Strange observation about Minolta meter.

[Chan Tran]

I have 3 Minolta meters and they exhibit the same thing which leads me to think that because of the sensor they use. The 3 meters are Autometer II, Flashmeter III and Flashmeter VI. The Flashmeter VI in spot mode and the other 2 in reflective mode. I measure my LCD monitor to make sure the white point luminance is correct (just a cross check against the I1 Displaly Pro puck). I set it for 120 cd/m^2 and the 3 meters measures the same which is EV9.7 @ISO100 which is 116cd/m^2 which is good. Turned the meter 90 degrees and it read 8.9 or 9 so it's about 66cd/m^2. Almost a full stop lower. I know the LCD has polarizer but do the meter sensor have polarizer too. You need 2 polarizers to have that effect right? It's not room right reflection because the room is dark.

 

[xkaes]

I'm not sure about your setup or what "Turned the meter 90 degrees" means, but the meters do not have PL filters. You can take whatever you have on the top of meters off, and you can see the cell.

Since the meters all read the same in one setting, it looks like they are A-OK.

What are you trying to accomplish?

 

[koraks]

I know the LCD has polarizer

That was my first thought too. I don't know the physical layout of the silicon cell, but I can see how it might act as a polarizer as well. Since you're losing 'only' one stop, if this is an interaction between the polarizer of the screen and a polarizing effect within the sensor, it's evidently not complete as you'd expect several stops of loss.
I've done some quick searching but it would take me some more time to figure out if there's a (partial) polarizing effect to some types of Si photo diodes. Again, I can sort of imagine how this could be the case as a result of the physical makeup of the photodiode and any protective glass on top of it.

You can take whatever you have on the top of meters off, and you can see the cell.

Keep in mind that the cell isn't bare; you're looking at a protective overcoat or layer, which can have a filter function as well. It depends on the specific type and exact specifications of such a top layer are also not always made explicit in the part's datasheet.

 

[Chan Tran]

The cell of the Autometer II and Flashmeter III look the same and the reflective attachment is only a mask with a hole in the middle so if there is a polarizer it's on the cell. Of course the Flashmeter VI in spot mode has optics to make is a spotmeter and I can't see the cell.
When I say turn 90 degree means normally you hold the meter vertically and whatever on the face of the meter is upright. Turning 90 degree mean turn the meter so that the face is sideway. I did this because I have 2 monitors one is set in landscape mode and the other in portrait mode. At first I measured the 2 displays and found they are different I thought the 2 monitors are not the same but turning the meters I found it's the orientation that makes the different. At first I use the spotmeter and I guess it's something in the optics that caused it so I used to other 2 meter and found it's the same.

 

[xkaes]

If that's a polarizer over the silicon cell, it's the thinnest polarizer I've ever seen. What could be the reason for putting a polarizer over a metering cell -- whether in a hand-held meter or in a camera?

 

[reddesert]

Polarization sensitivity can occur if there's something like a reflection or a half silvered mirror in the light path. I would not have expected that in an Autometer II / III, but I don't actually know the orientation of the sensor in the head of the meter. That's the reason linear polarizing filters can give odd results with cameras (like AF cameras and some others) depending on how the light is reflected to the meter cell.

If you want to test this further, try metering an unpolarized source (like a lit lampshade) through a linear polarizing filter, or a CPL filter turned backwards, and rotate the filter to see if you can measure the amount and direction of the orientation effect.

I didn't understand at first what you meant by rotating the meter 90 degrees, either, but I get it now, you mean keeping the meter input pointed at the source but rotating 90 deg around the line of sight (in aircraft terms, roll, but not yaw or pitch).

 

[koraks]

If that's a polarizer over the silicon cell, it's the thinnest polarizer I've ever seen.

1: A polarizing filter doesn't have to be thick. A polarizing pair of sunglasses or a camera filter are thick because of the glass (or plastic) support. The actual filter is a tiny fraction of a millimeter thin.
2: It's not necessarily a filter as such, but a polarizing effect that's inherent to the micro-geometry of the sensor.

 

[xkaes]

I didn't mean the physical thickness of the filter, I was referring to the transparency.

 

[Chan Tran]

I am sure the sensor has filter because it has to correct for the spectral response.

 

[koraks]

I was referring to the transparency.

The transparency of what?

I am sure the sensor has filter because it has to correct for the spectral response.

Quite so.

 

[xkaes]

I am sure the sensor has filter because it has to correct for the spectral response.

It certainly has a filter that adjusts the spectral response. All you have to do is look at it. It's yellow-green with minimal density like you would get from a polarizer.

If there were a polarizer over the cell, and I looked at the cell through another polarizer, I could see a change in density. I just did -- and I didn't.

 

[wiltw]

Ordinary LED displays do exhibit an effect characteristic of polarization of light emitted by them!
I turned on my smartphone display, and looked at it while wearing a pair of polarized sunglasses...and the LED lightened and darkened as I rotated my phone!!! The same lightening/darkening could be observed while looking at the LCD display of my Minolta Autometer Vf! If I wear polarized sunglasses in my car, I cannot read the station information displayed by the radio LCD, either.
I just looked at my Samsung OLED TV thru a CPL and it lightened and darkened as I rotated the filter, as expected...

1. But if I removed the hemisphere from my Minolta meter and aired its reflected light sensor at the TV, I got the same reading whether the meter was vertical or horizontal in orientation to the screen (I had thje screen frozen on a fixed image)
2. Yet when I did the same thing with my Minolta Spotmeter F, the reading off the screen did indeed change by 1EV!