Light meter and exposure
- Thread starter xhridhar
- Start date
KurtF
Well-known member
It's an over simplification. Meters give you the correct exposure for an 18% reflective surface (a mid tone gray). But what if you're metering a field of snow? That won't be 18% gray. So you have to adjust from the meters reading based on some common brightness values, testing, and experience.
Dave ©
Well-known member
Assuming that the goal is to expose the shot correctly (ignoring other reasons to intentionally over/under-expose the shot)...
It depends on whether you are using a reflected light meter (e.g. in camera) or an incident light meter (e.g. handheld). When using reflected meters, you may have to adjust or compensate for the reading depending on what metering mode you're using (e.g. evaluative, center-weighted, spot) and what's in the frame (e.g. subject in middle of white snow). You don't typically need to adjust when using an incident meter since it is basically telling you how much light is falling onto your subject.
It depends on whether you are using a reflected light meter (e.g. in camera) or an incident light meter (e.g. handheld). When using reflected meters, you may have to adjust or compensate for the reading depending on what metering mode you're using (e.g. evaluative, center-weighted, spot) and what's in the frame (e.g. subject in middle of white snow). You don't typically need to adjust when using an incident meter since it is basically telling you how much light is falling onto your subject.
sonofaresiii
Well-known member
There's a lot of confusion and misinformation going on in this thread. Do some more research, underexposing by a stop is absolutely not a general rule you want to work by. Don't know where you saw that or why it said to do it, but you misunderstood.
cyclone
Well-known member
Although light meters give you an indication of where 18% grey falls (as noted earlier), many cinematographers intentionally OVEREXPOSE the scene when shooting with negative stock to better saturate the negative, knowing that they will then correct this in the timing of the prints later on in post.
Which brings us to one of the key factors in knowing how to correctly expose in the digital world - how does the particular camera youre shooting on perform in the upper and lower ranges?, what codec is it being shot to? and how will the intended workflow effect all of this?
j1clark@ucsd.edu
Veteran
Such advice, without context, is wrong.
You probably did misinterpret the statement.
Meters are calibrated to allow one to expose 'properly'. What is in the scene, determines what is 'proper'. There are a couple of methods used to determine a proper exposure.
1) incident meter, and meter the general incident light, set camera accordingly. This method does not interpret anything in the scene
2) reflectance mater, and meter a so called 'middle grey' or "18%" grey card, set camera accordingly.
3) Go through a process where one's meter and camera are 'calibrated' to deliver known 'grey' values when presented on a screen, (or in the olden days a print...).
When using a reflectance meter, there is a bit of interpretation required in determine what to meter, to give a proper exposure.
If one has a Kodak 18% grey card, and if one has calibrated one's meter to it, then taking a reading off the grey card, should give a proper exposure.
If one does not have a Kodak 18% grey card, then one has to 'interpret' the scene, take a meter reading from something that is sort of approximately 18% and use that exposure setting...
Or for people photography use the following 'guidelines'.
1) Caucasian skin toned people, open up 1-1.5 stops from the meter reading.
2) Mediterranean skin toned people, use the meter reading directly.
3) Dark skin toned people, close down one stop from the meter reading.
There is some amount of debate as to what value of 'grey' a meter is calibrated to from a 'standards' point of view, other than it is calibrated to an ANSI standard...
So, the principle for a practical calibration is to get a Kodak 18% grey card, and use that to determine a 'personal' ISO number that delivers the desired 'grey' tone.
While that may seem like there will be a lot of variance, everyone having their own 'special' number, it really does work out that if one compares the original grey card to either print or now screen,
one can have a high degree of assurance one will get a 'good result' once the camera, meter, and display system is 'calibrated'.
Think of this like the average mileage listed on a new car sticker... someone, somewhere may have gotten that 30 or 40 miles to the galleon, but for your practical estimates, you need to 'calibrate' to your driving environment and style... and just live with the fact you have a lead foot and only get 15 miles per galleon no matter what you do...
Last edited:
JoeRawlins
Active member
A full stop is too much in my estimation, but coming from an editor/colorist, I like the old photography maxim "Expose for the shadows, develop for the highlights". The general rule of thumb is a white pixel that is truly white = 0 data. so if you are in a situation where you can't control lighting (e.g. landscapes) I would much rather have a shot underexposed half a stop that I can bring up in post, if it means I have 0 blowouts in the highlights. If it's a cloudy day, though, and there's no risk of blowouts, that's another issue.
If you're in an environment where you control the lighting, like on set, i would rather you shoot as the light meter reads and solve your highlights in your lighting, naturally. But even then, if you're going to expose wrong, give it to me half a stop under rather than a half a stop over, every time.
The rule, in my opinion is this:
"If you must err, err on the side of slightly underexposed."
My opinion.
Joe
If you're in an environment where you control the lighting, like on set, i would rather you shoot as the light meter reads and solve your highlights in your lighting, naturally. But even then, if you're going to expose wrong, give it to me half a stop under rather than a half a stop over, every time.
The rule, in my opinion is this:
"If you must err, err on the side of slightly underexposed."
My opinion.
Joe
A lot of the responses above are theoretically based. I set this as a challenge. Using your zebras and histogram, set your exposure. then get out your exposure meter and take a reading. You will most likely find that you have set your camera about 2/3 stop under your exposure meter. Honestly check it, I have been amazed. I'm not saying one is more correct than the other however, just that there is a discrepancy.
My guess is that you need to do the opposite of film (which was: expose for the shadows, dev for the highlights). If you expose to avoid the whites clipping and then look after the blacks in the grade, then you'll be ok.
However, I may be idiosyncratic in my approach since my main approach has been to bring the exposure up to the point of where the whites are at 100 and then bring them back a few percent to keep the detail. That depends on a calibrated meter or a good understanding of the meter, vs sensor. A meter and a histogram are a great combination for achieving this in my opinion.
My guess is that you need to do the opposite of film (which was: expose for the shadows, dev for the highlights). If you expose to avoid the whites clipping and then look after the blacks in the grade, then you'll be ok.
However, I may be idiosyncratic in my approach since my main approach has been to bring the exposure up to the point of where the whites are at 100 and then bring them back a few percent to keep the detail. That depends on a calibrated meter or a good understanding of the meter, vs sensor. A meter and a histogram are a great combination for achieving this in my opinion.
j1clark@ucsd.edu
Veteran
If you look out there, you will find people who claim that the famous Kodak 18% grey card is in fact, about 1/3 to 1/2 'high' relative to where actual light meters are calibrated such that 12% reflectance is in the middle...
And as a note, my GH-1 doesn't have zebras... just a histogram that goes 'white' when the image values 'match' the settings (I use mainly manual settings...).
Dave ©
Well-known member
We're probably violently agreeing but I think most of the responses above are based on both theoretical and practical knowledge and experience. That's why we have communicated collectively to the OP that there is no hard-and-fast rule about how to use a meter. Some stated it explicitly while others communicated it implicitly.
Again, there is a difference in the type of meter used (incident vs reflective) so I think it's important to mention the type of meter to which you are referring when making comparisons.
It is also important to recognize that there are different metering modes, which influence the readings dramatically.
I think we are agreeing and as far as you say in summing up -that's very true
I think a meter can be a complete waste of time if you aren't careful (mind you, you'll never learn what application it has without wasting the time eh?).
I had a very rough and ready chroma to do the other day (not enough lights, unsuitable lamps) and I found it really handy to quickly use my spot meter to check the evenness of the light and balance the light on the face and the light level on the chroma.
old school Dop's who light by the meter use them in all sorts of ways for which they weren't designed and they certainly get the shots!
gonzo_entertainment
Veteran
Yep.
The 5D is excellent in the shadows, but is terrible (like most digital) in the highlights. So you expose for the highlights even if it means going under a little on the shadows.
starcentral
Veteran
Hi xhridhar,
If you read your source again, the book is most likely referring to how much light you should use to expose the skin on your subject. What they are likely referring to is the idea to expose skin 1 stop under what is required to expose middle grey on your camera. For digital, middle grey is 50% on your camera. So they are suggesting to expose your skin 1 stop less at about 40%. On Barry's lighting DVD they suggest exposing skin at 60-70 IRE depending on the subjects skin color but in the end the level you expose it at is all based on artistic preference. I think of 40-50 to be more film like, and 60-70 more broadcast TV like.
If you don't have a light meter or have no plans of using one you can use a waveform monitor sometimes built into your camera or external monitor to adjust your lighting until you get 40% on your subject. If you want to use a lightmeter instead you can determine the working ASA of your camera with a grey card and this way start using your light meter to determine light levels. One caution though is you are metering for backgrounds, cars, or other things you have to take in account the materials lightness or darkness and even reflectivity, etc.. so a lightmeter is not always clear cut whereas the waveform monitor is exactly what the camera sees and records. For example if you metered your light at f4 then shot the blue sky or snow covered ground it would likely be over exposed by 1-2 stops. You have to know to compensate your meter reading based on your subject, and in that example instead of using f4 you might go to f5.6 or f8 on your iris, or keep it at f4 but add a 1 stop ND filter.
I personally like a lightmeter because it can be used to for pre-scouting a location and once you learn to work with it and trust it you can start lighting a scene before your talent even gets there, and also you don't have to worry about having a camera there to judge your lighting.
There are a lot of mixed feelings here in these forums on how to get exposure, but your best bet is to try different approaches and determine what works best for you. A lot of my work is graded in post so I always shoot to get consistent exposure on a subject and background as much as possible to make the post process faster and easier. ie. if all you ever do is expose for your scene and not the subject then I guarantee when your footage is edited there will be inconsistencies between your wide, medium and even close up shots.
Bottom line is if you don't plan on lighting your scenes, or just making use of ambient light then go ahead and forget the lightmeter and use your camera to tell you the best exposure for the scene. If you want to shoot using controlled lighting so shots are repeatable between takes, setups, or even shot over multiple days then a lightmeter will be your best friend.
Here's a great thread I wrote in the F3 section on how I've been using my lightmeter specifically to gauge exposure on a variety of projects recently: http://www.dvxuser.com/V6/showthrea...se-with-a-f3&p=2506068&viewfull=1#post2506068