PowerBook G4 + Cinema Display Color Temperature

[lukas_jenkins]

I just received my 20" Apple Cinema Display this evening. I set it up and the first thing to

do was calibrate it with my Spider2 colorimeter made by Color Vision. I bought the screen

for future G5 use but for now I'll be using it with my PowerBook as a second display. I plan

to fill the 20" screen with the image I'm working on and have all my dialogs and toolbars

on the 17" laptop screen. The first thing I noticed was that I had to lower the Cinema

display to its lowest brightness to be similar to my PowerBook. Since it takes a relatively

long time to calibrate both displays I decided to post my question. How should one handle

matching the color temperature between the two monitors. I currently have them both set

at 6500k at gamma 2.2. Technically this is all fine and dandy but visually both screens do

not look the same as they should. The PowerBook screen is cooler. Would you suggest

setting both monitors to the native color temperature of the backlight or try to alter them

through icc profiling with my Spider2? I know the Cinema display is 6300k native but I am

unsure of the PowerBook.

 

Thanks,

 

Lukas

[andrew_rodney1]

Yup, try Native Gamma and of course, be aware that you'll have to compare the color

matching IN an ICC aware application like Photoshop (not something like the desktop). And

setting the luminance so the two match is pretty critical.

[lukas_jenkins]

The luminance is the closest when the PowerBook is at maximum brightness and the Cinema

display is at its minimum brightness. I am blown away by how bright the Cinema displays can

get if one were to crank them up to maximum brightness. So I'm going to try both displays

set to a gamma of 2.2 and the color temperature native. I'll report back.

 

Lukas

[andrew_rodney1]

-->The luminance is the closest when the PowerBook is at maximum brightness and the

Cinema display is at its minimum brightness.

 

That actually doesn't surprise me.

[serge c]

Ideally both should be calibrated to Native WP. However normally native white points will not match, especially when displays are that different.

 

I would suggest setting one of them to measured Native white point and then using the x and y CIE numbers (it's more accurate than the K number in this case) from the generated profile as a white point target for the second monitor (assuming you use Spyder2 Pro). You can see those numbers by running validation and then going to info (or printing it out for better illustration).

 

As a shortcut you can load (and then modify if you wish) a profile created for the first monitor (with a Native target) as a target for the second monitor (if it's Spyder2 Pro).

[serge c]

Colorvision software does not have a Native Gamma option, so 2.2 is good.

[serge c]

Now, adjusting white point on an LCD monitor to match the target other than Native negatively affects image quality (unless you have a DDC setup and a monitor with 12 or more bits LUTs - which you don't). So if native white points don't match decide which monitor quality you are prepared to compromise. Don't worry about it too much - usually it's relatively unnoticeable.

[lukas_jenkins]

Ok the monitors are now calibrated as follows:

 

1. Apple 20" Cinema display

- Gamma = 2.2

- Color temperature = native (6,300k I believe)

 

2. PowerBook display

- Gamma = 2.2

- Color temperature = native

 

I put the displays on mirroring which I have found to be the easiest way to compare the

two screens. The PowerBook screen is significantly warmer and this is very apparent side

by side. This really surprises me I must say. I wonder if the PowerBook's LCD backlight is

going bad or its native setting is just really warm. It is also possible that the Cinema

display is just that much better (which would explain why it costs $800). So far it does not

look as if the screens will be matched as far as color balance.

 

Lukas

[lukas_jenkins]

I think I'm just going to go with both monitors being profiles which is MUCH better than

nothing so I'm happy. I have Spider2, I bought it over a year ago and now many versions

exist. I have a choice between 1.8 and 2.2 gamma and 5000k, 6500k, and native for color

temperature. I guess a more advanced colorimeter would be beneficial in this case but I do

not see it worth the money. native is where I should be anyway because quality is

important. Just wanted to make sure I wasn't doing something wrong.

 

Thanks for your help,

 

Lukas

[serge c]

Nothing's wrong. Just Native white points are different. It's normal.

 

Spyder2 (at least the pro version, but perhaps other version as well if they let you specify custom targets) lets you match white points if you choose to do so. There's no advantage to using other packages in this case.

[serge c]

Come to think of it Pro is the only Spyder2 that lets you have custom targets... So in this case setting the same K target for both would be the only option to match them. Which is not ideal for LCDs, but possible.

[Uhooru]

Shouldn't they both use the same white point? How can you match color if you have different

settings for white. That doesn't make sense.

[andrew_rodney1]

Mirroring is going to mess it all up as BOTH signals go through a single profile and video

LUT.

[serge c]

"Shouldn't they both use the same white point? How can you match color if you have different settings for white."...

 

They "should" if your goal is matching colors. Only to reach the same white point target on LCDs you need to adjust videocard LUTs which leads to RGB values loss (unless they have 10 or mere bits internal LUTs and so on...). So you have to decide what your priorities are. In Likas's case he'd have to adjust both from their native white point (to 6500K for instance) because his software does not have custom white point targets.

 

Note that even if they have different white points they are still calibrated - their output is measured and recorded in the profile. Color managed applications see both profiles.

[serge c]

Sorry, probably not the "videocard" LUTs, but the effect is still the same - altering 8 bit per channel color.