Puzzling Selenium meter behavior

[Retro-Grouch]

A little while back, I scored a Sekonic-Master L-104, the same as a Weston and made under license by Sekonic. It was only a few bucks at the flea market, so I took a gamble and was very pleased to test it at home and see that it was dead accurate, unusual for old Selenium meters. I tested it on "low" in bright room light.
I just recently decided to start using it for landscape work, but upon using it in bright outdoor light, I realized that the "high" setting consistently gave me readings 1 EV higher than my other meters, both in reflected and incident modes. I'm puzzled, since my experience with Selenium meters is that, when they fail, it is with a gradual loss of sensitivity. Clearly, I won't be using this meter, as it can't be trusted, but I'm curious to know why it should show a non-linear increase in sensitivity.
I'm completely familiar with the use of this meter, having used one for years before I dropped it off a breakwater (ouch!), so user error can be ruled out. Anyone have any theories about this strange behavior?

 

[santino]

First I really don’t know. What does the low-high switch actually do? Maybe dirt/oxidation in a circuit? I‘m really interested to understand the cause too.

 

[raydm6]

I went online to find a manual to understand the "low/high" settings. Looks like the hinged baffle changes the sliding low/high scales. Could this high-scale perhaps be off a bit since this is a mechanical operation and maybe something has shifted the high scale position registration?

I have an older Weston Master Model 715 with a baffle and tried it on mine. My theory doesn't seem to hold much water as it looks like the scales are on one rolling integrated scale; unless yours is mechanically different. Also with the baffle in place (high lighting situations), I'm wondering if that, coupled with the selenium cell sensitivity being non-linear because of age - and perhaps also, something more, electrically, going on in the circuit could all be contributing factors?

See Page 5:

https://www.cameramanuals.org/flashes_meters/sekonic_master_iv.pdf

Japanese version at Sekonic:

https://www.sekonic.co.jp/product/meter/download/pdf/manual/L-104.pdf

Also, a little peek at innards and some tips although these don't appear to be the case here:

Manualsand repairs

Weston Master IV Disassembly & Notes (Page 1)

A look inside a Weston Master IV meter

www.jollinger.com

 

[Dralowid]

I had/have a Weston that was behaving irregularly. I fiddled around with the adjustment screw on the underneath and then made my best effort to reset the meter. As if by magic the shock of being messed with settled it down.

 

[Coldkennels]

That adjustment screw is really just for setting the zero point, @Dralowid - you shouldn't be tinkering with it for anything else.

@Retro-Grouch, at the risk of stating the bleeding obvious... is it just reading too large an area? I've always found selenium meters (whether "on-board" or handheld) tend to be easily tricked by the sky and things like that, so I always angle them down (or even point them directly at the ground).

 

[jankap]

I think your problem is due to the old-fashioned technic of a selenium cell.
A selenium cell looks wide, so it does not measure your subject alone. They measure perhaps 160 degrees wide.
Modern cells and such in cameras measure your subject or even less in spot mode.
The baffle is meant to use the meter pointing in the direction of the camera. In the case of portraits for instance measuring the incoming light on the subject.

 

[Dralowid]

That adjustment screw is really just for setting the zero point, @Dralowid - you shouldn't be tinkering with it for anything else.

@Retro-Grouch, at the risk of stating the bleeding obvious... is it just reading too large an area? I've always found selenium meters (whether "on-board" or handheld) tend to be easily tricked by the sky and things like that, so I always angle them down (or even point them directly at the ground).

Yes, I am aware that the screw is only for 'zeroing' but whatever, it did the trick in that case. In the past I have been inside a few Westons, complicated they aint!

 

[Retro-Grouch]

I appreciate everyone's effort to puzzle this out. I made sure to eliminate all possibilities for error in angle of view; the meter was pointed (in reflected mode) at a very large expanse of evenly lit wall, so it was reading what my other meters read. Mechanically, the meter is like new, so there is no leakage around the baffle. And, to respond to jankapp's post, the baffle is to switch between high and low sensitivity ranges (the baffle covering the cell is used in bright light). For incident readings, pointed toward the camera, one uses the white "invercone", a translucent piece that serves the same function as the white dome on any incident meter. The baffle is used alone for reflected readings on "high", and in conjunction with the invercone for incident readings on "high". Both incident and reflected readings on "high" are inflated by +1 EV.
Another forum member, in a lengthy PM, has very kindly offered a suggestion for calibration that he has used: he suggested taping over the holes in the baffle, one by one, until I reduce the +1 error to zero. This sounds quite logical, and will be a project in the near future. When I eventually get to it, I'll post to let others know if the trick works. Thank you all for your suggestions!

 

[markjwyatt]

I appreciate everyone's effort to puzzle this out. I made sure to eliminate all possibilities for error in angle of view; the meter was pointed (in reflected mode) at a very large expanse of evenly lit wall, so it was reading what my other meters read. Mechanically, the meter is like new, so there is no leakage around the baffle. And, to respond to jankapp's post, the baffle is to switch between high and low sensitivity ranges (the baffle covering the cell is used in bright light). For incident readings, pointed toward the camera, one uses the white "invercone", a translucent piece that serves the same function as the white dome on any incident meter. The baffle is used alone for reflected readings on "high", and in conjunction with the invercone for incident readings on "high". Both incident and reflected readings on "high" are inflated by +1 EV.
Another forum member, in a lengthy PM, has very kindly offered a suggestion for calibration that he has used: he suggested taping over the holes in the baffle, one by one, until I reduce the +1 error to zero. This sounds quite logical, and will be a project in the near future. When I eventually get to it, I'll post to let others know if the trick works. Thank you all for your suggestions!

I just noted yesterday that both my Kodak Retina meters (Reflex IV and IIIc) read 1 stop higher than my hand helds (all are consistent). I just did it again with my Sekomic L-308 and the two Retinas, and the Retinas are both a stop higher (condition: Sunny-16 per hand held).

 

[Retro-Grouch]

I just noted yesterday that both my Kodak Retina meters (Reflex IV and IIIc) read 1 stop higher than my hand helds (all are consistent). I just did it again with my Sekomic L-308 and the two Retinas, and the Retinas are both a stop higher (condition: Sunny-16 per hand held).

This is not at all unusual. If one uses multiple meters, it is most often necessary to calibrate them (via ISO adjustment) to one standard, usually that of your known accurate meter. What is unusual is the non-linear response and increased high sensitivity of my old Selenium Sekonic. Even new and out-of-the-box, meters can vary a good deal.

 

[Deleted member 82967]

This is not at all unusual. If one uses multiple meters, it is most often necessary to calibrate them (via ISO adjustment) to one standard, usually that of your known accurate meter. What is unusual is the non-linear response and increased high sensitivity of my old Selenium Sekonic. Even new and out-of-the-box, meters can vary a good deal.

Agreed. I've owned several Sekonic L-398 selenium models and they consistently recommended less exposure than I would have liked for my Sigma SD9.