Confused about monitor calibration and color consistency

[Jerry van Dijk]

I recently commented on this excellent Landscape image by Rudi van den Heever that it was too saturated to my taste. I was quickly corrected by Ron Bernstein and Morkel Erasmus who saw nothing wrong with the saturation on their calibrated (and high end) monitors. I recently switched to a new laptop with wide gamut IPS screen and also noticed that my own older images which, like Rudi's image, had the sRGB color space embedded, looked much more saturated also. This got me confused about the color management on my new computer.
I'm working on the HP Elitebook 8760w with the wide gamut Dreamcolor screen, with a slightly wider native gamut than Adobe RGB. It has an onboard HP color management system, with a number of preset and factory calibrated color spaces for the screen, including Adobe RGB and sRGB (see screenshots below). I viewed Rudi's (sRGB) image with the screen set to Adobe RGB colorspace, which showed me an oversaturated image. When I switched the monitor preset to sRGB colorspace, the image looked fine (as it does on my other, standard screens).

However, from what I understand from color management in the digital workflow (a.o. from Roger Clark's website), when the color profile is embedded with the image, the software used to view the image allows for color management and monitors are correctly calibrated, everyone should roughly see the same colors on the image. But that does not seem to be the case here, which further confused me.
Even more confusing is that the internal color management seems fine. A test image saved in Adobe RGB colorspace and the same image converted to sRGB in CS5 look quite consistent in different viewers (Windows viewer, CS5, IE9, Safari) and even on different platforms (my laptop and iPad4), although you the difference between the aRGB and sRGB is more pronounced on the latter. Also, this consistency exists in both the Adobe RGB preset and the sRGB preset on the Elitbook, although in the sRGB setting, both test images have much less saturated colors. I haven't had the chance to check a print (I'm waiting for new cartridges).

This leads me to a number of questions:
- is it normal that an sRGB image looks oversaturated on a wide gamut monitor?
- could it be that everthing is fine, except for the factory calibration of the Adobe RGB color space by HP so I should just recalibrate my screen with a third party calibration unit?
- is there another possibility or flaw in my settings that I'm overlooking because of my limited (basic) understanding of color management?

Any help is greatly appreciated, I'm currently a bit reluctant to do any further image editing (or commenting on images) using the new screen before this issue is resolved !

[Tom Graham]

I agree with all you said.
Firstly, do all web/BPN comparisons in sRGB color space. And here is how I "correct" my postings in processing to look like how they will appear on BPN.
For BPN here is what I do. Download/save an image ( lot of color and dynamic ranges). View this image in a viewer (your choice, even PS if you with). Does that image in you viewer look exactly the same as when you bring it up on BPN? Compare side-side on monitor.
In other words, we will use BPN as the standard for postings on BPN.
Adjust/tweak, color-contrast-gamma, on the VIEWER (PS or other) to look like what it looks like when viewed directly from BPN.
You now have a viewer that is calibrated with BPN. But for prints or whatever else, you'll likely use different color settings.

Tom

[Jerry van Dijk]

Thanks for your response Tom! The problem is that in both the Adobe RGB and the sRGB setting, the image downloaded from BPN and opened in PS or other color calibrated viewer is consistent with how I view it in my webbrowser. Only in Adobe RGB both are oversaturated and in sRGB mode both are not. So the viewer calibration is not the problem, but the monitor calibration is.
I think it should be possible to have the "print workflow" as the central workflow, while converting images edited for print to sRGB colorspace for web presentation, while the reverse (viewing sRGB webbased images in an Adobe RGB based workspace) should also be possible without sudden changes in saturation and color representation. But I'm happy to stand corrected if this is just digital workflow utopia .

[Don Lacy]

Hi Jerry, from your description of your computer the problem seems to me to be the factory calibration of the screen is off compared to what others with calibrated system are seeing ( On my calibrated iMac the image you are referring to also looks fine ) I would recalibrate it using a good third party calibration system. I will also say that while lap tops have come a long way there screens are still not good enough for critical image processing or viewing IMHO.

[Linwood Ferguson]

I am very curious about this as well. I have two NEC monitors, a LCD2490WUXi (sRGB) and a PA271W (wide) that I just added. I noticed one day that the same face was terribly different on them. This is a cell phone image of the two side by side (wide is on the left, the right is in portrait mode). Both are carefully calibrated to the same white point, gamma and brightness with the same puck and software (both use hardware LUTS and Spectraview).

I found with trial and error that this problem goes away if I set the wide gamut monitor to sRGB emulation. Then they match. IT is as though the software + monitor + profile isn't smart enough to display properly.

Note in this case I actually split the window across the screens, but it does the same thing if there are two separate windows, each wholly in one monitor or the other.

It's more than just not matching -- the one on the left is just plain wrong, if I printed it, it would not be over saturated like that. I've printed FAR too many off the right hand monitor and know they match nicely.

So I don't have an answer for you but there are at least two of us confused.

[Tom Graham]

LF says -
"Note in this case I actually split the window across the screens, but it does the same thing if there are two separate windows, each wholly in one monitor or the other."
Your images are screen shots, yes? How did you capture the screen shots?
Both images you show are from the same camera jpg, right? This comparison would, to me, be better understood if you presented exactly the same image on both screens.
Or, are you saying that if you have two windows on the same monitor, each window shows such color differences for the same image???
Or am I getting windows and monitors confused???
Tom

[Linwood Ferguson]

LF says -
"Note in this case I actually split the window across the screens, but it does the same thing if there are two separate windows, each wholly in one monitor or the other."
Your images are screen shots, yes? How did you capture the screen shots?
Both images you show are from the same camera jpg, right? This comparison would, to me, be better understood if you presented exactly the same image on both screens.
Or, are you saying that if you have two windows on the same monitor, each window shows such color differences for the same image???
Or am I getting windows and monitors confused???
Tom

The actual photo is of two monitors sitting side by side almost touching. I then took a windows-type window displaying the photo, and dragged that window so it crossed between the two photos. What you are seeing is one image, split between the monitors. The man's face is quite literally the same image in the same windows window, cut in half where the monitors meet.

What I've also done is take the same image and open it up in two separate windows (i.e. two separate instances of the display program), and have one sitting on monitor #1, and one on monitor #2 (this to be sure I am not somehow confusing the color management of the application by having it split between monitors). Same type of result.

Everything is consistent - when the wide gamut monitor is configured wide, the images look oversaturated when displayed in supposedly color managed applications. What I did not do is take an inventory and try a bunch of applications, which I should probably do at some point, it may be program specific.

Since 98% of my work I do for the web I'm happy with it in sRGB emulation, but one day I need to figure this out better, as it might be desirable for some printing to have the wider space.

[Jerry van Dijk]

Hi Linwood, seems you have the same problem as I do. Still haven't found a solution, but didn't have the chance to run a screen calibration either. Will post any progress in the matter.

[Don Lacy]

I think you guys are confusing color space with calibration. You do not Calibrate a monitor to a color space you calibrate a monitor to properly reproduce colors to a set standard. Think of it this way if your monitor was a box of crayons when its calibrated your red crayons would be the right shade of red, now your color space correlates to the amount of crayons in your box a small color space like sRGB would be equal to a 12 pack of crayons while ProRGB would be the big 48 color box. All monitors need to be calibrated no matter what color space it is set to. Now to confuse things just slightly A wide gamut monitor in a wide gamut color space might display a color differently then a monitor set to a smaller gamut color space if that color falls out side the gamut of the smaller space. In Linwood's case one of his monitors might not be properly calibrated or the wide gamut monitor might be displaying colors outside the other monitors color space.

[Jerry van Dijk]

Thanks for your response Don! What I do not understand (given that the monitors are correctly calibrated in their different color space settings) is why an image that was created in sRGB colorspace should change in appearance when displayed in a wider colorspace. To stay in your analogy: the extra crayons from the wider gamut colorbox were not used to create the sRGB image, so enabling the display of these extra colors should not change the appearance of the sRGB image, unless the sRGB image is somehow converted or "stretched" to the wider colorspace (i.e. by interpolating the missing colors).

[Don Lacy]

Thanks for your response Don! What I do not understand (given that the monitors are correctly calibrated in their different color space settings) is why an image that was created in sRGB colorspace should change in appearance when displayed in a wider colorspace. To stay in your analogy: the extra crayons from the wider gamut colorbox were not used to create the sRGB image, so enabling the display of these extra colors should not change the appearance of the sRGB image, unless the sRGB image is somehow converted or "stretched" to the wider colorspace (i.e. by interpolating the missing colors).

Thats exactly what happens and the method used for the conversion depending on the colors also plays a role in how the image would be rendered when converted to a different color space both relative and perceptual handle out of gamut colors differently during a conversion. Here is a good tutorial on color spaces and conversions from one space to another read all three sections http://www.cambridgeincolour.com/tut...anagement1.htm