamut - if you get an Apple with their Apple LCD Displays,
the color match is almost perfect.
Most of these monitors can be set to sRGB mode. If that's not
present, usually you can set it to factory default, then press an AUTO
button to get it mostly there matching wise. Look for color temperature
and sRGB mode...it helps to buy a LCD panel that already has these.
(That said, my Winbook 15" was that easy - press factory default,
press auto, and I'm 95% matched to sRGB right away according to my
ColorVision Spyder.)
* Next, you'll want to set your color workspace (if you're using a
program that handles this - most windows programs simply assume sRGB;
Adobe products should be set for sRGB), and your printer mode to sRGB
(eg. most Epson printers have this option; other printers, well, suck
because they often don't have this option), and your scanner as well
(eg. Epson, you can set in advanced mode the scanning target color space
to sRGB).
There, now that everything's locked to sRGB, what you see on screen
will generally match what you scan in, and what you print.
PLEASE!!! Keep in mind still that what you see =will never
perfectly match= what you print (point #1 above). So even now that
everything's working just peachy, prints still won't look just 100% like
the screen. Sorry!
5) So what to do? Take a test target image like the Photodisc Test
Target (free, search for it), open it up on screen, print a copy in your
locked down sRGB workflow, and compare the print to monitor -- this is,
for the most part w/o tweaking, about as good as you'll get even after
some heavy duty color calibration/management tools.
Now, what you want on print will have to be adjusted for on screen
based on what you see, experience, and what you've seen printed. But,
because you have this standard test target on screen and on paper, you
will know that when your own picture looks good vs. the test target,
it'll look good in print - you will use the test target as a standard
reference point for rapid image adjustments and tweaks ot make it look
right.
That's right! If you adjust =your= images while viewing a =standard=
test target image that prints great, you can make great prints faster
because you have something to adjust your own images to! If you only
tweak an image by itself, you'll soon find yourself wondering is that
blue really this blue or so so blue? It's because when your eyes look
at certain colors too long, they lose any objectivity and reference as
to what's neutral, good, and correct -- and it's very difficult to
adjust correctly without some reference image (like driving at night
without any points of reference - just pure darkness).
6) You'll quickly find that with the above, you can easily reduce the
process down to one or at least two tweak then print steps before
achieveing a perfect print by using an onscreen reference image and a
locked down sRGB workflow with a little practice and experience.
What you'll do is simply adjust your image to look perfect next to
that test target, print and move on to the next image.
==================================
That said, what about the color management/calibration/correction tools
that cost a bundle ala ColorVision Spyder, MonacoEZ, etc?
They do have their place -- usually, the reason is for consistency --
ie. no matter which monitor you drop in, or printer, or scanner, they'll
all give you as close of a representation of color to what can be
expected of them.
It =does not= mean they won't look different after being
managed/calibrated/corrected!!!! Only that they'll produce as true of a
representation of color as possible.
Why is this useful? If you're matching your red to CocaCola's special
red, you'd naturally want the most exact match possible. But when
printing images, that's totally different - you usually don' t care if
that color or this is perfectly matched, but rather, only that they look
stunning in combination with the rest of the picture. Here, you're an
artist, and you can choose to make the colors a bit more dramatic to
achieve the effect you want in a final print. Yes, if you're doing
catalogs for clothing, you'd want the colors of that dress to match as
close as possible to the real thing - good for the $$$ tools, but most
of the time, you can do without these tools in a sRGB workflow.
---
Q: But isn't color m/c/c supposed to give you stunning prints?!?
A: NO!!!!
Color m/c/c/ only gives you the most accurate production of colors
possible, not the best image possible -- this is why we still have
humans running all of the print machines, to double check what's coming
out and to make 'subjective' adjustments to make a nice print into a
stunning print.
What does this mean?!?
=No matter how much you spend= on a color m/c/c/ setup, your
'average' images will still print 'dull and flat', your 'poor' images
will print 'poorly', and rarely will your 'great' images print
'stunningly' well.
( You can run right over to any press shop today and ask them, can
you print my book of 'stunningly' perfect images I've taken w/o doing a
test run, trial print, or tweak? They'll all laugh at you when they
ship you some poor looking books!
Every press on the planet has to adjust and tweak even though
they're all color m/c/c to produce the best looking prints! )
---
Of course, you don't want the monitor to change colors from one day to
the next -- makes it very difficult to know what colors look good, so
that's why you'd want to lock it down to an sRGB 5000k/5500k state, and
just don't touch it (here, a LCD panel will do this 100% better than a CRT).
======================
That said, is the Colorvision Spyder useful? I though so, bought one,
used it on my PC, and took it right off after weeks of use and never
looked back.
Why?
My LCD panel, like most sRGB gamut LCDs, was already 90%+ there matched
to the sRGB gamut. What little difference I saw on screen was like the
choice between cool white and warm white - something I can easily do
with my head, and a good test target on screen. Otherwise, what I saw
was basically just like I saw before, only a touch warmer overall.
(Don't even need to use Adobe Gamma here - a very subjective and 'bad'
tool IMO on a sRGB LCD monitor that's already accurate enough.)
Did the colorvision make better prints? No! I still had to tweak each
image for the best 'subjective' print - it didn't make better prints all
by itself.
I also have Epson everything - Epson 1200/925 printer in sRGB mode,
Epson 1200S scanner in sRGB mode, and the whole input to output pipeline
simply gave me pretty good base colors from one end to the other w/o
having to touch anything.
That said, I can firmly state today that these silly color m/c/c devices
perhaps are of use to a CRT monitor user (where most CRTs vary greatly
and aren't locked to sRGB gamuts), and to professionals wanting exacting
color matching, but for the most part, most consumers today wanting a
great print will 1) still have to hand-tweak 2) can use a sRGB workflow
to produce generally accurate colors 3) not have to spend $$$ on these
color m/c/c/ tools at all.
You can also easily reduce the cycle of tweak & print down to one or two
prints by simply having a standard test target visible on screen as a
reference to adjust your image to in a locked down sRGB workflow.
Whew!
=)
ps. that said, I've happily wasted 6+ months researching this whole
silly color m/c/c/ thing, spent $$$ more than I should, and found that
I'm back to where I began - hand-tweaking images for the best prints,
but now, in only one or two tweak & print cycles - and thrown the
Colorvision Spyder and Adobe Gamma out the door for a locked down, free
sRGB workflow.
honestly! Never once produced a better looking print with those color
m/c/c/ tools vs. those made without. As good, yes. Better, never.
