Incident Light Meter

[markjwyatt]

I do use incident metering sometimes. Basically I point the dome towards the camera (or the primary light source in some cases) and use the reading directly (Gossen Luna Pro or Luna Pro SBC). For practical purposes, what follows does not really matter, but might be interesting to understand.

The light meter does not know when I slide the dome over over the meter lens. Assuming the subject is 18% grey (which many believe the meter does, some say 12%, etc.), then

1. Call the light incident on the subject is 100%
2. The light reflected from 18% grey back to the camera is 18%

So, I guess we must conclude that the dome is designed to cut the light reaching the meter lens by close to 2.5 stops more than the actual incident light, so we get the correct reading relative to 18% grey reflected? I pointed the Luna Pro at the screen I am writing on, and other areas, and the dome cuts closer to 1.5-3 eV. When I point into the room, where the light is more hemispherical it comes closer to the 1.5 eV, while pointing at the screen (more directional) it is closer to 2, and a distant light bulb more like 3.

Thoughts?

 

[grain elevator]

So, I guess we must conclude that the dome is designed to cut the light reaching the meter lens by close to 2.5 stops more than the actual incident light

I think this is in a way correct. Your error is in the empirical part, or the ideas behind it. Without the dome, you're not measuring "the actual incident light", hence the difference from your expectation. Without dome, it's a normal meter and you measure the average (or weighted average) brightness of an area, which happens to include your light source.
I the case of the computer screen, an incident metering just doesn't make sense, it's backlit. Of course you can use it as a light source (in which case you need to put the incident meter at the position where the subject will be, remember light falls off with distance), then the reflective metering simply doesn't make sense. Remember you're either measuring exposure for a picture of the screen, or for an object somewhere else, why would they be the same?
Maybe the two measurements of computer screen and incident light produced by it will coincide in the way you propose if you're in an otherwise dark room, the screen exactly fills the metering angle of the reflective measurement, and you have a (sort of pointless) incident device which limits the angle of incidence to the same as the reflective metering, at the same distance. That's a way to think about it - the incident dome has 180 degree angle of incidence and, due to its opacity, isn't irritated by point light sources. So it does a lot more than just dim the light down. You really can't compare the two in this way - compare them used correctly for the different measuring modes.

 

[ic-racer]

Reflected meter:

Aperture squared / Shutter Time = (Luminance * Film speed) / (K = about 12)

Incident meter (light at 40 degrees):

Aperture squared / Shutter Time = ( Illuminance * Film speed ) / ( C = about 300)

These constants are determined by the meter manufacture. For example Minolta Meter III has C=330 for the dome and C=250 for the disk.

 

[markjwyatt]

@grain elevator my first sentence kind of addresses the intent aspect. The rest of the OP wanders into design apspects of light meters. My only purpose in pointing the meter/dome at certain lighting conditions was strictly to determine the transmittance characteristics of the dome.

@ic-racer thanks, I think I have seen these equations before here on Photrio), and they get to the heart of the matter.

In the end, the meter does not know (at least Gossen Luna Pro models) what my intent is (i.e., measuring luminance or illuminance), and it does not know whether the dome (or disk) is over the sensor window or not. Those equations are enforced by the meter designer/manufacturer by the geometry (hemisphere, disk, possibly others) and optical material properties of the dome/disk. Really the geometry of the dome effectively converts from luminance (reflected light entering window) to illuminance (integrated light flux directed into window via the dome/disk geometry), aided possibly by the diffusiveness, then calibrated to 18% reflected by the total transmittance of the dome/disk. It seems a lot rests on that little plastic part!

 

[Danner]

The meter reads the absolute light level on the sensor, and outputs a corresponding signal. The 18% gray calculation comes later. Because 18% gray is not an absolute light level, it's a function of the scene illumination, and further manipulated with exposure information. The dome scatters light from a 180º hemisphere onto the sensor. The lens under the dome, with the dome out of the way, looks at about a 30º cone of light in front if it.