Spectral Response of Meter Cells?

Tom1956 · Feb 14, 2014

There is currently another thread broaching the subject of spectral response of meter cells (link following this sentence), but I decided to introduce a new one so as to be more focused on the question. Link: (there was a url link here which no longer exists)

Hopefully I can draw some knowledgeable responders on the subject. Perhaps put another way, black and white film tungsten ratings, or metering in other lighting conditions than daylight. I believe it is universally accepted that exposure meter sensitivity is planned to be in the 5600K neighborhood of color temperature. But what about metering in other lighting? What sort of allowance should be made for the meter's non-linear behavior in different kinds of light? I post here a link leading to the spectral response of one single distributor's product, a CdS cell: http://www.token.com.tw/resistor/photo-cds.htm

But I have never undertaken a study on this subject. I'm just a person who has experienced a lifetime of poorly exposed negatives and never considered why. Now interested in the subject, I put forth this thread fishing for responders who may be better versed in the matter. This one link I just posted is my first step in the study of the matter. Hardly a definitive treatise. Any input welcomed.

dorff · Feb 14, 2014

Well, the obvious goal is to match the spectral response of the film and the meter. In simplest terms, that means you could calibrate your meter by metering off specific colours and doing step-wedge exposures and seeing what EI for what colour gives the best exposure. Newer matrix-type SLR metering systems are colour-aware, and therefore do the compensation for you, as do the more advanced studio light meters. If as you say your negatives are perpetually badly exposed, I can only suggest that you consider more sophisticated equipment or do a series of tests with what you have and compensate accordingly. Last but not least, bracket whenever practical to do so.

Xmas · Feb 14, 2014

As well as the/a meter material data sheet a meter may have a sharp cut IR filter - like a M8 needs to avoid purple sheens. DSLRs can have hot filters alas I don't do DSLRs...

Then you need the each films data sheet spectral response.

The films I use have significant differences from near infra red traffic surveillance to ortho!

The light does vary in colour temperature you can get colour temp meters.

Tungsten light (colour) is dependent on mains voltage (your enlarger may be stabilised), bulb type, dimmer setting. Movie film can have a meter setting for tungsten and daylight 5222 has.

But setting meter to the pre 1961 speed rating is simple or using C41 mono like XP2 at 200 ISO.

Bill Burk · Feb 14, 2014

Here's a fun thread cross-reference:

(there was a url link here which no longer exists)

So you have human eyes which see a peak at green, film which sees a peak at blue, meters which see a peak at red*... Good luck lining any of them up.

*the Weston Photronic cell (I picked a fairly red-sensitive meter cell deliberately to present the worst case scenario).

dorff · Feb 14, 2014

If there were to be a meter with colour awareness into the IR region, it would be dandy to be able to select the film type (loaded as a curve into the meter) or at least be able to define a custom curve. But it means a meter manufacturer needs to make a programmable meter, possibly augmented with built-in or add-on filter sets. Imagine, you don't set the ISO, you select the film itself. Everything taken care of.

Xmas · Feb 14, 2014

dorff said:
If there were to be a meter with colour awareness into the IR region, it would be dandy to be able to select the film type (loaded as a curve into the meter) or at least be able to define a custom curve. But it means a meter manufacturer needs to make a programmable meter, possibly augmented with built-in or add-on filter sets. Imagine, you don't set the ISO, you select the film itself. Everything taken care of.

An Leica M8 and a steep cut IR and steep cut Red filter would do, they are getting cheaper, the steep cut IR was free with new ones (M8) for a while. If the meter fails trash can.
A custom meter might be more expensive.

cowanw · Feb 14, 2014

Do we have to distinguish between reflective and incident meters in this thread. I ask because I can't imagine the point of specifying the 5600 k incoming light if the metre is measuring reflected light.

Tom1956 · Feb 14, 2014

I wonder what Brian Shaw said before he deleted it. He's one of the reputable names on here. I'd have liked to heard it. I wish the one site I found a data sheet on the CdS cell had not passworded their PDF, so I could bring it into photoshop and crop the response curve to post here. It was very enlightening. But suffice to say it was VERY peaked in it's sensitivity range. Enough to make you wonder how these meters are calibrated and what film speed compensation must be considered outside that peak. And another thing--back before about 1970, B&W film had a daylight rating and a tungsten rating. I always assumed it was the film. Now I wonder if it was because of meter color sensitivity. And if so, why doesn't the film manufacturer have a tungsten rating now? Why did they drop it?

Xmas · Feb 14, 2014

HiTom

Double-X Kodak's cine mono still has a tungsten and daylight ISO.
Probably for three reasons
Think it is quite blue sensitive
The cine people are more critical on exposure than stills
Not many people use tungsten studios?

Noel

Gerald C Koch · Feb 14, 2014

Xmas said:
HiTom

Double-X Kodak's cine mono still has a tungsten and daylight ISO.
Probably for three reasons
Think it is quite blue sensitive
The cine people are more critical on exposure than stills
Not many people use tungsten studios?

Noel

I think your second reason is the most probable. The difference between the two values is only 1/3 of a stop. The fine grain positive stock on which the negatives are printed has little latitude. Still films probably have similar values but it is not critical for still exposure.

Tom1956 · Feb 14, 2014

Ahh... I've finally glommed on to a graph of the sensitivity of a CdS cell. Look at this, it's astounding. If this is the truth, then many built-in 35mm meters can be WAY off in different kinds of light.

Xmas · Feb 15, 2014

Tom1956 said:
Ahh... I've finally glommed on to a graph of the sensitivity of a CdS cell. Look at this, it's astounding. If this is the truth, then many built-in 35mm meters can be WAY off in different kinds of light.View attachment 82306

Hi Tom

Yes...
Only the CdS cell (probably the left graph - Im on a phone) was used in cameras and light meters until superceded by silicon cells.
The CdS meters would work in moon light and their response could be improved by a tailored filter, like Leica M8s needed.
People used to use tables, then Selenium, then CdS, ... etc., they work good in 98 % of most peoples shots.
You need to understand their limits for the other 2%.
I still use a table and a Selenium.
Before tylight I'm ODing on latte and muffins, in a WiFi cafe.
Ansell used Selenium but his phots are better then mine.
Dont forget about the mercury battery problem... and CdS aging...

Noel

RalphLambrecht · Feb 15, 2014

dorff said:
Well, the obvious goal is to match the spectral response of the film and the meter..

Is it?or ;if you believein and work withprevisualization is it the match betweeneye and meter?:confused:

Bill Burk · Feb 15, 2014

Ralph, you're right.

Visualizing - making what your mind sees... understand what spectrum your eyes are sensitive to, understand the spectral sensitivity of the film. And understand the meter's sensitivity.

Throw filters into the mix and you have four different charts to comprehend.

Boggles the mind if you let it. But my favorite approach is to consider the "folklore" of photography advice, and see how it fits in with the specific combination I am using.

Then it becomes fun. I can bring out my Weston Master II and see the world in a different light.

Sirius Glass · Feb 15, 2014

Tom1956 said:
Ahh... I've finally glommed on to a graph of the sensitivity of a CdS cell. Look at this, it's astounding. If this is the truth, then many built-in 35mm meters can be WAY off in different kinds of light.View attachment 82306

This aligns with what I remember about CdS meters. They were blue sensitive and not so good for reds or IR film.

BrianShaw · Feb 15, 2014

I always thought that the obvious goal was to understand any significant interactions between the spectral response of a meter and the color of light, scene, and filter to best determine the desired exposure values. Some would call that 'visualization". I call it "anti-formularization" but that lacks panache and doesn't roll off the tongue as good.

BrianShaw · Feb 15, 2014

p.s. I'm with Bill B on the "folklore" and tribal wisdom. Between that, the proven smarts of the camera/meter manufacturers (in most cases), and the smarts of a smart photographer it is possible to make good images without dissecting the minutia of engineering details surrounding well-proven historical equipment. Not that that can't be interesting, of course.

Dan Fromm · Feb 15, 2014

BrianShaw said:
p.s. I'm with Bill B on the "folklore" and tribal wisdom. Between that, the proven smarts of the camera/meter manufacturers (in most cases), and the smarts of a smart photographer it is possible to make good images without dissecting the minutia of engineering details surrounding well-proven historical equipment. Not that that can't be interesting, of course.

Ignorant barbarians insensitive to the fine points (that's me) just use their equipment as directed and nearly always get good results (technically, not artistically).

I've hit only two situations out-of-doors where having a color temperature meter and CC filters might have been useful, and then only with color film. Once upon a time in the Big Cypress at the Sweetwater Strand ... I shot an alligator in deep shade. Nikon FM2n, ISO 100 E6. Roses are red, gators are blue, and my little blue gator was well-exposed. Score one for Nikon. Once upon a time in my backyard I was trying out a B&L Tessar IIb. Acceptance testing, done with ISO 100 E6 in a N8008S. When I took the test shots the sun was behind a cloud. White subject came out a well-exposed blue. Same problem as my little blue gator. Score one for Nikon. On retest with the sun not hidden behind a cloud the blue cast didn't appear.

When my LunaPro (= LunaSix III) and Master V were both freshly calibrated they never agreed exactly. No matter which I followed I usually got good exposures on E6. The exception was with a 160/5.6 Pro Raptar in the infamous blue-faced Rapax, top speed 1/200. Turns out that the shutter ran very very slow at its top three speeds.

In other words, some posters here worry too much and test too little.

Tom1956 · Feb 15, 2014

I'm not sure I'm gleaning what I had hoped from starting the thread. Specifically, shouldn't I start paying more attention to making allowance in my camera settings for other color temperatures of lighting than daylight, for which it is understood the film speeds are dialed in for on the meter dial? Look again at the sharp cut-off in the graph for color temps outside the peak. Until I began this research I had no idea meter cells were THIS peaked in their sensitivity. They quickly become downright insensitive outside this peak.

Bill Burk · Feb 15, 2014

Hi Tom1956,

The vendor's chart you found seems to illustrate three different products they offer, which seem to be designed to peak at different colors.

The reference I found in Todd-Zakia Sensitometry gives a single CdS curve, probably an average of cells available at the time, that is a little less abrupt...

It seems to indicate CdS has 80% sensitivity or more over the range from blue to orange.

And then the Weston (I assume Selenium Solar Cells) leans slightly more into the red.

Folklore-wise, that tells me to use a "Tungsten" rating when evaluating indoor incandescent exposure with a Selenium-based meter.

And it would also tell me when using that meter, I would be better off to meter without a filter and then apply a filter factor when using a deep red filter.

And I think that will help you avoid underexposed negatives when using a deep red filter... Which is what I think should be your primary concern... It is mine.

Tom1956 · Feb 15, 2014

Thank you Bill for your input. I can see now how I have been making more something out of nothing for the most part. I suppose though it would serve me well to wind off a roll from my bulk loader and do a little bit of gray card testing. I've been in photography since Methuselah finally kicked the bucket, but only here in the past couple years I've decided to set myself a pretty strict accuracy standard. !/4 stop would be nice, but 1/3 stop is a reasonable standard I hope you'd agree. Anything more just means more test stripping and foolishness in the darkroom.

Mr Bill · Feb 15, 2014

Bill Burk said:
Hi Tom1956,

The vendor's chart you found seems to illustrate three different products they offer, which seem to be designed to peak at different colors.

The reference I found in Todd-Zakia Sensitometry gives a single CdS curve, probably an average of cells available at the time, that is a little less abrupt...

Hi, I don't know if Todd and Zakia told what the (RCA) S-3 and S-11 are, but the designations sound like they may be photomultiplier tubes. If so, they would have been more at home in an densitometer from that time period than in a hand-held exposure meter.

All of those sensors could have had their response manipulated just about anywhere with filters, so one should keep in mind that an actual exposure meter might not respond the same as the sensor data indicates.

Mr Bill · Feb 15, 2014

Tom1956 said:
Perhaps put another way, black and white film tungsten ratings, or metering in other lighting conditions than daylight. I believe it is universally accepted that exposure meter sensitivity is planned to be in the 5600K neighborhood of color temperature. But what about metering in other lighting?

I don't know what current ISO standards call for, but going back to 1970s ANSI standards, the reference standard was 4700 K. (see the following)

From ANSI PH3.49 - 1971, Appendix B:

The primary standard for color temperature of a 4700 K radiation consists of a tungsten lamp operated at 2854 K in combination of the Davis and Gibson filter. This source not only matches closely the spectral-energy distribution of a blackbody at 4700 K, but is also reproducable from specifications.

All they had to say about spectral response requirements for the meter was that it had to be "continuous in the range between 350 nm and 700 nm." But the test seemed less complete, only requiring that the light detector match readings (at fixed luminances) at 2854 K vs 4700 K.

Bill Burk · Feb 15, 2014

Tom1956 said:
Thank you Bill for your input. I can see now how I have been making more something out of nothing for the most part. I suppose though it would serve me well to wind off a roll from my bulk loader and do a little bit of gray card testing. I've been in photography since Methuselah finally kicked the bucket, but only here in the past couple years I've decided to set myself a pretty strict accuracy standard. !/4 stop would be nice, but 1/3 stop is a reasonable standard I hope you'd agree. Anything more just means more test stripping and foolishness in the darkroom.

I agree 1/3 stop is a good goal. I personally adopted 2/3 stop for my goal. Just seems to me that a mistake in any assumption would lead to 1/3 stop "error" - so the best I hope for is that a couple of 1/3 stop errors cancel each other out... leaving me within 2/3 stop of what the exposure should be (if I make four errors).

Spectral Response of Meter Cells?

Privacy & Transparency

Privacy & Transparency