I think I see the light (meter): how to buy one

With the stumbling self-taught revival of film photography, there does not seem to be any really simple, practical treatment of how light meters are supposed to work or why you might want one type over another. This article will probably not be that treatment, but why not try? It will definitely have information density; you can read this in 12 minutes and save yourself the trouble of watching 200 Youtube videos. And hopefully minimize the number of missteps in selecting a meter.

Do I need a light meter? Maybe not.

If a light meter is to serve any purpose other than being neck jewelry, there must be a use case. If your camera has a working internal meter, there is probably no situation, for hand-holdable light levels, in which a handheld meter is going to add a lot. Well, as against a well-tuned, in-camera meter and experience in using it, using a handheld may add a lot of delay and inconvenience.

If your camera lacks a light meter, you need (i) a meter or (ii) enough practice not to need one. Film is not free, and making pictures takes a lot of time.

In the ancient days of rollfilm, people used very rudimentary methods to measure light: experimentation and extinction meters. An extinction meter was a piece of developed film (or printed acetate) a black background and numbers of progressively decreasing transparency. As the numbers got higher numerically, the density of the film was higher. So when you looked at it, you would see that say “7” was the highest number you could see (i.e., light would pass through the number), you would use that to compute the exposure. This was still inaccurate because the effectiveness of extinction meters depends on the type of scene and human eyesight. So people experimented. All of this was totally ok because black and white film was insensitive to red light – so you could “develop by inspection,” in other words, watch for the highlights (darkest parts of the negative) to become visible. So an extinction meter might be good enough.

That was if you were a professional. Mainstream amateur cameras lacked the exposure control to make metering useful. The No. 2 Brownie camera – the first thing to use 120 film – had a fixed aperture and shutter speed, exposure control basically demanded only that the camera be used outside in bright sun, with the sun shining on the subject. So developing by inspection (and printing as best you could) was the only way to go. You can still do this with ortho film. Pictures were remarkably good.

With progressively more sophisticated folding cameras came variable apertures (like f/6.3-16) and some need to tell one lighting condition from the next. This could still be done with extinction meters or settings based on general light conditions (see the “Sunny-16” rule below).

Two things put more pressure on the development of light meters. One was panchromatic black-and-white film, which has to be developed blind because it can see any color of visible light. The other was color transparency film that in comparison to b/w negative was fantastically intolerant of exposure errors.

Light meters were not always common for amateurs, who often read suggested exposure setting off the inside of a film box. The instructions expressed the “Sunny-16” rule. It seems simplistic, but you realize, after metering thousands of different situations and noting the results, that the “Sunny-16” rule is right almost all of the time in outdoor situations. The Sunny-16 rule looks at the lighting condition and suggests an exposure:

1. Bright sun, distinct shadows – 1/ISO shutter speed + f/16 (so with 100-speed film, 1/125 and f/16)
2. Bright sun, soft shadows/partly sunny – f/11
3. Bright overcast – f/8
4. Medium overcast – f/5.6
5. Sun setting – f/4

…And outdoors, there may be little point beyond that unless you have a camera on a tripod.

There is a reason why the Sunny-16 rule was printed inside Kodachrome boxes: it works. So the box says that these are “starting points,” but depending on your needs, they may also become “ending points.”

What about my phone?

No matter how cute a phone app is for light metering, it combines the elements of using a second camera, taking a spot reading and not an average, being a handheld device with poor haptic feedback, having a tendency to lock itself when least convenient, and exhibiting a tendency to be hard to read in bright light. By contrast, the Sekonic shown at the top of this article has no pretty graphics, no A13 processor, and no color. You press the button to take a reading, and you simply spin the thumb wheel to see all of the equivalent exposure combinations. It is ergonomically simple and supremely functional. Of course, Sekonic had to get rid of it in favor of an L-478 that looks and works like an iPhone. Because some executive decided that it was what the “kids” wanted.

Cell technology

There are essentially three meter cell technologies.

Selenium. Meter cells made of this bioaccumulating metalloid emit electricity when exposed to light. Meters with these cells need no batteries. Selenium cells degrade over time with exposure to light (and more commonly, moisture), but when they are in their prime, they are bulletproof. Selenium cells have poor low-light accuracy but have very good accuracy across different color temperatures. Selenium cells are sensitive to EV +4. Don’t worry. That’s fairly low light.

Cadmium Sulfide (CdS). These cells were used in most electronic cameras well into the 1980s. CdS cells require a 1.35v input voltage, which previously was supplied by long-lasting mercury cells that kept the same voltage until they died. Many CdS meters lack voltage regulation, and sticking a 1.5 alkaline or 1.55v silver battery in one is a great way to experience underexposure. Most CdS cells have experienced significant degradation due to chemical breakdown. They lose their resistance, pass more current, and make the meter think there is more light than there actually is. Many CdS meters will need calibration, no matter what the eBay seller/photo pimp says. CdS cells do not do as well as selenium in incandescent light. CdS and other modern cells can meter to EV 0 or even lower.

Silicon blue cell or silicon photo diode (SBC or SPD). These operate like CdS cells, except that almost every meter or camera that has these takes modern batteries and has voltage regulation. This is the current state of the art in handheld meters. These also have issues with incandescent light, so bracket.

Types of electric/electronic meters

There are three types of electric/electronic meters for ambient light. This guide will ignore flash metering, which is simply using any of these types of meters to measure an instantaneous light impulse rather than a continuous one. Flash metering is important for studios, but for most handheld uses, people most commonly rely on a flash’s internal metering or TTL flash provided by the camera.

Reflected light meters. Most handheld meters are this type. They generally accept light from 10 to 30-degree angle, average some or all of a scene, and try to expose that scene for middle grey. For example, if you pointed a reflected meter at an all-black scene, it would attempt to make the whole scene 18% grey. Same thing if you pointed it at a white wall.

If your scene is perfectly average mix of light and dark, or all concrete or limestone, this type of meter works without any thought at all. If not, you need to practice to understand how much brighter or darker than middle grey various objects might be.

Reflected light meters were used by Ansel Adams and pretty much anyone who needed to measure the brightness range of a scene. Point the meter at the dark parts. Take a reading. Point it at the high. Take a reading. Average or compute the scale. This type of meter (and preferably a spot meter, below) is absolutely essential for the Zone System.

Spot meters. These are a species of reflected light meter that typically accepts from 1 to 5 degrees. 1 degree sounds small, but on a distant subject, it can be quite big. Spot meters help you sort out the various tones in a scene so that you can spend 1000x the effort to get a picture that is 10% better than an averaging reflective meter used correctly. This is being cynical, but spot meters, like communism, seem like a great idea until you try to use them on an everyday basis. You pretty much always have to take more than one reading of a scene, unless you use a grey card to cheat (and at that point, you might as well use a cheaper reflected or incident meter). For landscapes or slow-moving subjects, careful spot metering is workable and with appropriate finishing can result in brilliant output. For everything else, it can seem punishing.

Incident meters. Incident meters are used by people with many skill levels and personal creeds, but they are especially popular with three groups of people: (1) cinematographers who cannot afford to make exposure mistakes; (2) people who think that incident light measurement is somehow a universal shortcut and are willing to clean up any unexpected consequences in post; and (3) people who deal with complex flash or studio lighting situations. Incident meters are very helpful, but read on to understand why you may not want to buy a meter that does only incident. One important thing to understand about incident metering is that the Sunny-16 rule is essentially a rough-cut system of incident metering: it cares about what the light is, not what the subject is.

Incident vs. reflected

The choice of incident versus reflected metering seems to be the subject of much ignorance confusion. You might have noticed that most older meters regard incident metering as an afterthought; the sliding white domes are tiny. Incident light measurement is not something most people needed with negative film. It has its ups and downs.

You’re not fooling me with that sheepskin! Incident measurements care about the light source and assume a normal range of reflectivity for a scene: no Vantablack and no white phosphorous. Dark objects look dark; light objects look light. Exposure is based on the expected reflection by the subject of a particular quantity of light. The meter is not fooled by the color of the object or the reflectivity of any particular object.

Reflected metering measures light falling on a scene (meter pointed at the subject) and assumes that it is an “average” scene, which is one where the actual reading is normalized to that 18% brightness.

What’s the difference? Consider backlighting.

• The incident meter will not be fooled by the bright backlight because it is pointed toward the camera position.
• The reflected meter may read two stops low if it averages the entire scene.

Or consider bright overcast where you expect an exposure of f/8.

• The incident meter will not be fooled because the huge bright sky is out of its field of view – and it is the light source.
• The reflected meter will underexpose if there too much sky in the measurement (reading f/16, for example) and will overexpose the sky if the meter is pointed toward the ground (reading f/4).

Tonal range. But that ability to read (or misread) parts of the scene is what makes the reflected meter useful. Incident meters are blind to what the subject actually is. This is the flip side of not being fooled by subject reflectivity. Incident metering tells you nothing about how you might need to push or pull developing. Pushing and pulling is not such a big part of color film photography, and here, an incident meter and a reflected meter are very similarly capable.

Sightlines. One requirement of successful incident metering is that if you cannot stand at the subject, you at least need to be able to replicate the subject’s sight line to the light source. For example, if you are standing inside a dark structure taking a picture out a window, an incident reading is going to see a dark “sky” and force a very bright exposure. If, on the other hand, you can walk up to a subject, it tends to work better. You can approximate an incident reading at the subject by standing somewhere in which the meter cell can receive approximately the same light that the subject is, for example, on the roof of said building.

The approximation works well outside because the difference between your position and the subject’s position, given the brightness and distance of the sun, is insignificant. With the sun 150 million kilometers away, what’s another 100 meters between you and the subject? Nothing for purposes of light metering. For artificial light, which is far weaker and less uniform, you have to be at the subject for incident metering to work correctly. This is due to the inverse square law: doubling the distance from the light quarters its brightness. This is a factor when the difference is 3m or 6m from a 60w light; it is not a factor for natural light, which is literally a hundred million miles away.

Reflected metering can be done anytime you can see the subject with your eyes.

Splitting the difference. So why weren’t incident meters more popular over history? Because taking a reflected reading of an 18% grey card lit the same way as the subject gives you the same reading as an incident meter. This trick can be used even with spot meters that have no incident capability at all. One of these cards can fit in your pocket.

Another trick. But wait! You don’t actually need a grey card if you know the reflectivity of your own skin. For example, if you meter a grey card at f/8 and your skin at 11, then you can assume for the future that reading a reflected meter off your hand and adding a stop (or reading your skin with half the ISO) gets you to middle grey. This works for every skin tone, so long as you know the offset. Measure off the palm of your hand. Unless you spray-tan your palms, this is a very stable reference point. This trick also works if your reflected meter is the one in your camera. Fill the meter’s field of view with your hand, maintaining the same lighting as the subject, i.e., don’t shadow your hand with the camera.

Reading your meter

Meters have various ways of displaying the exposure.

Direct read (shutter/aperture – digital): a digital display shows a shutter/aperture pairing. This is best if you shoot predominantly with one shutter speed or aperture and let the other variable float. Otherwise, direct-read meter displays fail the human-machine interface test for the same reason digital watches do – humans for some reason are much better at reading graphic displays than numerical ones. That’s also why digital speedometers are disfavored. A direct-read digital display also requires you to scroll through the combinations if you switch up apertures, for example, between shots. Direct-read does have the advantage of being able to display an averaged reading instantly if your meter supports that. Some also can show you what percentage of an exposure is ambient and what percentage is coming from flash.

Some of these meters have an analog-looking scale at the bottom that can be used to show the spread of multiple readings and their average. Not all of these things show on the dial at once; this illustration just shows how many things a Sekonic digital would tell you, depending on mode.

Many of these meters will also show exposure value (EVs), which correspond to pairs of shutter speeds/aperture settings.

Direct Read (f/stop, analog): a few older meters do this, like the Weston 853 and one new one (the Sekonic L-398). These involve setting your shutter speed and having the needle swing to the right f/stop for the light level. This is very similar to how shutter-priority autoexposure works in SLRs. This type of reading requires something to moderate the light compared to account for different shutter speeds; the 853, for example, has a rotating mask that covers part of the photocell to simulate faster shutter speeds.

Light intensity (exposure value/footcandles): here, a needle points to a number (EVs, footcandles, etc., sometimes both!), or an LED shows a number, and you rotate the calculator dial to that number (having previously rotated a pointer to your ISO value), and you see the shutter/aperture pairings on a ring or a slide-rule. Most selenium meters work this way; few selenium meters lock the needle. Below is the Sekonic L-398 (selenium); the other all-time greats for light-intensity meters were the Gossen Luna Pro S (or Lunasix) (CdS) and the Weston Master series (selenium). They require computation using a calculator wheel to get to a camera setting. The L-398 below includes both direct read for f/stops and light intensity (footcandle) readings:

For reference, this is a table of EVs (Exposure Values) from Wikipedia.

As you can see, EV +4, where selenium meters mostly peter out, is not even remotely hand-holdable. And this version of the chart is helpful because it goes down to -6; your AF SLR maker may claim the ability to AF down to EV -4 (minus 4), but you would not be able to see much yourself in that kind of light. EV +15 is about the highest light level most people would encounter in everyday life, and EV +9 would be the darkest that most people would take pictures. EV +21 might be expressing a nuclear explosion; most tables don’t go this high. Like a lot of things, there is quite a bit of math behind how this works but for operational purposes, it’s not important. EVs are just one of the ways that meters read.

One thing that is fantastically confusing about EVs is that EV is typically expressed for ISO 100 when cited in camera specs. This is just like Guide Numbers for flash, which are also always expressed at their ISO 100 values. EV +15 for 100-speed film is actually EV +17 for 400-speed film. This is why most handheld meters show the ISO-corrected EV on a calculator dial – so it can change with ISO. Some meters don’t show the actual EV corrected for the ISO – for example, the Pentax Digital Spotmeter’s viewfinder display only shows the EV for ISO 100, and that number, combined with the ISO setting gives you all the pairs of camera settings, not the actual EV.

Match needle: meter needle moves (no scale), and you turn the calculator dial so that the two pointers match (the pointer moves when you turn the calculator dial). Then you read the same type of scale as on an light-intensity meter. This is far, far, faster to operate than a light-intensity meter. All you do is match the Meßwerkzeiger (German for “meter needle”) with the Nachführzeiger (German for “ring thingy…” actually, “tracking pointer”) by turning the calculator dial (the thing with shutter speed/aperture combinations). Below is a Gossen Super Pilot (or “Sixtar” as they called it in Europe and “Systar” as they call it in Lego) (ok, bad joke…). Match needle probably gets the highest score for human-machine interface because you don’t have to read and comprehend a number, remember it, and then transfer it to the calculator dial.

Null: basically a variation of match needle (and on a camera, you would actually call it “match needle”), but the thing you are doing is making the needle hit a fixed zero mark by turning the calculator dial. In practice, this is simpler. This is better than a match-needle for measuring brightness range. The Luna-Pro SBC is one of the few good meters that have this.

Filters

Few things are more complicated exposure-wise than dealing with the use of filters in metering. There is no use in reinventing this wheel. Go to this article on the old site, and it will tell you what you need to know about exposure compensation. Maybe not everything, but there is a lot of data.

Good and bad meters

Well, you were going to ask for recommendations, right? How about some snarky short-takes? This does not cover every possible meter, but it will cover the ones you commonly see.

Note: this does not cover any of the many types of electronic meters designed solely to fit in a flash shoe; those have too much interaction with the ergonomics of a particular camera to make even somewhat accurate snap judgments.

1. General Electric DW-58: a perennial Bakelite favorite at yard sales and one of the most popular meters ever made, this 1940s dual-range light-intensity-style meter (reads in foot-candles) has a grille to change from low to high range. It has a calculator dial from which to compute exposure combinations. You can also take the snout off for incident readings – though it is not a dome – it’s just a flat cell.
2. Gossen Digi-Six: tiny lightweight digital light intensity meter that also fits in a flash shoe. Has a thumb wheel to compute aperture. Crude but tough and effective. Competitor to the Sekonic Twinmate.
3. Gossen Luna-Pro S (CdS): the king of the old school, this meter is awesome, assuming yours is calibrated correctly and you use the right batteries. You read a numerical value from a low or high range and set that into the calculator dial (which in turn will compute actual EV). The meters are cheap used; factory calibration costs megabucks; some independents will do it too, but you know it’s been done right if the top of the low range matches the bottom of the high range. Most (maybe all) use 1.35v cells unless they have adapters or have been converted. This can be kludged with attachments into measuring anything but flash. There are probes, domes, waist-level finders, you name it.
4. Gossen Luna-Pro SBC (silicon): This Luna-Pro variant is larger and uses a null dial instead of starting with an arbitrary light value (you can read the EV off the calculator dial). It also fixes the one thing that challenges the Luna-Pro S: it uses a modern photocell. The Luna Pro SBC can take a flash measuring attachment that is best ignored in an era when flash meters are made much better.
5. Gossen Super-Pilot (CdS): this is basically a baby Luna-Pro with a match-needle.
6. Gossen Scout 2: Cheap, functional, selenium.
7. Gossen digital meters (silicon): all of them are good. User interface is about as good as Minolta digital, which is not that good. Dieter Rams and the Braun design philosophy do not translate well to light meters.
8. Metered Light Pocket Spot (silicon): interesting little handheld meter that uses a hole through the meter as an aiming reticle. Built of anodized aluminum and tough!
9. Metrawatt Metrastar (CdS): The coolest-looking CdS meter ever made, this one has a tiny waist-level viewfinder built into the dial.
10. Minolta digital meters (silicon): all of them are good. User interface is about as good as Gossen digital, which is not that good.
11. Pentax Digital Spotmeter (silicon): many of these are not actually Zone VI modified (even if they have the sticker with the grey squares – you need the main label). The holster-like Zone VI case could get you shot at a traffic stop. It takes 40.5mm filters. It is accurate, durable, and expensive. Unless you become a Zone System afficionado, you will use it twice, stick it in a drawer, think you need it for about two years, and then sell it to the next guy on eBay.
12. Sekonic L-208 Twinmate (silicon): what a cute widdle Luna Pwo! Tiny match needle meter that also fits in a flash shoe and runs on a lithium coin cell.
13. Sekonic L-308: basic incident and reflected meter. Classic functionality. Runs on one AA. The only inconvenient thing is that it has up/down buttons instead of a thumb wheel. Electronically, it’s a baby L-358 that is more versatile out of the box. But its one big vice for reflected metering is that the meter cell is on the same side as the display.
14. Sekonic L-358 (silicon): with a reflected or spot attachment, it’s great. With an incident dome, it’s a great studio tool, but I suspect that most people buy this configuration by accident. The L-358 with the Lumigrid reflected meter lens is ace.
15. Sekonic L-398 (selenium): the king of selenium meters by default, this studio favorite has been in production longer than most readers here have been alive.
16. Sekonic L-438 (silicon): a cross between Luke Skywalker’s binoculars and a 110 camera, this pocket spotmeter runs on a single AA cell and works like a baby camera. The novelty wears off quickly. No incident or reflected capability, so it’s a one-trick pony.
17. Sekonic L-478D (silicon): the supposed replacement for the L-358, this is extremely competent but ergonomically poor.
18. Sekonic L-558 (silicon): imagine an L-358 that also includes a viewfinder for taking reflected measurements. Replaced by the L-858D, which is the touchscreen version.
19. Shepherd/Smith Victor DM-170 Light Meter (silicon): solid and serviceable, it has a bright 80s-style LED light-intensity display (some displays are red, some are a cool green). Runs on a 9v battery. Dependable.
20. Soligor digital spot meter (CdS): this will not get you killed at a traffic stop, but it will get you killed if Kirk and Spock have already drawn down with phasers.
21. Vivitar 45 CdS meter (CdS, of course!): cheap, small, match needle.
22. Weston Master series (selenium). These used to be the top dog before the Luna-Pro displaced them. These measure light intensity.

Conclusion

Meters are subject that is both simple and complex. The best way to choose a meter is to borrow one and see if you can get into the rhythm of using it the way it is designed. I am a big fan of match-needle meters, but having used them all, there is no truly bad design. The bigger question is how you will meter (incident? reflected? spot?), and you may not learn what works best for you on the first try. So take your best guess, try it, and if it doesn’t work, try again!