The visual evaluation is based on a test image which is displayed in fullscreen on the monitor. I recommend to use the KowoDisplayTest program to create such an image. In Visual qualification of monitor using test images the test image is described.
The different letters show the different visual evaluations.
a) Groups of lines can be seen in the center and in the four corners; the recognizability of the individual lines in the high-contrast group and the low-contrast group are noted.

The result is evaluated in two notations for the best resolution of the 100% contrast line groups and the best contrast recognizability at a line width of 2 pixels.

For example, the result 1P-4G indicates that the lines with a spacing of 1 pixel are recognizable and that the four gray levels of the 2-pixel-wide lines can be resolved.


b) Below the 0% box is a box with horizontal and vertical lines 1 pixel width that are 5% brighter than the base brightness. The visibility of these lines is noted.
c) Below the 100% box is a box with horizontal lines 1 pixel width that are 5% darker than the base brightness. The visibility of these lines is noted.
d) To the right of the box with the dark horizontal lines is a box whose center is 5% brighter; the visibility of this box is noted.
e) To the left of the box with the light horizontal lines is a box whose center is 5% darker; note the visibility of this box.
f) Above and below the percentage brightness boxes are high-contrast fields. If these produce artifacts (e.g., fringes), note them.
g) At the bottom center of the image are two gray scale wedges that differ in gradation by a factor of 2. Using the switch (Alt – S), select the gradation at which no gradations are visible in the upper wedge and they already appear in the lower wedge. The number of distinguishable steps is noted.
h) From the three fields with numbers in black, gray, and white blocks, note the highest numbers that can be seen in the three fields.

You should use a table like this to note the results and count the points:


If an individual result deviates by more than 2 levels from the reference in (a) or (h), or if there is a deterioration in any of the results from (b) to (g), error correction measures must be initiated.
In addition, it must be checked whether the result (the number of points) still meets the required monitor class in the fourth column.

I think I should give some examples. The first evaluation was done with a cheap office monitor.

Sharpness and contrast are poor; only lines 2 pixels wide with 8 gray scale contrast could be resolved horizontally; vertical resolution is slightly better, but the poorer of the two values should be taken for the result.
All the line pairs with 5% contrast in area (b) and (c) could not be resolved. The box on the black ground (d) was not resolved, only on white ground (e) the 5% contrast was visible.
There were some artifacts at hard edges (gray shades) (f).
Only 32 steps from the gray value stepwedge could be resolved (g).
Of the numbers in the black area (hB), only the “2” can be recognized, in the gray area (hG) at least the ‘5’ and in the white area the “3”.

The sum of the point is 19 only. This monitor could only be used for image display (MK4) and not for image evaluation which would require in minimum 32 points. Okay - it is an office monitor and adequate for word processing.

Here is a second example - the monitor in front of me. It is a Samsung U32D970 with hardware calibration. When I used the monitor for the first time for image evaluation I did the test with the test image and filled the column "Reference". The result was 78 points out of a possible 81. In the last evaluation the monitor still archived very good results with 77 points.
(a) The monitor has perfect sharpness and contrast in the center and lower right corner; in the other corners I saw somehow a little bit less - this may be a feature of my eyes .
(b) to (g) achieve a perfect score.
(h) the only deviation of the test after 3 years of operation was in the dark area with the numbers - 3 years ago I saw the "7", now only the "6".

Finally the result is very close to the reference - the hardware calibration does a good job!
For the best monitor class MK1 in minimum 65 points are required.
I may continue image evaluation with my monitor.