How would this interesting monitor be after calibration?

[bikelike]

Iiyama is using Imac screens to make an affordable glossy 5K monitor, the ProLite XB2779QQS. I love the glossy Imac screen so I am higly interested. But I do not quite know what to expect from it. sRGB is not available...

In a review I read: "The panel does have one weakness, and that’s colour accuracy. With an average πDelta E of 3.34, and a maximum of 6.22, this isn’t a monitor you can rely on for colour-critical photo or video workflows."

So what does this mean? Is it referring to an incapability of the hardware and / or firmware of displaying certain color nuances, which cannot be solved by calibration? Or would all be fine after a Spider 5 job?

[digitaldog]

http://digitaldog.net/files/Delta-E%20and%20Color%20Accuracy%20Video.mp4

[bikelike]

http://digitaldog.net/files/Delta-E%20and%20Color%20Accuracy%20Video.mp4

I saw that video, but it does not bring me much, because of I have no idea how calibration will affect the accuracy.

My guess is hat right now it is a wide gamut device and that calibration will lower the gamut and the deltaE will drop, but really, I don't know. Can a calibration be done towards sRGB?

To be clear, I have never used calibration software.

[daicehawk]

1. It is not a wide gamut display. The review says 99 % sRGB.
2. It is not mentioned under which conditions the color accuracy is measured. IME, most modern IPS monitors can be calibrated to a sufficient color accuracy.

[degrub]

I saw that video, but it does not bring me much, because of I have no idea how calibration will affect the accuracy.

My guess is hat right now it is a wide gamut device and that calibration will lower the gamut and the deltaE will drop, but really, I don't know. Can a calibration be done towards sRGB?

To be clear, I have never used calibration software.

But calibration is not really about accuracy i think. It is about matching something - points in a standard colour space, prints on a printer, etc.. After all, we deal mostly in a reflective real world and not a transmissive world. So a screen is very different from what we see around us.   
Isn't it mostly about getting the screen to match the print device output ?

[daicehawk]

But calibration is not really about accuracy i think. It is about matching something - points in a standard colour space, prints on a printer, etc.. After all, we deal mostly in a reflective real world and not a transmissive world. So a screen is very different from what we see around us.   
Isn't it mostly about getting the screen to match the print device output ?

Please do not think typing-wise without a clear understanding of and better yet some experience with the entity in question. It is dangerous in disseminating incorrect information.
Monitor calibration is just an adjusting a black to white axis (grey one) to the given targets:
1. White point - defines the color and brightness of the white (255 in every channel r/g/b/ of the 8bit range)
2. Consistency of that white point color from white to black
3. The gamma curve - the uniformity of the brightness increment from black to white, subjectively judged by the place of the middle grey along the grey axis. In more consumer terms, contrast curve.
4. In some cases, the black point, both in terms of hue and brightness. In relation to white point brightness, it defines contrast.
You can and should adjust the white point and gamma curve by means of the OSD menu controls as close to the target values as possible before calibrating the display with a spectrophotometer\colorimeter. The calibration by using these devices will "equalize" each channel as to be in accordance to the target white point and gamma curve. Thus a big shift from actual white point and gamma curve from the target ones will need a bigger correction coefficient and lead to a deficiency of bit resolution resulting in a non-linear behavior (read: bad gradients) of any non-grey (read: hue having) color of the display, that will lead to color inaccuracy.
 

[degrub]

So what did i type incorrectly ?
You provided the nuts and bolts of the concepts that i expressed.
The OP indicated that he was not familiar with what could be accomplished by calibrating the monitor.

[Doug Gray]

But calibration is not really about accuracy i think. It is about matching something - points in a standard colour space, prints on a printer, etc.. After all, we deal mostly in a reflective real world and not a transmissive world. So a screen is very different from what we see around us.   
Isn't it mostly about getting the screen to match the print device output ?

The term calibration, refers to putting something (a printer, monitor, etc) in a known state. Profiling, which maps points to/from a calibrated device to a defined colorspace, is done after calibration. When calibration is well defined, such as setting CYMK densities on a printer that allows it, the profile doesn't need to be redone as things age. One just re-calibrates. Generally, cheap monitors can't be calibrated, better ones are done automatically by the profiling software so you effectively have both done.  Some early monitors would specify setting the contrast/brightness in certain ways before profiling and that is kind of a calibration and you would have to reset it if you changed them later to use the profile correctly.  Some monitors offer more extensive calibration such as daicehawk outlines. Most of these will also internally adjust the monitor so that it has a well defined colorspace, gamma, whitepoint, and luminance too. If all of these are set inside the monitor by profiling software the monitor's calibration is to the requested colorspace and the computer video driver just outputs RGB values without change. Applications then used the icc profile that matches what the monitor was profiled to. That's created by the profiling software although icc profiles like sRGB can be used if the monitor was profiled to match the icc sRGB. This can be useful if you profile a monitor on one system and move it somewhere else using a standard icc profile.

[digitaldog]

But calibration is not really about accuracy i think. It is about matching something - points in a standard colour space, prints on a printer, etc.

And if it's far from that 'standard' aim point, how can accuracy not be part of the consideration? And why would the OP's question about software and other's report a deltaE metric? So yeah, accuracy is absolutely part of the consideration in calibration and profiling.

[degrub]

Sure, accuracy in the matching of the color space representation, i agree. What i was referring to, and not clearly, was "accuracy" relative to the real world we see versus what is represented on the screen and print.

[JRSmit]

Sure, accuracy in the matching of the color space representation, i agree. What i was referring to, and not clearly, was "accuracy" relative to the real world we see versus what is represented on the screen and print.

Accuracy implies a reproducable measurable condition. Very difficult tot measure real world, as we cannot measure how we see. We can only simulate to a certain degree. Personally i believe only a limited degree. Yet we have the ICC results of close to a 100 years ago. So calibrating and profiling does bring the monitor closer to reality.

[digitaldog]

Sure, accuracy in the matching of the color space representation, i agree. What i was referring to, and not clearly, was "accuracy" relative to the real world we see versus what is represented on the screen and print.

The accuracy discussed is (should be) the dE between reference values and resulting values. So yeah, those high dE values are not good.

[bikelike]

The accuracy discussed is (should be) the dE between reference values and resulting values. So yeah, those high dE values are not good.

My question is: is it likely that those high values will decrease after calibration? Or are those gaps just there, whatever calibration / profiling will be applied?

[digitaldog]

My question is: is it likely that those high values will decrease after calibration? Or are those gaps just there, whatever calibration / profiling will be applied?

You have to calibrate to something (and measure) to even get such a report. Again (and I provided a URL video that I was told was seen and understood), one has to produce a set of measurements that are compared to a reference to even produce a dE metric. The URL defines what colorimetric accuracy is, how that metric is created.