Color Problem

| So you are working with an uncalibrated monitor? 'Default' is not the
| same as calibrated, and has nothing to do with having a calibrated
display.
|

Yes. It's uncalibrated. I adjust it until it
looks right.

That's an interesting issue. For printing the
monitor needs to be coordinated with the
printer, so that the printer prints what you see.
For screen there isn't a realistic a standard.
If I create an image that's, say, H0378F1 that's
a delicious sky blue. I can paint it in a graphic
editor or spec it in a webpage. But I can't control
how other people see it. Other people with other
monitors will see if differently. Most people will
see differences depending on their viewing angle.
Many people will see the color washed out because
their monitor is set at high brightness. Apple
monitors may dither the color by a few points
away from what Windows full 24-bit displays
show. So what is the true color H0378F1? In
practice it's not an absolute color.

On Apr 2, 2016, Mayayana wrote
(in article ):


So you are working with an uncalibrated monitor? 'Default' is not the
same as calibrated, and has nothing to do with having a calibrated
display.

Yes. It's uncalibrated. I adjust it until it
looks right.

....but you don’t have a calibration eye-ball.

That's an interesting issue. For printing the
monitor needs to be coordinated with the
printer, so that the printer prints what you see.

Not quite. You need to start with a calibrated display, then you need to use
printer/paper specific icc profiles. Do that and your prints should be a
faithful representation of your intentions.

For screen there isn't a realistic a standard.

Hence the required calibration.

If I create an image that's, say, H0378F1 that's
a delicious sky blue. I can paint it in a graphic
editor or spec it in a webpage. But I can't control
how other people see it. Other people with other
monitors will see if differently. Most people will
see differences depending on their viewing angle.

Most people don’t bother to calibrate properly.

Many people will see the color washed out because
their monitor is set at high brightness. Apple
monitors may dither the color by a few points
away from what Windows full 24-bit displays
show. So what is the true color H0378F1? In
practice it's not an absolute color.

However, Apple displays can be calibrated.

--

Regards,
Savageduck

On Sat, 2 Apr 2016 09:19:56 -0400, Alan Browne
wrote:

On 2016-04-01 18:34, Eric Stevens wrote:
I am tempted to recover some ancient photographs which I have scanned
from slides. I am having particular problems with
https://dl.dropboxusercontent.com/u/31088803/GS5.jpg which has an
excessive red tinge. I am particularly trying get rid of the red from
the label while retaining the underlyiing cream color.

I have tried all sorts of things in both Photoshop and Lightroom but
without success. I would appreciate any help in solving this problem.

Did you look up the colour components of the "cream" colour so you would
be forearmed with info about what to change - and what not to?

Various online "cream" definitions show significant red content - so
reducing the red may require adjustments to the G and B as well to
maintain balance.

I knew that and did that.

Offhand: Did you try setting the Image|mode to LAB and then
de-saturating the red from there?

I must say I never thought of that.

I was doing most of my work in Lightroom. I did think of using PS to
enable Savageduck's piecemeal approach but I was trying to avoid it.
Lab never occurred to me.
--

Regards,

Eric Stevens

On Sat, 02 Apr 2016 14:36:55 GMT, "MC" wrote:

Eric Stevens wrote:

I am tempted to recover some ancient photographs which I have scanned
from slides. I am having particular problems with
https://dl.dropboxusercontent.com/u/31088803/GS5.jpg which has an
excessive red tinge. I am particularly trying get rid of the red from
the label while retaining the underlyiing cream color.

I have tried all sorts of things in both Photoshop and Lightroom but
without success. I would appreciate any help in solving this problem.

15 minutes using "Replace Color" and "Hue/Saturation"

https://www.dropbox.com/s/npmvlx3a9m6stf5/GS5.jpg?dl=0

However, not knowing the true underlying colours it is based on
guesswork. Also, a bit more time and effort will make the things look
more refined. My post is just to show you that it does not take much
to correct colors in PS.

That's very good.

Unfortunately, as I have already explained, ever since I got CC I have
been concentrating on processing several thousand photographs of
various types in LR and my PS experience has fallen far behind. I can
see I must get up to speed.
--

Regards,

Eric Stevens

|You need to start with a calibrated display, then you need to use
| printer/paper specific icc profiles. Do that and your prints should be a
| faithful representation of your intentions.

Personally I don't print. I work with digital.
If I did print, I'd be concerned with getting
the printer to print what I see on the monitor.
If I need to get color x in photo z by shifting
R by, say, 8 points and you do it by shifting
R by 7 points, so what? Your method is right
because you calibrated your monitor? If it
prints right then the color is right. Most people
don't even have color perception sufficiently
subtle to deal with such things. Whether that's
physical or psychological is hard to say, but few
people can do what an interior designer does
in discerning subtle color qualities.

So this is all very relative. Also, people who
don't work with the technical end like to think
digital can somehow be made static, but it can't.
Your strong feelings about EXIF data are similar.
As is the attitude in the corporate world that
DOC files are official. They're not. They're highly
mutable.
A JPG is just a map of image pixels. If you edit
it it's no longer the same map. EXIF data can
be useful for your records, but it's also just
mutable, digital bits. It doesn't authenticate
your image.

Anyone who's worked with web design knows
the whole thing is maddeningly relative.
Originally there were "web safe" colors that
everyone decided would be as standardized
as possible. I think it was any color with 00
33 66 99 CC FF, and maybe a few with 80.
So you could use #669966 for your webpage
background and feel *a little bit* confident
that most people were seeing something in
the ballpark of what you saw. If you specced
#669960 you'd be less certain.

In printing there are Pantone colors, which
can be verified by the human eye. (Even though
that's also very relative. Even if one has perfect
vision, cream next to yellow looks white, while
cream next to white looks yellow.)
There are even alleged Pantone equivalents to
computer color display standards. For example,
Pantone 2925 is #008ED6. So I suppose you
could buy a Pantone chart and hold those tabs up
to your monitor while you display a webpage set
with a background of #008ED6, or whatever other
color you want to check. Even then, the matte of
the paper and reflectiveness of the screen will
probably show as 2 different colors. And even if
you get them all matched, prepare an image and
put that image on your website, your viewers will
have Mac or PC; Samsung, Acer, or AOC; ATI or
NVidia. I find that colors look very different on
the exact same hardware when I install Linux on
a Windows computer. Linux seems to have more
sophisticated graphic drivers. The colors look richer.
Then there are different user settings. Different
room lighting. Matte or gloss screens. There's the
background you display your image on. It will look
very different against red than it will against blue.
(Thus we have colored matte board for framing.)
Finally, many of your viewers actually do not have
the same color vision that you do. I used to have
a friend who took up oil painting. He was surprisingly
talented. But.... "Hey, David, what's with the purple
color pears in that still life?" It turned out David was
color blind. He'd never known it! He agreed with
other people about what green or red was. He just
wasn't seeing the same thing. What exactly does
he see? I have no idea.
Even the physical function
is relative. If you stare at a red flag and then look
at a white wall you'll see green. Yet the green light
waves needed to create that perception do not
exist. It's a trick caused by the eye's structure.

All of which is to say that while I don't think it
hurts to calibrate, the main thing with print is:
Do your photos print the way you expect? And
the main thing with digital is far more amorphous.
Does your website seem to look good on the
monitors where you can view it? Are you comfortable
using your monitor with the current brightness
setting? Then it's probably OK. But if you ask
someone, "What did you think of that incredible
bittersweet hue in that photo I posted?" you
shouldn't be surprised if they answer, "What? You
mean the cranberry?"

Then the question is, what's different? Their
eyesight? Their level of articulation in perceiving
color? Their monitor? OS? Graphic card? Color
profile?

I find that most flatscreen TVs are far too saturated.
I'm guessing that's probably because people are
impressed with that. They think a richer blue is a
better blue. A richer yellow is a better yellow. So
the TV companies crank up the color and it needs
to be fixed in order to display properly. But most
people don't make that adjustment. They don't know
how and/or they like the rich colors. (Then we get
into issues of personality types, among other things.)
So, then you get famous and go on TV to show your
wonderful photo with the incredible bittersweet hue.
And the next day people stop you on the street:
"Wow" Such beautiful photos! How did you ever capture
that incredible magenta hue?"

In article , Mayayana
wrote:

|You need to start with a calibrated display, then you need to use
| printer/paper specific icc profiles. Do that and your prints should be a
| faithful representation of your intentions.

Personally I don't print.

that much is clear.

I work with digital.

so what? digital can be printed or not printed. analog can be printed
or not printed.

one has nothing to do with the other. they are two independent
statements.

If I did print, I'd be concerned with getting
the printer to print what I see on the monitor.

that's what a colour managed workflow does.

If I need to get color x in photo z by shifting
R by, say, 8 points and you do it by shifting
R by 7 points, so what? Your method is right
because you calibrated your monitor? If it
prints right then the color is right.

trial and error is an ineffective, inefficient and inaccurate way to do
something.

Most people
don't even have color perception sufficiently
subtle to deal with such things. Whether that's
physical or psychological is hard to say, but few
people can do what an interior designer does
in discerning subtle color qualities.

that's why people use a hardware device to properly calibrate their
display and printer.

So this is all very relative. Also, people who
don't work with the technical end like to think
digital can somehow be made static, but it can't.
Your strong feelings about EXIF data are similar.
As is the attitude in the corporate world that
DOC files are official. They're not. They're highly
mutable.
A JPG is just a map of image pixels. If you edit
it it's no longer the same map. EXIF data can
be useful for your records, but it's also just
mutable, digital bits. It doesn't authenticate
your image.

not relevant to colour management, and they *can* be made static anyway.

Anyone who's worked with web design knows
the whole thing is maddeningly relative.
Originally there were "web safe" colors that
everyone decided would be as standardized
as possible. I think it was any color with 00
33 66 99 CC FF, and maybe a few with 80.
So you could use #669966 for your webpage
background and feel *a little bit* confident
that most people were seeing something in
the ballpark of what you saw. If you specced
#669960 you'd be less certain.

those days are gone.

In printing there are Pantone colors, which
can be verified by the human eye. (Even though
that's also very relative. Even if one has perfect
vision, cream next to yellow looks white, while
cream next to white looks yellow.)
There are even alleged Pantone equivalents to
computer color display standards. For example,
Pantone 2925 is #008ED6.

no it isn't.

pantone 2925 is a specific colour.
#008ED6 is a triplet that can be any colour, depending on the colour
space, the gamut of the display and its calibration, among other
factors.

So I suppose you
could buy a Pantone chart and hold those tabs up
to your monitor while you display a webpage set
with a background of #008ED6, or whatever other
color you want to check. Even then, the matte of
the paper and reflectiveness of the screen will
probably show as 2 different colors.

that's essentially what calibrating and profiling a monitor does,
except that it's a helluva lot faster and far more accurate to use a
hardware device than holding up a pantone chart and eyeballing every
patch.

And even if
you get them all matched, prepare an image and
put that image on your website, your viewers will
have Mac or PC; Samsung, Acer, or AOC; ATI or
NVidia. I find that colors look very different on
the exact same hardware when I install Linux on
a Windows computer.

that's because you don't have a colour managed workflow.

if you did, then it wouldn't matter what hardware or software you used.

read this book and learn something for a change:
http://colorremedies.com/realworldcolor/

Linux seems to have more
sophisticated graphic drivers.

definitely not. linux wishes it was sophisticated.

the state of the art software is on mac and windows. linux does fine
for headless tasks where there's little to no human interaction, such
as a server.

The colors look richer.

which typically means less accurate.

Then there are different user settings. Different
room lighting. Matte or gloss screens. There's the
background you display your image on. It will look
very different against red than it will against blue.
(Thus we have colored matte board for framing.)
Finally, many of your viewers actually do not have
the same color vision that you do. I used to have
a friend who took up oil painting. He was surprisingly
talented. But.... "Hey, David, what's with the purple
color pears in that still life?" It turned out David was
color blind. He'd never known it! He agreed with
other people about what green or red was. He just
wasn't seeing the same thing. What exactly does
he see? I have no idea.

colour blindness can be simulated for those who are not colour blind.

Even the physical function
is relative. If you stare at a red flag and then look
at a white wall you'll see green. Yet the green light
waves needed to create that perception do not
exist. It's a trick caused by the eye's structure.

yet another irrelevant issue.

On Apr 3, 2016, Mayayana wrote
(in article ):

You need to start with a calibrated display, then you need to use
printer/paper specific icc profiles. Do that and your prints should be a
faithful representation of your intentions.

OK! Due to your lengthy argument I am going to address a few things, but snip
to make it less of a scroll-fest. Anybody who wants the full context can
refer to your original response.

Personally I don't print. I work with digital.
If I did print, I'd be concerned with getting
the printer to print what I see on the monitor.

If you consider color management in anyway important then you have to start
with a calibrated display.

If I need to get color x in photo z by shifting
R by, say, 8 points and you do it by shifting
R by 7 points, so what? Your method is right
because you calibrated your monitor?

If you need to get color x, the only place that color is going to meet your
intentions is on your uncalibrated display, as far as anybody else goes it is
going to be dumb luck if they see anything close to your intended final
color. I have no control over anybody else’s display or eyeball.

If it
prints right then the color is right. Most people
don't even have color perception sufficiently
subtle to deal with such things. Whether that's
physical or psychological is hard to say, but few
people can do what an interior designer does
in discerning subtle color qualities.

Even with a print you have to start with a standard palette (appropriate
colorspace and calibrated display) and if that standard palette is
contaminated, your print process is going to be problematic. Also consider
that every other display digital copies of your images might be viewed on are
probably not going to be a representation of the color rendition you
intended.

The visual acuity or color perception of any viewer is something totally out
of your control, so again all you can or should deliver is an image which can
be viewed by the majority of your target viewership with colors consistent
with your intent, regardless of them seeing it on their display, or as a
framed print.

Le Snip

All of which is to say that while I don't think it
hurts to calibrate, the main thing with print is:
Do your photos print the way you expect? And
the main thing with digital is far more amorphous.
Does your website seem to look good on the
monitors where you can view it? Are you comfortable
using your monitor with the current brightness
setting? Then it's probably OK. But if you ask
someone, "What did you think of that incredible
bittersweet hue in that photo I posted?" you
shouldn't be surprised if they answer, "What? You
mean the cranberry?"

...and another Snip

So all I have to say is, for $100-$200 you can get the basic calibration
tools you need to deliver your images with your intended colors regardless of
how outlandish and garish they might be.

....and if you consider the investment each of us has made in equipment,
including computers and software, the cost of ensuring the final step is as
close to the intended goal is not a big one. So my recommendation is, step up
buy a Datacolor Spyder, or an X-Rite ColorMunki, calibrate your display
properly and you might be surprised at what you see right in front of you.
You have no guarantee of what you will get without some sort of consistent
color management.

--

Regards,
Savageduck

On Sun, 3 Apr 2016 10:36:42 -0400, "Mayayana"
wrote:

|You need to start with a calibrated display, then you need to use
| printer/paper specific icc profiles. Do that and your prints should be a
| faithful representation of your intentions.

Personally I don't print. I work with digital.
If I did print, I'd be concerned with getting
the printer to print what I see on the monitor.
If I need to get color x in photo z by shifting
R by, say, 8 points and you do it by shifting
R by 7 points, so what? Your method is right
because you calibrated your monitor? If it
prints right then the color is right. Most people
don't even have color perception sufficiently
subtle to deal with such things. Whether that's
physical or psychological is hard to say, but few
people can do what an interior designer does
in discerning subtle color qualities.

So this is all very relative. Also, people who
don't work with the technical end like to think
digital can somehow be made static, but it can't.
Your strong feelings about EXIF data are similar.
As is the attitude in the corporate world that
DOC files are official. They're not. They're highly
mutable.
A JPG is just a map of image pixels. If you edit
it it's no longer the same map. EXIF data can
be useful for your records, but it's also just
mutable, digital bits. It doesn't authenticate
your image.

Anyone who's worked with web design knows
the whole thing is maddeningly relative.
Originally there were "web safe" colors that
everyone decided would be as standardized
as possible. I think it was any color with 00
33 66 99 CC FF, and maybe a few with 80.
So you could use #669966 for your webpage
background and feel *a little bit* confident
that most people were seeing something in
the ballpark of what you saw. If you specced
#669960 you'd be less certain.

In printing there are Pantone colors, which
can be verified by the human eye. (Even though
that's also very relative. Even if one has perfect
vision, cream next to yellow looks white, while
cream next to white looks yellow.)
There are even alleged Pantone equivalents to
computer color display standards. For example,
Pantone 2925 is #008ED6. So I suppose you
could buy a Pantone chart and hold those tabs up
to your monitor while you display a webpage set
with a background of #008ED6, or whatever other
color you want to check. Even then, the matte of
the paper and reflectiveness of the screen will
probably show as 2 different colors. And even if
you get them all matched, prepare an image and
put that image on your website, your viewers will
have Mac or PC; Samsung, Acer, or AOC; ATI or
NVidia. I find that colors look very different on
the exact same hardware when I install Linux on
a Windows computer. Linux seems to have more
sophisticated graphic drivers. The colors look richer.
Then there are different user settings. Different
room lighting. Matte or gloss screens. There's the
background you display your image on. It will look
very different against red than it will against blue.
(Thus we have colored matte board for framing.)
Finally, many of your viewers actually do not have
the same color vision that you do. I used to have
a friend who took up oil painting. He was surprisingly
talented. But.... "Hey, David, what's with the purple
color pears in that still life?" It turned out David was
color blind. He'd never known it! He agreed with
other people about what green or red was. He just
wasn't seeing the same thing. What exactly does
he see? I have no idea.
Even the physical function
is relative. If you stare at a red flag and then look
at a white wall you'll see green. Yet the green light
waves needed to create that perception do not
exist. It's a trick caused by the eye's structure.

All of which is to say that while I don't think it
hurts to calibrate, the main thing with print is:
Do your photos print the way you expect? And
the main thing with digital is far more amorphous.
Does your website seem to look good on the
monitors where you can view it? Are you comfortable
using your monitor with the current brightness
setting? Then it's probably OK. But if you ask
someone, "What did you think of that incredible
bittersweet hue in that photo I posted?" you
shouldn't be surprised if they answer, "What? You
mean the cranberry?"

Then the question is, what's different? Their
eyesight? Their level of articulation in perceiving
color? Their monitor? OS? Graphic card? Color
profile?

I find that most flatscreen TVs are far too saturated.
I'm guessing that's probably because people are
impressed with that. They think a richer blue is a
better blue. A richer yellow is a better yellow. So
the TV companies crank up the color and it needs
to be fixed in order to display properly. But most
people don't make that adjustment. They don't know
how and/or they like the rich colors. (Then we get
into issues of personality types, among other things.)
So, then you get famous and go on TV to show your
wonderful photo with the incredible bittersweet hue.
And the next day people stop you on the street:
"Wow" Such beautiful photos! How did you ever capture
that incredible magenta hue?"

How do you know you don't have a color blind monitor?
--

Regards,

Eric Stevens

| Couldn't be bothered to read the whole thing

I suppose your comments might be more
relevant and more interesting if you knew
what you were responding to, no?

|
| A printer does not and cannot know how you view the colours on your
| monitor. It only knows profiles.

I never said anything about ignoring profiles.
What I'm saying is that if I can print what I
see then it doesn't much matter if my monitor
is officially shifted slightly one way or another.
And when it comes to digital online, which is
mainly what I deal with, the "truth" of color is
far more nebulous. In that case I'm designing
layouts and preparing images for other people to
see. They all see something slightly different.

You and I both did an edit of the photo. Did you
think yours was good and mine was poor? I thought
they were both good and very similar to each other.
Yours is a bit more blue-green. Mine is a bit more red.
Not because my monitor is batty. Because I liked
the look of mahogany on the steering wheel and
what I thought was a more realistic steel color.
(Your steering wheel looks more like walnut to me.)
Which is closer to "reality"? We don't know.

| You cannot calibrate by eye ans different people interpret colours
| differently.
|
That makes sense in a limited way, but it's like
saying your oven must be calibrated before
baking bread because people have different
taste preferences. Just because everyone loves
the bread that doesn't mean it was cooked
properly.

In any case, anyone who can't discern a slight
tinge toward R, G or B from the norm probably
shouldn't be working with color photos.

| How do you know you don't have a color blind monitor?

Colors that I know look right. Did you think
my edit of the photo was off? If not then what's
the problem? I mean... I'm not the one posting
pictures or garishly green steel.