What is the downside to using an incident light meter to set my exposure for video? I've been doing it for about a month and it has been working well for me so far; but I'm a rookie—those of you with much more experience—what can go wrong in video when measuring light with an incident meter rather than using the various camera exposure tools? I'm mainly concerned with ensuring that faces look the same from shot to shot.
The problem I'm trying to solve is that my style generally includes a fair bit of back and edge lighting which often freaks out the waveform monitor, the RGB parade, and the histogram so I have been getting inconsistent brightness levels.
Using a light meter is also quicker because I can adjust lights and get a reading immediately rather than running back and forth to the monitor to see what has changed.
If it works for you, then keep doing it. Just make sure to follow your footage through post to see if you need to tweak your exposure method. For example, I find that when I use my meter, I tend to expose more brightly than I intend other than for shots where the face is clearly in a strong key light at full exposure.
Just keep in mind that your exposure technique is not "the shot", it's just one way of getting towards your shot, so if you don't spend time actually looking at the results, you aren't taking advantage of one of the main benefits of shooting digitally, which is seeing the results on set, live. There's no reward for being good at using a light meter, only in the images you create, so don't make it about the process more than the results.
That said, occasionally you don't always have a good reference monitor set-up, so knowing that your metering technique is accurate is useful.
Keep in mind that incident meters aren't going to tell you about what areas of the frame are clipping.
"Love Witch" was shot on film so I had to use my meter, but I also took some reference stills with my digital camera just to get a rough overall sense of the balance since the light levels were so high on set.
I used an incident meter to measure the key and often the fill (for a master shot) then balanced by eye, same goes for fill on coverage, once I had the look set in the wide. Sometimes I'd go around and meter the walls since I was trying to spotlight those separately from the faces. I know from past experience that 3-stops under for fill would give me shadow detail without things looking flat.
As for bright overexposed highlights, I had more problems creating them than I did with worrying about things blowing out -- generally on film, I don't worry too much about losing highlight detail.
At 100 foot-candles for an f/2.8 on 200 ASA film rated at 100 ASA, getting things to fall-off to black wasn't the hard part.
Richard.. it may seem obvious but you didn't mention using zebras in your first post.. most cameras these days you can set two or more levels with different "hatch" or color .. rather than histograms and parades these are a really basic and easy way to judge exposure in your shot.. there is usually one set to 100%..and above. to show your clipping and then the other set for skin tones.. (or any other level that you want.). often you can set a window also for the measuring levels from 10% to 1%.. e.g. set for 70 at 10% will read from 65-75.. but set to 1% will be just 70 exactly..
Personally I would never have then on the whole time,although some people do.. but have them on an assign switch to quickly check.. false color does pretty much the same thing.. which can often be set on monitors.. but zebras AFAIK are only generated in the EVF.. but a very handy tool.. and pretty much all you need really..
Edited by Robin R Probyn, 25 June 2017 - 06:37 PM.
Zebras is a great idea. It's probably an issue of experience—I need to learn to trust them. All I really want out of life is for faces to be the same implied brightness from shot to shot. Sort of like having a reference white and black in every shot.
I'll give zebras a try and see how it goes, and you're right about switching them off—they are fairly distracting when used throughout.
Lighting purely off an incident meter for video is admirable, but there is a breath of caution:
If your meter isn't calibrated, then it's giving you a false reading. However, unlike film, you'll notice on set that something is gravely off when shooting digital.
I always recommend checking your meter with the camera during prep. If it's off, you can compensate for it and still use it.
The waveform, histogram, zebras, etc are all just tools designed to measure exposure. Like David said, there's no reward for using an incident meter. The reward is effectively using any and all tools at your disposal to get the exposure you want.
I shoot with a Panasonic GH5, and most stuff is shot in V-log external to 10-bit 4:2:2 on the Atomos. I still have a light meter, but I rarely use it. The Atomos has great built in scopes, zebra patterns, etc - that really make it much easier to get proper exposure from. Most higher-end cameras have these scopes and patterns built in as well. Plus it helps you can actually see what is being recorded. My fear with relying only on a light meter is that I'm gonna 'eff something up, and it can't be recovered in post. As others have said, film is a lot more forgiving (I have only ever shot still film, but its the same difference - I still use a light meter for stills). When you are exposing video - especially some flat profiles like unprocessed or vlog/slog/, you're really running with a fine margin of error on exposure - even more so if your shooting to a 4:2:0 and/or 8-bit format, where noise, artifacts, and banding will come into play.
Recently, I have been shooting some promos for a personal film project using a lot of green screen - and I find that the camera and/or recorder based exposure scopes and zebra's are a lifesaver in this area - makes sure your green screen does not have any dark zones that will interfere with keying.
If you have the experience with a light meter, and you are happy using it, and you are getting good results from it - there is little reason to switch your work flow. I only caution to make sure your playback looks good on tricky shots. That is one of the great advantages of shooting on video formats - you get instant gratification of seeing what your final output will be. It kind of defeats that advantage to ignore that in my opinion.
Yes agree.. if the metre is telling you one thing.. but the EVF and a decent monitor is really dark or totally blown out.. and that isn't the effect you want !!.. I would reset to the monitor/EVF.. that would be like driving over cliff because the sat nav says go straight..
I am passionate about light meters, and I think that using a properly calibrated meter, properly calibrated meter for the camera or LUT that you are using is the best way to work, but you need to know what you are doing. It would be helpful to know which camera you are using, since you refer to it just as a video camera.
If you are using a good entry level cinema camera, or even better a good cinema camera you can treat it as a pro camera, as long as you have your facts straight. You will need to know how to use a light meter, your cameras dynamic range and the zone system... But you should not assume that your meter knows where your camera is placing middle gray. You need to test the camera, the meter and the LUT's. Would I shoot a scene that looks way over exposed in the EVF because I'm following the meter information? If I know my numbers, my camera's limitations and my meter, yes I would. Why? Because I know that I can bring it back in post, and if I do it that way I would get less noise or grain compared to shooting an underexposed scene just because the EVF tells me that before any post-processing it looks right in the camera. That's just an example.
You should not assume that your camera is properly calibrated with the meter. They don't know each other, they are calibrated in their own world, and not all manufacturers use 18% gray, You have to test your camera and compare it to the light meter to know where they meet and compensate for the difference.
I'm sure you have taken pictures and you have noticed that if you select a different profile in your Nikon camera (I shoot Nikon) the pictures look different than other profiles. Let's say you take a picture on "Standard", one on "Neutral" and another one using the "vivid" profile. The "vivid" in the Nikon example is going to look more colorful, more contrasty and you are going to loose detail on shadows and highlights faster because it is more contrasty than the other ones, but that doesn't mean that the sensor is not capable of capturing more detail on those areas, so learn about your cameras dynamic range.
It's about perception. If you look at a LCD screen at the desert in the middle of the day your perception of "properly exposed" is going to be different than if you look at the same LCD screen or EVF at night in a darkroom. A light meter bypasses that factor and gives you information that allows you to pre visualize the final image in your mind.
Zebras, false color and wave form scopes will give you information as well. They are great tools and all of them are telling you information about brightness on the scene, but you need to be in front of the monitor and looking at the scopes in order to get that information. Test your camera and scopes and find common ground between the scopes and the meter, that way you can be next to your subject, getting info straight from the light meter, translating that information on your mind and knowing how it is going to look at the end, when you apply a LUT or color grading.
If you know the limitations of your camera and you adjust the setting inside and outside the camera (lighting) to make everything fit into your camera's dynamic range, you'll make any DSLR look like a Red.
It's about perception. If you look at a LCD screen at the desert in the middle of the day your perception of "properly exposed" is going to be different than if you look at the same LCD screen or EVF at night in a darkroom. A light meter bypasses that factor
But so does a histogram. And a lot of cameras have colour histograms, at least for stills review, which is quite adequate for setting exposure. Your light can look fine on a meter but you can still be blowing a channel and polluting highlights. Probably the best way to make any digital sensor look "like a Red" is to use ETTR and build your lighting around getting everything you care about in the top 4 stops of DR. But a meter won't give you adequate information to do so optimally.
(Warning: I'm extrapolating from stills experience! In particular I often shot bayer and foveon sensors on the same shoots, which made for both tricky metering problems and a unique opportunity to see what happens to highlights if a channel blows.)
Edited by David Mawson, 14 September 2017 - 09:53 AM.