Calibration software for an iMAC

wrote:
I have an iMAC about a year old. I am trying to calibrate the monitor
to assist in making photos. I have used the Spyder3 software and
colorimeter in the process. The problem is that the prints all come
out noticeably darker than the image on the screen. The Spyder people
say this is Apple's fault. The brightness control (really the
backlighting control) on the display will not sufficiently reduce the
brightness to get an accurate calibration. The only work around I see
is to reduce the brightness in Photoshop below what looked good on the
screen and hope the printer responds by yielding a print of the
correct brightness. But this adjustment is completely ad hoc and is
just the kind of fiddling I hoped to avoid using the calibration
routines.

Two questions: 1) Does anybody have a better way to use Spyder3 to
compensate for this problem? 2) If not, what alternative calibration
system might be used that will avoid (or at least mitigate) this
problem? I read good things about the Macbeth systems, but will they
run into the same problem?

Thanks for the help.


First ignore all the garbage in troll replies - the iMac 24 inch screen,
matt, if that's what you have, is one of the best for colour accuracy
and actually doesn't need calibrating at all. I have a Spyder 2 Pro, a
ColorMunki and a Huey Pro. Guess how I calibrat? I use Apple's excellent
visual utility in Advanced mode, making sure I do the following:

Set the monitor to minimum brightness (not excessive, and very
comfortable in my office with normal lighting, or with my window
shutters part open on a sunny day like this). Calibrate following
Apple's sequence, being sure to sit around two feet from the monitor,
and being sure to blur your eyes slightly and avoid using the central
foveal zone of the eye (do not look at or focus on the target when
adjusting - defocus by looking 'through' the target, and look an inch to
the side of the matching patch).

This will ensure that visual accommodation does not screw up your
ability to spot when the density/colour matches.

Finally, performing the profile creation Native Gamma and Native White
Point. Don't adjust these to 1.8 or 2.2, or 6500 or D65 or 5000 or
whatever. The monitor is already fairly close to a gamma between Mac
(1.8) and PC (2.2) and very closely to a D65 whitepoint. Never, ever use
a 5000/D50 white point as this is an old standard which will make the
screen look murky and yellow.

Save/apply the resulting profile. It will have fewer tone breaks or
unwanted colour shifts than a Spyder 3 profile, because the contrast
scale of the monitor and its WP have not been messed with. Calibrators
can never add accuracy - they can only approximate it by REDUCING the
range of the monitor; any change they make to levels in one channel or
brightness zone will be reflected in a loss of bit depth elsewhere in
the tone scale. The Apple manual/visual calibration method loses less
bit depth than most hardware calibrators.

After having done this, ensure your printer driver settings and paper
profiles all work correctly. I have a laser printer, the second I've
had, and no matter what I do - always darker, always poor colour. I also
have an Epson 3800 and the colour is a perfect match for the screen
without the slightest effort.

I do have a Colour Confidence GrafiLite, a small D65/6500 type viewing
light closely matched to the native colour of the iMac screen (and most
good LCD/TFT panels). This can help when fine tuning anything critical,
but most of the time, room or daylight is enough to confirm I've still
printing correctly.

See:

http://www.dphotoexpert.com/2008/01/...viewing-light/


David

David Kilpatrick wrote:
wrote:
I have an iMAC about a year old. I am trying to calibrate the monitor
to assist in making photos. I have used the Spyder3 software and
colorimeter in the process. The problem is that the prints all come
out noticeably darker than the image on the screen. The Spyder people
say this is Apple's fault. The brightness control (really the
backlighting control) on the display will not sufficiently reduce the
brightness to get an accurate calibration. The only work around I see
is to reduce the brightness in Photoshop below what looked good on the
screen and hope the printer responds by yielding a print of the
correct brightness. But this adjustment is completely ad hoc and is
just the kind of fiddling I hoped to avoid using the calibration
routines.

Two questions: 1) Does anybody have a better way to use Spyder3 to
compensate for this problem? 2) If not, what alternative calibration
system might be used that will avoid (or at least mitigate) this
problem? I read good things about the Macbeth systems, but will they
run into the same problem?

Thanks for the help.


First ignore all the garbage in troll replies - the iMac 24 inch screen,
matt, if that's what you have, is one of the best for colour accuracy
and actually doesn't need calibrating at all. I have a Spyder 2 Pro, a
ColorMunki and a Huey Pro. Guess how I calibrat? I use Apple's excellent
visual utility in Advanced mode, making sure I do the following:

Set the monitor to minimum brightness (not excessive, and very
comfortable in my office with normal lighting, or with my window
shutters part open on a sunny day like this). Calibrate following
Apple's sequence, being sure to sit around two feet from the monitor,
and being sure to blur your eyes slightly and avoid using the central
foveal zone of the eye (do not look at or focus on the target when
adjusting - defocus by looking 'through' the target, and look an inch to
the side of the matching patch).

This will ensure that visual accommodation does not screw up your
ability to spot when the density/colour matches.

Finally, performing the profile creation Native Gamma and Native White
Point. Don't adjust these to 1.8 or 2.2, or 6500 or D65 or 5000 or
whatever. The monitor is already fairly close to a gamma between Mac
(1.8) and PC (2.2) and very closely to a D65 whitepoint. Never, ever use
a 5000/D50 white point as this is an old standard which will make the
screen look murky and yellow.

Save/apply the resulting profile. It will have fewer tone breaks or
unwanted colour shifts than a Spyder 3 profile, because the contrast
scale of the monitor and its WP have not been messed with. Calibrators
can never add accuracy - they can only approximate it by REDUCING the
range of the monitor; any change they make to levels in one channel or
brightness zone will be reflected in a loss of bit depth elsewhere in
the tone scale. The Apple manual/visual calibration method loses less
bit depth than most hardware calibrators.

After having done this, ensure your printer driver settings and paper
profiles all work correctly. I have a laser printer, the second I've
had, and no matter what I do - always darker, always poor colour. I also
have an Epson 3800 and the colour is a perfect match for the screen
without the slightest effort.

"Perfect match"? Really David, I expect more accuracy from you. There
is no way a transparent (monitor) will match a reflective (paper).

Having said all the above. Unless one 'simulates' the paper while in
(eg) Photoshop, then what you see on the screen is never going to match
what prints. And it is that fateful step of simulating the paper as I
described in my other post that gets you there.

I too have an iMac 24 an Epson 3800, and as closely as I have calibrated
it (very minor change from the apple visual calibration), unless one
views the photoshop image with 'simulate paper' on, one will not have
anything close to the print.

Cheers,
Alan

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: Remove FreeLunch.
-- usenet posts from gmail.com and googlemail.com are filtered out.

In article , David Kilpatrick
wrote:

First ignore all the garbage in troll replies - the iMac 24 inch screen,
matt, if that's what you have, is one of the best for colour accuracy
and actually doesn't need calibrating at all. I have a Spyder 2 Pro, a
ColorMunki and a Huey Pro. Guess how I calibrat? I use Apple's excellent
visual utility in Advanced mode, making sure I do the following:

although it's an excellent screen, there are better, although much more
expensive. while the calibration is pretty good out of the box, it
helps to profile it (many people don't understand the difference).

Set the monitor to minimum brightness (not excessive, and very
comfortable in my office with normal lighting, or with my window
shutters part open on a sunny day like this). Calibrate following
Apple's sequence, being sure to sit around two feet from the monitor,
and being sure to blur your eyes slightly and avoid using the central
foveal zone of the eye (do not look at or focus on the target when
adjusting - defocus by looking 'through' the target, and look an inch to
the side of the matching patch).

if it doesn't need calibrating as you say, why are you calibrating it?

furthermore, calibrating by eye is at best, slightly better than doing
nothing at all, and often much worse. using a hardware calibration
puck is *much* better because it eliminates the huge variability of the
eye.

This will ensure that visual accommodation does not screw up your
ability to spot when the density/colour matches.

exactly why a hardware device is better.

Finally, performing the profile creation Native Gamma and Native White
Point. Don't adjust these to 1.8 or 2.2, or 6500 or D65 or 5000 or
whatever. The monitor is already fairly close to a gamma between Mac
(1.8) and PC (2.2) and very closely to a D65 whitepoint. Never, ever use
a 5000/D50 white point as this is an old standard which will make the
screen look murky and yellow.

Save/apply the resulting profile. It will have fewer tone breaks or
unwanted colour shifts than a Spyder 3 profile, because the contrast
scale of the monitor and its WP have not been messed with. Calibrators
can never add accuracy - they can only approximate it by REDUCING the
range of the monitor; any change they make to levels in one channel or
brightness zone will be reflected in a loss of bit depth elsewhere in
the tone scale. The Apple manual/visual calibration method loses less
bit depth than most hardware calibrators.

nonsense. a profile from a spyder or other device will be *much*
better than one done by eye.

Alan Browne wrote:

I too have an iMac 24 an Epson 3800, and as closely as I have calibrated
it (very minor change from the apple visual calibration), unless one
views the photoshop image with 'simulate paper' on, one will not have
anything close to the print.


Have you got a viewing light? If you use a viewing light which provides
about the same reflected brightness as the apparent brightness of the
monitor, the match can very close.

Of course, I am used to viewing proofs and prints, and monitors. I've
been doing it since the first colour monitors appeared (we had the first
Radius with calibration and also one of the very early Barcos which I
never liked much). I do not expect a 'match' in the unrealistic sense
some users may. Also, I'm not messing with out of gamut colour
adjustments and to some photographers my pix must look very plain.

It was maybe three years ago I did a talk for Hawick Camera Club, and at
the end of the evening, they asked for some quick help with printing
problems. They had a stack of very bad prints. It took me about twenty
minutes - not being familiar with Windows 2000 which I think was the
system - to find the Adobe Gamma utility, set Photoshop colour prefs
properly, load an image, implement and note down the correct driver
settings (for their future use) and get a print out which almost raised
a cheer - it matched the screen!

At that stage none of the users of their new digital photo system
understood how all the components in a CMM work, or how the printer
driver (it was an Epson, not sure which one) worked. I think everyone is
better informed now but I still deal with a lot of calls helping readers
get this stuff right.

http://www.pbase.com/davidkilpatrick/image/110547036

Here's a shot of the Grafilite with a print and a monitor image - in
order to get the scales right, I used a small old monitor, Apple 15
studio, which actually gives crazy shifts in contrast with vertical view
angle and of course the print is seen with sheen from the gloss in the
photo. In real life the 'lit' print can be a very neat match to the
screen placed near it, especially with my iMac 24 on minimum brightness
at native WP.

David

nospam wrote:

if it doesn't need calibrating as you say, why are you calibrating it?

To create a profile - as you suggest. In fact any of the visual
adjustments made on this screen are so small they may not be worth making.

There is a very small difference between the default profile and the one
created by the Apple utility, but there's a big difference if I use a
hardware calibrator - and generally, the results are worse.

It would be more accurate to describe my approach as profiling the
monitor, rather than calibrating it. Calibrating is what we used to do
with hardware adjustment of the guns in CRTs or loading new LUTs into
display cards.


exactly why a hardware device is better.

Far from true. Hardware devices vary greatly and so does the software.
I've tested dozens of them. Some are disatrous (the Pantone Huey is a
good example - next to useless, especially on most laptops for which it
is a natural companion).

nonsense. a profile from a spyder or other device will be *much*
better than one done by eye.

Most of the calibration problems I have had brought to me by
professionals have been caused by hardware devices.

When I use a hardware device (or test one) I will use the recommended
procedures which often involve changes in brightness, match to specific
gamma, match to WP. It is very common to find - especially with 6-bit
depth LCD screens and laptops generally - that these adjustments create
tone breaks including banding in sky gradations. Brightness may also be
reduced to the point that the display becomes unusable.

I'll also use my own standards, which involve working to a reasonable
brightness, using native WP and gamma. In theory, the hardware device
should then create a profile which corrects only for non-linearity of
the RGB curves of the monitor. Some are better than others. The extreme
20-minute profiling sequence of the highest end software for the Spyder
2 Pro (I had three or four different programs for it) never produced
results I liked. The simple Colormunki approach of reading just eight
levels per colour, very quickly, seems just as good in terms of final
profile.

Even so, I use the visual/manual profile creation method. It produces a
better profile. I get stacks of RGB black and white images from my
professional readers (mono accounts for 70 per cent of all UK social
photography at awards/exhibition level) and with my profiles, they all
look just as they should with no tint shifts at any density.

That is unless the photographer has used a hardware calibrator, ended up
with a tint shift in the shadows, then messed around in Photoshop trying
to remove it and ended up with a red or green bias showing up in
gradations.

A good test of whether your monitor is correctly calibrated is simply to
take any Photoshop RGB image, execute the Desaturate command, and view
it. If there is absolutely no hint of any tint shift at any density,
your monitor is fine and just needs a profile, not a calibration+profile.

David

David Kilpatrick wrote:
Alan Browne wrote:

I too have an iMac 24 an Epson 3800, and as closely as I have
calibrated it (very minor change from the apple visual calibration),
unless one views the photoshop image with 'simulate paper' on, one
will not have anything close to the print.


Have you got a viewing light? If you use a viewing light which provides
about the same reflected brightness as the apparent brightness of the
monitor, the match can very close.

No. After printing I typically walk them into my dining room as it has
a mix of south east light and incandescent lighting. But glad to see
you've changed to "match can be very close".

Of course, I am used to viewing proofs and prints, and monitors. I've
been doing it since the first colour monitors appeared (we had the first
Radius with calibration and also one of the very early Barcos which I
never liked much). I do not expect a 'match' in the unrealistic sense
some users may. Also, I'm not messing with out of gamut colour
adjustments and to some photographers my pix must look very plain.

Then you must have been aloof or lazy when you wrote: "the colour is a
perfect match for the screen without the slightest effort." as that is
simply not possible.

It was maybe three years ago I did a talk for Hawick Camera Club, and at
the end of the evening, they asked for some quick help with printing
problems. They had a stack of very bad prints. It took me about twenty
minutes - not being familiar with Windows 2000 which I think was the
system - to find the Adobe Gamma utility, set Photoshop colour prefs
properly, load an image, implement and note down the correct driver
settings (for their future use) and get a print out which almost raised
a cheer - it matched the screen!

At that stage none of the users of their new digital photo system
understood how all the components in a CMM work, or how the printer
driver (it was an Epson, not sure which one) worked. I think everyone is
better informed now but I still deal with a lot of calls helping readers
get this stuff right.

http://www.pbase.com/davidkilpatrick/image/110547036

Here's a shot of the Grafilite with a print and a monitor image - in
order to get the scales right, I used a small old monitor, Apple 15
studio, which actually gives crazy shifts in contrast with vertical view
angle and of course the print is seen with sheen from the gloss in the
photo. In real life the 'lit' print can be a very neat match to the
screen placed near it, especially with my iMac 24 on minimum brightness
at native WP.

Well, that image as you present it above is the least convincing ... It
took me several weeks to get the Epson 3800, my iMac and CS3 to play
together properly - which is not an exact match. And that is limited to
a few papers to date.

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: Remove FreeLunch.
-- usenet posts from gmail.com and googlemail.com are filtered out.

David Kilpatrick wrote:
nospam wrote:

if it doesn't need calibrating as you say, why are you calibrating it?

To create a profile - as you suggest. In fact any of the visual
adjustments made on this screen are so small they may not be worth making.

There is a very small difference between the default profile and the one
created by the Apple utility, but there's a big difference if I use a
hardware calibrator - and generally, the results are worse.

It would be more accurate to describe my approach as profiling the
monitor, rather than calibrating it. Calibrating is what we used to do
with hardware adjustment of the guns in CRTs or loading new LUTs into
display cards.


exactly why a hardware device is better.

Far from true. Hardware devices vary greatly and so does the software.
I've tested dozens of them. Some are disatrous (the Pantone Huey is a
good example - next to useless, especially on most laptops for which it
is a natural companion).

nonsense. a profile from a spyder or other device will be *much*
better than one done by eye.

Most of the calibration problems I have had brought to me by
professionals have been caused by hardware devices.

When I use a hardware device (or test one) I will use the recommended
procedures which often involve changes in brightness, match to specific
gamma, match to WP. It is very common to find - especially with 6-bit
depth LCD screens and laptops generally - that these adjustments create
tone breaks including banding in sky gradations. Brightness may also be
reduced to the point that the display becomes unusable.

I'll also use my own standards, which involve working to a reasonable
brightness, using native WP and gamma. In theory, the hardware device
should then create a profile which corrects only for non-linearity of
the RGB curves of the monitor. Some are better than others. The extreme
20-minute profiling sequence of the highest end software for the Spyder
2 Pro (I had three or four different programs for it) never produced
results I liked. The simple Colormunki approach of reading just eight
levels per colour, very quickly, seems just as good in terms of final
profile.

Even so, I use the visual/manual profile creation method. It produces a
better profile. I get stacks of RGB black and white images from my
professional readers (mono accounts for 70 per cent of all UK social
photography at awards/exhibition level) and with my profiles, they all
look just as they should with no tint shifts at any density.

That is unless the photographer has used a hardware calibrator, ended up
with a tint shift in the shadows, then messed around in Photoshop trying
to remove it and ended up with a red or green bias showing up in
gradations.

A good test of whether your monitor is correctly calibrated is simply to
take any Photoshop RGB image, execute the Desaturate command, and view
it. If there is absolutely no hint of any tint shift at any density,
your monitor is fine and just needs a profile, not a calibration+profile.

Thank you, David. This is one of the most illuminating (pun intended)
threads in quite a while.
I've yet to re-profile my iMac, as I've seen no problems so far with
prints on my 3800. Previous Macs' monitors have been done with the Spyder.

--
John McWilliams

In article , David Kilpatrick
wrote:

if it doesn't need calibrating as you say, why are you calibrating it?

To create a profile - as you suggest. In fact any of the visual
adjustments made on this screen are so small they may not be worth making.

true, the differences from out of the box and a hardware profile are
small.

There is a very small difference between the default profile and the one
created by the Apple utility, but there's a big difference if I use a
hardware calibrator - and generally, the results are worse.

then you're doing something wrong. for me the results are dramatically
better with a hardware device.

exactly why a hardware device is better.

Far from true.

it's exactly true.

Hardware devices vary greatly and so does the software.

while there is some variance among the different hardware devices, they
are *far* more consistent than the human eye, which varies much more
than even the worst hardware device, unless of course, there's a defect
in the puck. for example, some of the early spyders had an issue that
produced very odd results, and they would replace it with one that
didn't have the problem.

David Kilpatrick wrote:
wrote:
I have an iMAC about a year old. I am trying to calibrate the monitor
to assist in making photos. I have used the Spyder3 software and
colorimeter in the process. The problem is that the prints all come
out noticeably darker than the image on the screen. The Spyder people
say this is Apple's fault. The brightness control (really the
backlighting control) on the display will not sufficiently reduce the
brightness to get an accurate calibration. The only work around I see
is to reduce the brightness in Photoshop below what looked good on the
screen and hope the printer responds by yielding a print of the
correct brightness. But this adjustment is completely ad hoc and is
just the kind of fiddling I hoped to avoid using the calibration
routines.

Two questions: 1) Does anybody have a better way to use Spyder3 to
compensate for this problem? 2) If not, what alternative calibration
system might be used that will avoid (or at least mitigate) this
problem? I read good things about the Macbeth systems, but will they
run into the same problem?

Thanks for the help.


First ignore all the garbage in troll replies - the iMac 24 inch screen,
matt, if that's what you have, is one of the best for colour accuracy
and actually doesn't need calibrating at all. I have a Spyder 2 Pro, a
ColorMunki and a Huey Pro. Guess how I calibrat? I use Apple's excellent
visual utility in Advanced mode, making sure I do the following:

The 24" Imac screen isn't "one of the best for color accuracy". It's no
better - and probably worse - than other currently available IPS panel
screens of that size.
The panel is made by LG China (formerly LG Philips) presumably made to
Apple's specs - including price. The panel itself is probably fine.
I understand that the Imac 24" backlight is an array of 3 x CCFL tubes.
That's noteworthy because while (cheap) TN panel displays of that size
might have 3 CCFL tubes, IPS panels displays usually have 6.
Here's an article which gives a very plausible explanation for the
excessive brightness:
"Apple tries to compensate for this with ridiculously high level of
brightness"
http://www.silvermac.com/2008/24-inc...n-do-about-it/
Note the comment: "Even the lowest setting is still way way too bright
for most of us". I understand that lowest setting gives around 180cd/m2.

Next step up in apple hardware is probably the the 24" LED backlit ACD.
Reviews state that white level can be set to appropriate levels (for
"most of us" - I assume), and would hope that LED backlighting gave less
non-uniformity, but AFAIK it's a white LED (as used in laptops - not the
vastly better RGB LED backlit display used in professional grade LCD
monitors) display, has lousy connectivity options (not a problem perhaps
if you have the right mac), and lousy ergonomic adjustments (tilt only -
no 180 deg swivel, and no height adjustment) and it's expensive for what
you get.
Please write this off as a troll post if the truth hurts.

nospam wrote:


while there is some variance among the different hardware devices, they
are *far* more consistent than the human eye, which varies much more
than even the worst hardware device, unless of course, there's a defect
in the puck. for example, some of the early spyders had an issue that
produced very odd results, and they would replace it with one that
didn't have the problem.


I have 98/100 Farnsworth-Munsell test vision and Shirley has 100/100
(plus the disturbing ability to remember colours - that is, to memorise
a colour and match any time later) which I guess is one of the things
which got her ahead in her Masters degree in colour science.

So I may be as accurate as some calibrators. I've used many calibration
systems because we are sent them to test. We did have (and still have,
but unused) a neutral grey room and I'll admit I no longer use that for
retouching (it is now my studio instead), but I have solid wooden
shutters on the windows so my environment can be totally controlled.

Putting it simply, if you obtain three calibrators and three packages to
run them, then calibrate your monitor with all three, they will not
produce identical profiles. I've tested this many times. Since any three
or more hardware calibration devices will never produce identical
results, and some are way different, I conclude that unless you have
limited colour vision and really can't do the visual thing, they may not
always be worth the money.

Of course, if your colour vision is in the lower quartiles (which is the
case for many men, especially those over 30) you won't be able to see
the differences after calibration for exactly the same reasons that you
couldn't use the visual calibration.

What's your Farnsworth-Munsell 100 score? It would be interesting to
know, because this test is something we recommend as a preliminary for
studios when selecting staff to do retouching/pp work. Most universities
with a colour science dept will test you.

Most colour management I see are not fine-tuning issues, they are
massive errors - prints far too dark compared with the screen, etc. That
is rarely a calibration issue, it is nearly always a printer profile and
driver settings issue.

David