Light Meter Calibration

Bill Burk · Jun 24, 2017

Chan Tran said:
And when meters agree within 1/2 stop or so they are within 40% of each other. Do you have another measuring tool that you could accept a 40% difference between one tool and the next?

How does a half stop work out to 40%? Seems a very minor difference to me.

wiltw · Jun 24, 2017

Chan Tran said:
Measure a Kodak gray card under the same lighting. Careful to avoid specular reflection.

Commonly repeated MYTH

benveniste · Jun 24, 2017

Kawaiithulhu said:
https://www.yelp.com/biz/quality-light-metric-hollywood
That company/service comes up a lot in light meter discussions. Haven't tried them myself, though.

I have, and they did a great job on my Minolta Color Meter. Recommended.

Chan Tran · Jun 24, 2017

wiltw said:
Commonly repeated MYTH

That's for you only.

Chan Tran · Jun 24, 2017

Bill Burk said:
How does a half stop work out to 40%? Seems a very minor difference to me.

1/2 stop more is 41% increase in brightness or 1/2 stop less is 30% decrease in brightness. It does seem minor. But even if we accept the logarithmic nature of the exposure meter think about it this way. Most meters measuring range is less than from 0 to 20EV @ ISO 100. (typically 1 to 17). But take it as 20 EV range and the error is 0.5 then it's error is 2.5% which is significantly less than a typical ruler, thermometer, scale etc.. we have around the house.

Chan Tran · Jun 24, 2017

dmr said:
No! And this is why I still don't get that warm fuzzy feeling when dealing with the accuracy of light measurement. My only consolation is that I know that the film I usually shoot (almost always color negative) can usually be overexposed or underexposed a stop or so with good results. I want to be within the tolerance and latitude of slide film, however, which can't take a joke, exposure-wise. I'm also in the habit of bracketing on any semi-important shot.

I am with you. I want my meter to be accurate to better than 1/10 of a stop (that's a whopping 7%) but I can't afford a light meter just exposure meter. I don't care how critical my film or digital sensor is. I want my meter to be accurate.

Leigh B · Jun 24, 2017

wiltw said:
Commonly repeated MYTH

How so?

Kodak put a lot of time and effort into designing a grey card with accurate reflectance.

Kindly elucidate your credentials for challenging their results.

- Leigh

Bill Burk · Jun 24, 2017

Wiltw has demonstrated that incident meter reading does not easily tie to a gray card exposure.

Because the angle you hold the card can dramatically alter the effective luminance of a gray card held in a given amount of incident light. It's a dramatic demonstration that you would need to be careful if you were trying to relate an incident reading to a gray card exposure.

But you can setup a reasonable test using a spotmeter to read gray card from camera position, and then take that picture without moving the card.

In this test, I used the spotmeter function of OM-4 to meter 18% gray of a target with several patches between 3.6% to 90.7% and seven patches closely surrounding 18%. I have marked all the patches on this graph. In the same shot I've got my dog and you can see where his eyes, ears and nose fall on the curve. The curve is from a sensitometer exposure on the same roll of film. All the marked densities may be considered to fall on the curve (but assume the test shot received the same development as the sensitometric exposure - a fairly valid assumption, however you must know that in small tanks sometimes different parts of film can receive different amounts of development).

I expected the density of the 18% gray patch on the curve to fall immediately under the -1.1 (10) mark on the film speed & log mcs scale.

Stephen Benskin and I looked at this and from the result, I conclude there is an 0.05 or 1/6 stop error in my meter (or experimental error resulting in 1/6 stop deviation between expected and actual results).

Bill Burk · Jun 24, 2017

So for the short story. That graph is how I would check an exposure meter if I didn't have a 4700K calibrated light source.

wiltw · Jun 24, 2017

Leigh B said:
How so?

Kodak put a lot of time and effort into designing a grey card with accurate reflectance.

Kindly elucidate your credentials for challenging their results.

- Leigh

Rather than credentials, let me produce scientific testing which are reproducable by others, not merely claims of my own which are not backed by evidence.

First of all let me point out: 18% tonality is simply 'the middle of the tonal range with 'white' and 'black' at the extremes of the range. Period. Mr. Adams so illustrates in The Negative.
There are discussions about the fact that the 'metering tonality' needed for correct exposure which places '18% target in the middle' is closer to 12% tonality. This can be derived via the ISO standard equations for calibration of reflective light meters and for calibration of incident light meters. The illustrious Ctein writes to this point about the fallacy of metering off 18% targets on p. 33 of his book, Post Exposure.

The following histogram shows where exposing +0.33EV (which is NOT where the 18% target would have you expose) puts a Photovision target (black-gray-white tritone panels) so that the 18% mid-tone indeed falls centered on the histogram.

and this next shot shows when you expose with the spotmeter centered on the 18% gray portion of the Photovision target, the peak for mid-tone 18% falls TO THE LEFT OF CENTER on the histogram

Bill Burk referred to a demonstration that I did, which illustrates the variability of light reflected back to the lens by an 18% tonality target, simply based upon the ANGLE of the card, a series of shots at identical f/stop and shutter speed within seconds of each other...

And Ctein even mentions the specific Angle to hold the card, as mentioned in Kodak instructions in Kodak Professional Dataguides

Leigh B · Jun 24, 2017

wiltw said:
Bill Burk referred to a demonstration that I did, which illustrates the variability of light reflected back to the lens by an 18% tonality target, simply based upon the ANGLE of the card, a series of shots at identical f/stop and shutter speed within seconds of each other...
{graphic removed}
... the specific Angle to hold the card, as mentioned in Kodak instructions in Kodak Professional Dataguides

All the photos prove is that you don't understand the instructions for proper use of the card.

That's not Kodak's fault.

- Leigh

faberryman · Jun 24, 2017

Given all the pitfalls and foils of using an uncalibrated lightmeter, it's amazing we can get properly exposed photographs at all. Yet they mostly come out that way.

wiltw · Jun 24, 2017

Leigh B said:
All the photos prove is that you don't understand the instructions for proper use of the card.

That's not Kodak's fault.

- Leigh

I fail to see how illustration of the variability of reflectance proves that I don't know now to use a card ?!

Illustrating the variability shows to others that a 'reference standard' card can be tremendously fooled by a variable result. The Kodak definition of holding the gray card has been missing from some editions of the Kodak gray card, so it is not necessarily 'improper use' that causes the problem, and simply avoidance of card surface sheen reflectivity does not ensure proper use either.

The explicit Kodak instruction in 2007 is "Aim the surface of the gray card toward a point one third of the compound angle between your camera and the main light." I bet most folks merely know only to avoid surface sheen. This was on the Kodak web site in March 2016, in comparison...

"Exposure
"When shooting the gray card, it is recommended that you meter the card with a spot meter. Point the meter at the center of the card and use the reading directly. It is a good practice to have the meter reading for the gray card match the shooting stop, which is typically determined by using an incident meter. In most cases - if the gray card is in the key light - the two readings will be very close. Tilting the card toward the key light may be necessary for them to match exactly. It takes practice, but this method yields accurate and repeatable results."

http://motion.kodak.com/us/en/motion/education/publications/shooting/gray_card/default.htm

One version of card purchased by someone in 1999 says only, ""Meter readings of the gray card should be adjusted as follows-
1) For subjects of normal reflectance increase the indicated exposure by 1/2 stop.
2) For light subjects use the indicated exposure; for very light subjects decrease exposure by 1/2 stop
3) If the subject is dark to very dark increase the indicated exposure by 1 to 1.5 stops"

So since Kodak itself is not consistent in its recommendations, how can you accuse me of not knowing how to use it?! I merely illustrated the variability of its reflecting light back to the lens.

Bill Burk · Jun 24, 2017

It is a good practice to have the gray card reading match the shooting stop. My advice for you is to read that paragraph a couple times.

I take it to mean that you hold the card in the key light and angle it until the spotmeter reading agrees with the incident reading.

Isn't that exactly what it says?

Well then let's try that and see if you can get an incident meter to agree with a spotmeter.

I would hazard to say if you had a 12% reference held at the "correct" angle in the same frame as the "gray card tilted" until incident agreed with spotmeter... the 12% card would match the 18% card

wiltw · Jun 24, 2017

Bill Burk said:
It is a good practice to have the gray card reading match the shooting stop. My advice for you is to read that paragraph a couple times.
I take it to mean that you hold the card in the key light and angle it until the spotmeter reading agrees with the incident reading.
Isn't that exactly what it says?

Well then let's try that and see if you can get an incident meter to agree with a spotmeter.

I would hazard to say if you had a 12% reference held at the "correct" angle in the same frame as the "gray card tilted" until incident agreed with spotmeter... the 12% card would match the 18% card

A nice experiment to do, but we can't buy an EZ Balance 12% card here in the USA. Someone in UK or Canada can buy an EZ Balance 12, I believe.

As for the spotmeter and incident reading agreeing, my Minolta Autometer Vf incident readings already matches my Minolta Spotmeter F pointed at an 18% gray card, and they both agree with my Canon 40D and 5D and 7DII bodies.
But then we have a complication that if I take all three bodies and shoot them with identical exposures, my 7DII puts the histogram peak about 1/3EV to the right of the 40D or 5D histogram peak...putting the 18% card peak perfectly at the mid-line of the histogram just where it should be!

RalphLambrecht · Jun 25, 2017

John Koehrer said:
The meter you have has standards that are a bit lax. I don't believe there's a standard
for K factor for all makers.
Re: consistency Once you've got your standard(for your developing method) As long as you don't
change anything, you're good.
Spot meter: The reading's are influenced by the subjects color and I think, reflectivity(shiny or matte).
Once you learn it it's good too.

I'm lazy & haven't used anything besides an incident meter for about 100 years

You clearly have the most experience of all of us!

Bill Burk · Jun 25, 2017

wiltw said:
A nice experiment to do, but we can't buy an EZ Balance 12% card here in the USA. Someone in UK or Canada can buy an EZ Balance 12, I believe.

As for the spotmeter and incident reading agreeing, my Minolta Autometer Vf incident readings already matches my Minolta Spotmeter F pointed at an 18% gray card, and they both agree with my Canon 40D and 5D and 7DII bodies.
But then we have a complication that if I take all three bodies and shoot them with identical exposures, my 7DII puts the histogram peak about 1/3EV to the right of the 40D or 5D histogram peak...putting the 18% card peak perfectly at the mid-line of the histogram just where it should be!

My Sekonic target has suitable patches. You're welcome to stop by sometime to experiment. I've got a 100 Ft Lambert standard to give us a point of reference. My meters tend to disagree incident vs. spot. But I figure some differences are due to calibration. I think they are all right.

AgX · Jun 25, 2017

Chan Tran said:
To calibrate the meter you need standard light source of varying intensity. You would need 2 type of light source for reflective and incident.

Why?

There is nothing to calibrate between incident and reflective mode. Best one can do is to test. That means testing the diffusor.
Irregular calibrating then would mean adding a ND-filter or grinding-off the diffusor.

Craig75 · Jun 25, 2017

Unless shutter speeds, aperture, and speed of yr film have also been checked (and perhaps speed of film with every box) then i cant see how a calibrated meter is going to be any advantage

Chan Tran · Jun 25, 2017

AgX said:
Why?

There is nothing to calibrate between incident and reflective mode. Best one can do is to test. That means testing the diffusor.
Irregular calibrating then would mean adding a ND-filter or grinding-off the diffusor.

Which meter are you talking about. I am talking about calibrating 2 different meters. You need one light source for reflective meter and one for incident meter.

Bill Burk · Jun 25, 2017

Craig75 said:
Unless shutter speeds, aperture, and speed of yr film have also been checked (and perhaps speed of film with every box) then i cant see how a calibrated meter is going to be any advantage

I see advantages of checking all of these. Of these, speed of film is the one that changes the least unless you make your own emulsion or play with expired film.

I don't know if they all need to be "calibrated" per-se, but checking to make sure they are working correctly could definitely be an advantage.

Sometimes with old light meters you find something wrong. Maybe just a loose screw or a bit of metal on the magnet that throws off your readings. In the older mercury battery meters you might have a definite problem using alkaline batteries (That would require special batteries/adapters/calibration).

But I would definitely want to make sure that my meter is "working".

I think one standard trans-illuminated light source would be enough. You can then adjust another light to match the standard (I move a photoflood closer and farther until the reflected light from a gray card agrees with the transilluminated light standard - I guess this depends on correct use of gray card, so it will be important to note the test layout so others can check/replicate your results). You can reduce the light to check the range, using scrims or light meter grid inserts.

dmr · Jun 25, 2017

Bill Burk said:
if I didn't have a 4700K calibrated light source.

So? Out of curiosity, how much ($$$) is such a calibrated light source?

John Koehrer · Jun 26, 2017

^^^I'm a bit curious too. When I servicing this stuff our tester used about 3200K
but were able to vary the k factor.

Bill Burk · Jun 26, 2017

Although 4700K is what you want, I think mine is 2850K. It was about sixty bucks (I'm a patient shopper). Add an 80B filter and you are on your way.

Cary Lee · Jun 30, 2017

My two cents....I had all my meters calibrated by Quality Light also...luckily they are close to me in Los Angeles. I've had a couple incident meters be off by a stop but it really doesn't matter if your shooting analog film when using the zone system. Since all meters measure for middle gray anyways knowing how exposure effects your overall picture is more important than any given measurement. When I shoot large format, I use my pentax digital spotmeter or my analog Pentax V manual meter and measure for the shadows and develop for the highlights all within zone 2-7. If it is outside that then I adjust in development and as long as it is within the zones then my pictures comes out good to go. Now if you're trying to calibrate your meters to your spot meter on a digital camera or light table...then it will be kinda hard to master. When I shoot digital i wouldn't start shooting anything until my first shot includes using my color checker passport as reference guide. With it, I find it easy to color correct in post since i have a reference tool to use as base standard.

Light Meter Calibration

Privacy & Transparency

Privacy & Transparency