Oh boy, here we go . . . Zone System. First of all, the above scale is a very general reference - I don't want any lawsuits from any Zone System Archbishops.
First of all, as Michael Langford said so well in Basic Photography, these things are called light meters, not exposure meters. They measure light, not exposure. Good?
Second, light meters are inaccurate. They’re glaringly inaccurate at the brightness extremes. I’ve never seen one that wasn’t. The aim is to find one that’s the least inaccurate. What are the extremes and what is the inaccuracy? Generally speaking, anything below 1 candle per square[1] foot (c/ft2) and above 100 c/ft2[2] will give you inaccurate readings. The low brightness readings will be too high, and the high brightness readings will be too low. You underexpose in the woods at dusk and overexpose light toned objects in bright sunlight.
What does candles per square foot mean? You don’t need to know. I’ll give the equivalent on a Pentax digital spot meter. EV 10 is 10 c/ft2. So 2 c/ft2 is EV 8 and 100 c/ft2 is EV 13. These indications are pretty much the same for Minolta spot meters (if you can read them), and the blue cell Luna Pro meters - at ISO 200.[3]
I’m sure you know all about these inaccuracies, right? I’ll bet you’ve looked at sheet film in the wash water or when unwinding rollfilm off the reel and thought “what the hell happened there?” It’s probably the meter. But we doubt ourselves and never blame the meter. We can’t do anything about it anyway – who is going to listen to us? Do we take the meter back? How do you prove it’s inaccurate? Even if you provide test results, are you going to explain this to a camera store clerk who says “just average the readings”?
What’s the problem? – the photocell
The Achilles heel in light meters is the photo cell. Among other things, colors, ultraviolet radiation, and extremes in brightness fool the photo cell. The old Soligor spot meters were the worst offenders. I remember trying a Spot Sensor II that defied the laws of physics: for my asphalt driveway it gave the same reading in both sun and shade! Unbelievable.
Photo cells that directly read subject brightnesses are always a wild card. They’re subject to many variables. I suppose that people doing precise work made sure their meters were checked by experts, if you can still find any.[4] The rest simply put up with it – unknowingly.
Zone VI used to make a modified Pentax digital spot meter that was supposedly accurate. I never used one but Richard Ritter told me they worked fine and if he says so, then I’m sure it’s true. But – only so many of them were made; eventually connections loosen, and they’ll inevitably break.
The best meter wouldn’t use a cell at all. It would compare a known brightness in the meter to brightnesses in the world. More on this later.
Think
Light meters are most accurate in the mid range. How do we use them in extremes of brightness; very high or very low? Easy – don’t read the extremes.
Let’s say you have an autumn hayfield in bright sun. It’s dry and of light value – bright. It’s probably 500 or 1,000 c/ft2 or higher. If I meter it directly I’m guaranteed an inaccurate reading. Here’s what I do. At the same camera angle, find a place where the field is shaded. That’s 2 stops lower. So I do that. It reads 200 c/ft2. That may be accurate, but still feels a bit high to me, and 200 c/ft2 is higher than the meters “mid range.” I have the black side of my focusing cloth. White objects reflect 5 stops more light than black objects in the same light . . . about. The black side of the focusing cloth in sun reads 50 c/ft2.[5] So the shadow reading for the hayfield was correct – this time . . . probably. Field in sun on VIII, so 1/200 at f 16 places the shadowed portion on VI, so 1/100 (1/125) at f 16[6] is the exposure (adjusted later for depth of field with the view camera). If the shaded black side of the focusing cloth was in the picture it would fall on III. That exposure will work.
Suppose I’m wrong? If I’m off a bit, it’s probably by such a small amount, that I’d still get a negative I could print easily. What if I don’t? So what? There are plenty of other good pictures out there.
If you’re at the other extreme – dark conditions, exposure calculation is easier because the latitude is so great. In dim lighting it’s almost impossible to overexpose. I’ve never seen it happen. Give plenty of exposure in dim light. Even with reciprocity failure, if the meter reading indicates 1 minute, I always give at least 3 minutes. If the indicated exposure is over 1 minute, I give at least 5 minutes. The difference between 3 minutes and 5 minutes is very slight – less than ½ stop.
Readings. If you’re in the woods at dusk or before dawn, your meter might indicate from 1/8 c/ft2 to 1/2 c/ft2[7] for the brightest tree trunk. (On my meter that’s EV 3 to EV 6 at ISO 200.) It’s very likely the meter is indicating more light than exists. If you read your open palm facing the sky that might read 2 or 5 c/ft2[8]. Place that value on Zone VI and double the indicated exposure. You’re sure to get a printable negative. If your palm is too dark, try a piece of white cloth or paper. If you can get that to 1 c/ft2 at least (EV 7), you’re probably in good shape. Place the value on Zone VIII and double the exposure. For that example the exposure would be 1 minute. If the subject is very dark – wet tree bark or swamp water – I’d give 3 or 5 minutes. You’ll be okay because it’s nearly impossible to overexpose in those conditions.
The only accurate meter
The most accurate light meter didn't used a photo cell. No, it wasn’t the SEI. For calibration, the SEI used a selenium cell and ammeter which introduced pesky variables.
The Ainger-Hall Photometer looked a bit like the SEI, and was also developed and manufactured in the UK. Like the SEI it was a comparison spot meter that used a bulb and a big battery. Also like the SEI, it compared a given brightness in the world with a calibrated and dimmable “light spot” visible in the meter’s field of view. When the values matched, you had an accurate reading. And the Ainger-Hall had a foolproof element – it required no photocell. Nothing could trick it.
But it was cumbersome to calibrate. You needed a darkened room and a candle. You adjusted a neutral density filter in the meter to match the lamp brightness to the candle brightness. When they matched, the meter was calibrated: 1 footcandle in the world equaled the meter’s indication for 1 footcandle. From there, figuring exposure was simple. You compared and adjusted the light spot to subject brightessess; when the spot matched the subject brightness, you read off the c/ft2 value and calculate the exposure. I think the Ainger-Hall had shutter speed and aperture scales too.
The Ainger-Hall meter wasn’t completely foolproof. As the bulb and battery aged, the readings varied. I suppose you would just recalibrate the meter occasionally. The SEI meter eliminated the need for the darkened room and candle by using a selenium cell and an ammeter to calibrate the bulb and battery output. But even though the selenium photocells weren’t exposed to subject brightnesses, they’d weaken over time. The bulbs were soldered in place, couldn’t easily be replaced and eventually failed. The ammeter would eventually fail also. But it was still a brilliant idea.
I like the Ainger-Hall concept best. It was the simplest, most accurate, and least subject to error. With modern technology, it would be relatively simple to replicate and improve upon. First, produce a known brightness for calibration – say an 8” x 8” LED panel that produced 10 c/ft2. Next, all you’d need is a dimmable LED spot projected onto the subject field. You set up the calibration panel, and view it through the meter. Then adjust the spot brightness until it matches the panel brightness and set 10 c/ft2. Lock that setting, and the meter is calibrated. No ISO settings, no photocell, no EV numbers, no ammeter, and certainly no zone dials – just a number.
Since footcandles are measured geometrically like shutter speeds, exposures would be child’s play. Once your film is tested and you have your key stop (mine is f 16), exposure calculation is quick: (1) the scale of c/ft2 readings are identical to shutter speeds; and (2) any c/ft2 reading, converted to a shutter speed, at the key stop, equals Zone V. So to place a meadow on Zone VI: the reading indicates 100 c/ft2. 1/125 at f 16 gives me V, so 1/60 at f 16 gives me VI.
Visualize a portrait in sidelight. Read the lit skin tone, it reads 50. The shadowed skin tone reads 10 and the black sweater in shadow reads 5. Exposure is 1/25 (1/30) at your key stop – for me 16. The skin tone would be VI, shadowed skin tone is IV, and the sweater is III. Easy. Suppose you’ve got a picture of a granite cave and want detail inside? If the interior reads .4 (1/4 c/ft2) and the outer rock reads 10. Exposure is 1 sec at the key stop, the shadowed interior will be III and the outside rocks VIII. What if the outer rocks are sunlit? They’ll read 50 c/ft2 at least and you’ll have some decisions to make. But – with a comparison spot measurement, you’ll know the readings are accurate. Unlike a photocell meter where the low and high readings are always suspect.
Why we ever veered from footcandle readings I’ll never know. Random EV numbers needlessly confuse exposure calculation, and introduced wholly unnecessary complications.
Recap - think and think again
For 35mm work, I don’t even use a meter. For Foma 400 film the brightest condition is 1/125 at f 16; hazy sun is one stop more (f 11); bright overcast, 2 stops more; dark overcast 3. For a very dark subject in very dark conditions, 4 or 5 stops more. All this applies as long as you develop the film a little less than normal. When it gets much darker than that, the camera can’t be hand held. I always give more exposure and reduced (minus) development for 35mm work. I don’t meter individual brightnesses individually with 35mm film because the film area is so small. It’s easier to just view the scene as a whole - and a little less development to give you more latitude.
For larger formats than 35mm - and more precise exposure - you must simply assume that your meter is inaccurate at the extremes. So for very low readings (below 1 c/ft2), assume it’s wrong. I meter something brighter and give more exposure anyway. For very bright subjects – above 200 or 400 c/ft2 – assume it’s wrong - especially for things like snow or white painted objects in sun, or sunlit clouds. Meter something darker. There are only so many possibilities anyway. Eventually you know what the actual brightness values are from experience. Check your proofs and assess.
Once in a while there’s a critical, unique situation that you’ve never seen before when you’re uncomfortable with your options. This is when you curse your meter and have to wing it. That’s life.
*
[1] On my meters, that’s anything below EV 7 at ISO 200.
[2] EV 13 at ISO 200.
[3] You may need to adjust your ISO depending on the meter so that EV 10 equals 10 C/ft2. See the exposure formula article.
[4] Quality Light Metric in Hollywood, CA is out of business.
[5] 5 stops less than the sunlit hayfield.
[6] F16 is the “key stop” for the film I use.
[7] EV 4 to EV 6 at ISO 200.
[8] EV 8 or EV 9 at ISO 200.