Lux-meter, candelas or lux, measuring nits in practice

[flying_bear]

TL;DR

My lux-meter can be used in two modes, candela and lux. How do I use it to accurately measure a display's luminance in nits?

My lux-meter can be used in two modes, candela and lux. How do I use it to accurately measure a display's luminance in nits?​

The screen isn't Lambertian, I have to assume this as an unknown. I'm interested in the luminance as seen by the eye looking at the screen perpendicular to it.

Now, first some definitions, so we talk about the same thing and we avoid lectures:
LUMINOUS FLUX: Measured in lumens.
ILLUMINANCE: Lux is measured in luminous flux/area, thus lumen/m^2.
LUMINOUS INTENSITY: Candela. Measured in luminous flux per solid angle (think of it as a cone), lumens/sr.
LUMINANCE: Measured in luminous flux per solid angle per area, candelas/m^2. Nit: Another term for luminance, 1 nit=1 candela/m^2.

My lux-meter is only a sensor that measures luminous flux per solid angle (candelas), and can also show lux, which is simply candelas/distance^2. This is why the distance has to be entered into the lux-meter. It's nothing magical, just a light sensor with certain properties.

Placing the sensor onto the screens in question, I get candela values that are pretty close to the display's rated nits. But luminance does not equal luminous intensity.

I find online conversion calculators that strangely just equates illuminance with luminance. They're not the same! It's like saying that kilos potatoes planted per square meter equals kilos potatoes shot from a tater in a certain solid angle (cone) per square meter. It would vary with the solid angle used! It's potatoes per angle^2 per meter^2. One cannot simply just ignore the angle^2! Or can one?

My question is simply: HOW DO I USE MY SIMPLE LUX-METER TO MEASURE MY DISPLAY'S NITS RATING? I have scoured YouTube and the web for this, have strangely found no clear and practical instructions on how to proceed.
The closest I have found is this:
... all that is involved to find stimulus luminance is measuring the illuminance produced by the stimulus at the location of the eye and dividing by the solid visual angle of the stimulus. However, two other matters need to be considered.
(1) Lacking a lens, an illuminance meter is not directionally selective of the light to which it responds. Therefore, one must ensure that only light from the stimulus reaches the detector.
(2) Calculating the solid angle of a stimulus is easy enough if it is round. If it is not round, use a slightly smaller circular aperture and place it directly in front of the stimulus. Then calculate this aperture’s solid angle. A little work and some math can save thousands of dollars.

So, would the procedure be to

1. Cover the screen with a circular aperture, radius r
2. Place the sensor at least 5*r away (so we assume the convex surface as flat), distance d
3. Calculate the solid angle sr as d/r^2.
4. Make sure there are no other light sources, point the lux-meter towards the center of the aperture at distance d
5. Read the lux value lx (having entered the distance into the lux-meter, or just using candelas?)
6. Luminance in nits=lx/r^2?

 

[trurle]

Most of lux meters have a milky-color hemispherical caps to make them insensitive to direction of arrival of light. Ideally, such caps have a 2pi steardian FOV, but cheap meters may have a reduced FOV.
Candela reading of meter is generated just by dividing lux reading of meter by its FOV.
As long as your light source (monitor) is fully in the FOV of meter, will be no loss of accuracy by using either lux or candela readout.

 

[WernerQH]

LUMINOUS INTENSITY: Candela. Measured in luminous flux per solid angle (think of it as a cone), lumens/sr.
LUMINANCE: Measured in luminous flux per solid angle per area, candelas/m^2. Nit: Another term for luminance, 1 nit=1 candela/m^2.

Read the lux value lx (having entered the distance into the lux-meter, or just using candelas?)

My understanding is that the candelas are a property of the source, and should be independent of the distance $d$ (as long as you remain in the same solid angle, as seen from the source). I think that what your device really measures is lux, and the candelas are computed from that as $\text{lx} × d^2$. At twice the distance you should find 1/4 of the illuminance (lux).

Luminance in nits=lx/r^2?

I think this should be $\text{candelas} / (\pi r^2)$.

EDIT: I would dispense with the circular aperture, and use $\text{candelas} / (A \cos \theta)$ to get the average nits of the screen of size $A$, assuming a big enough distance and possibly seeing it inclined at an angle $\theta$.