Here are my random thoughts and what I have learned so far (mainly through reading Bruce Fraser, Fred Bunting and Chris Murphy’s book on “Real World Color Management” and numerous web resources and discussions with friends. Insights are welcomed and appreciated.
Color confidence and color reproducibility across different devices depends on a combination of (1) using an accurate and reliable calibrator/profiler and (2) the capability and limitations of the monitor itself (e.g., is it calibratable and to what extent – RGB voltages, contrast, luminance; What is the gamut size/color space, LCD versus CRT, and how the monitor is ‘calibrated’ – LUT only, in hardware). But realizing this is not enough to solve my current color issues — the same image still looks different depending on the monitor that I am using – even if they are calibrated and profiled with the same device and in the same environment (light conditions). I do have LCD monitors (MacBook Pro with a LED backlight display – GeForce graphics card and a Power Mac G4 with a Cinema HD monitor – ATI graphic card).
For example, consider a photo of my daughter playing in our backyard where the backdrop is green foliage (see here). In one display the foliage is displayed as lime green, vivid and almost a nuclear green, while in another the background is rendered in a calm pleasing avocado green. Skin tones in one monitor look red and flushed in one but in another I see a yellow cast (around her right ear area). Lips look magenta in one and a rusty red in another. Brightness varies as well (e.g., blown highlights in hair in one display but not in another). This image is on flickr and has been converted to the sRGB color space I am viewing my images via color managed – Safari or Ominweb browsers (I have also tried firefox -unmanaged – browsers).
I am curious where my problem is, and how I can solve it. Is the problem my calibrator and profiler system or is it the limitation of my monitors – or is it, can it be, the catchall: an operator problem?
How do I know which monitor shows correct color? Is it possible to know?
If I had a reliable calibrator/profiler system enabling transformation to and from a device independent system such CIE XYZ or LAB will the image look similar across different devices even if one monitor is not as capable? How do I know which monitor is closer to showing accurate color, how do I know if the calibrator is faulty?
I have a Spyder2Pro (a hardware calibrator and profiler system) and it is supposedly accurate and reliable. I thought it would solve the color inconsistencies I have been seeing across my platforms both at home and work but I am still stuck with color inconsistencies.
One thing I have discovered – for sure- is that people see the web with many different colored spectacles – and there is no way they see what you see ‘unless’ you are using the ’same’ reference system. You can, presumably, control your own environment, even if your calibrator/profiler uses the ‘wrong’ reference system – it then ensures, I suppose, that the monitor outputs are consistent among themselves (albeit consistently wrong). But I am not seeing that either – my output differ. So I cannot make any conclusions: – wether my monitors are faulty or not as capable? Or is my calibration system not reliable?
So in conclusion: I think it is a combination of the capability of the monitor, its gamut, and the calibration system and right now I don’t trust my calibration system… help – I am getting deeper and deeper into the infinite depth of color spaces and I can’t get out.
HELP!
