REVIEW: Dell U2410 Part 8
sRGB and AdobeRGB are covered completely or almost completely. This also applies to the ISO Coated colour space. In colour management applications, this provides the basis for very precise with both working colour spaces as source colour spaces and for soft proofing based on ISO Coated.
At 88 percent, the coverage of ECI-RGB is lower, but this is not surprising. Thus, slight colour gaps and somewhat larger colour deviations can be expected here, depending on the image content. Nonetheless, the user can absolutely work with ECI-RGB as a working colour space.
We got a nasty surprise during the calibration and profiling of the Dell U2410 from the "Custom Colour" mode. This mode is important because the RGB Gain regulators are available, allowing the white point to be adjusted using on-board means. In light of the powerful electronics (12-Bit LUT incl. FRC implementation on the panel) on the Dell U2410, this is very useful for avoiding stronger losses of hues.
Unfortunately, we could not achieve an even satisfactory result, neither on a Revision A00 model nor on a Revision A01. The profile validation revealed considerable deviations, i.e. the current screen status was not captured correctly in the profile.
In order to examine the problem in more detail, we adapted the actual XYZ standard colour values of the primary colours with Bradford as per D50 and compared them with the standard colour values captured in the profile:
* based on D50
The deviation in green is glaring. The reason for the "incorrect" capture of the primary colours becomes clear when one looks closely. The actual luminance values for the primary colours (0.2769 + 0.4607 + 0.0566) do not amount to 1. Thus, here, there is a clear lack of linearity.
If we are using the chromaticy coordinates (x,y,z; however, only the XYZ tristimulus values are displayed in the table) of the primary colours and the white point for calculation and presume linearity, the resultant Luminance components match the values captured by iColor.
Measurement of the luminance components based on the standard colour value shares; the inverted values in the second vector are based on the standard colour value shares of the white point (here: D50).
Result of the calculation. Compared with the Y values under "Profile".
Thus, this mode is not usable for calibration or usage without calibration.
Explanation of deltaE deviation: We measure deviation in colour hues in DeltaE 94. This expands the original CIELAB formula from 1976 with correction colours, with which the colour saturation of the colour samples is considered. Here, we use the factors used in the graphics industry. For neutral grey colour samples, both formulae would deliver an identical result. The more saturated the colour samples are, the lower the colour status is in DeltaE 94 with regard to the first definition. However, the uniformity is not achieved even with the newer formula.
Therefore, a general recommendation for the recommended deviation range is difficult. At a DeltaE of 1, the user can generally not perceive any difference in colour for direct comparisons except with very lightly saturated colour samples. Beyond a DeltaE of 5, the deviation would generally be described as strong. We measured various hues (primary and secondary colours as well as some tertiary colours). Secondary and especially tertiary colours deliver good indicators with regard to the linearity of the screen. Compared to sRGB, some considerable deviations arise naturally as compared with the sRGB reference in monitors with extended colour spaces. On the other hand, a grey axis that is as neutral as possible in the factory settings is interesting.
For the white point, we indicate the DeltaE deviation from the target white point in the respective tables. When the monitor is uncalibrated, larger deviations can certainly arise. This is generally not problematic initially. The target value is purely a recommendation (and seldom suits precisely). However, the difference between the white point and the black body curve should already be as low as possible. This can be checked using the second value indicated, which should not exceed a DeltaE of 3. When the monitor is calibrated, both values should be considerably lower. Our target white points (D50 and D65) are only slightly off the black body curve.
Explanation of DeltaC deviation: We test the neutrality of the grey axis by measuring grey levels between 10 and 90 percent, which we consider with regard to the actual white point. The DeltaC indicates the difference in chroma and corresponds to the DeltaE without luminance share, with achromatic colours as a reference. In order to expose unpleasant fluctuations in the direction of various hues, we also measure the DeltaC range. It reflects the extent of the largest interval vector between the colour samples on the a*-b* level.
A monitor should be as neutral as possible in the factory settings. At the maximum, a DeltaC of 4 and a range of 3 should never be exceeded, since otherwise, considerable tints arise. The perception threshold is considerably lower for achromatic colours, with a DeltaC of about 0,5. For a good result, one can use the UDACT (UGRA test) criteria as a guide; this requires an average maximum value of 1 and a maximum range of 2. This result should of course not be exceeded by a significant amount when the monitor is calibrated. With monitors that are very neutral in the factory settings, a slight disimprovement may arise after calibration.
In addition, we also prepared a graphical representation of the test model’s gradation. When the monitor is uncalibrated, a gamma of about 2,2 or an sRGB gradation, stable across the entire grey axis, is desirable. It is entered as a target status for the relevant diagrams. Thus, the correct display of sRGB content with regard to the luminance distribution is achieved even without calibration. At the same time, the corrections during calibration (for target values of sRGB or 2,2) are low. In this context, a gamma setting on the model is a clear advantage, especially when sRGB and L* can be selected as well as values in figures. When the monitor is calibrated, the monitor should achieve the characteristics established before the calibration.
No Comments available