Light Meter to use with a f3

(I'm speaking strictly from a film-look perspective) I quite disagree with the approach of using zebras as an effective way to judge your exposure to any capacity.

Using zebras comes from the videocam days where you didn't want any portion of your image to clip as to preserve as much of the range in the image as possible. Those cameras had (and some still do) a limited dynamic range so going by your zebras was something that made sense to watch out for. Step into the modern world and cameras are now offering 6-7 stops below middle grey and 6-7 stops above.

If you use your zebras to expose for highlights in a scene then you are letting the rest of your scene fall where it may. If you use zebra on skin, how do you know exactly which part of the face to watch out for strips on when you might have high contrast lighting or lighting that isn't perfectly flat? The Zebra approach might be fine for a single interview or ENG type of work where you setup a shot and hardly move, but go shoot a music video or film with multiple scenes, setups, locations, days, lighting conditions and you will end up with nothing but inconsistent footage when its cut and your subject(s) skin levels are all over the place. Expect to spend a lot of time in post correcting this.

Instead of zebras, use a waveform monitor (too bad none on the F3) and failing time/money for an ext. device just use the marker function built into the F3 and aim it at your subjects face/skin for an IRE readout and use it to get consistent skin exposure between all your shots. Balance your scene lighting accordingly by adding or subtracting light from it. Use zebras to watch for 'unwanted' clipping but remember in 'real life' its okay for hot sources like bulbs to clip, or maybe you want an area to clip slightly like the view through a window to the outdoors.

Here's a guideline I use for exposing my subjects relative to middle grey being 50 (REC709 and Cine 3):

White or light skin: IRE 30-40
Light dark skin IRE 40-50
Very dark skin IRE 50-60

When you shoot in other cinegamma modes middle grey areas shift up or down which can cause you to over or under expose the image so you must factor this in when exposing for skintones relative to standard middle grey. ie. with Cine 1 (useful for bright sunny day or high contrast scene) middle grey goes down to 40 so you can place your 'white or light skin' between 20-30 (I'd go for 30).

Some other tips about cinegamma modes: Cine 2 is broadcast legal and clips over 100. Cine 1, 3 and 4 record to about 110. REC709 goes to 105.

Have a look at this guide I put together which shows you the actual gamma curves and where middle grey is located: http://www.starcentral.ca/Sony PMW.pdf (info courtesy of Abelcine)
 
I quite disagree with the previous post. Zebras are an excellent way of judging your exposure and can be used for much more than just to prevent clipping of whites. Anyone who thinks that's all zebras are good for doesn't really understand them. And their use has nothing at all to do with the dynamic range of the camera. It doesn't matter if the camera has 2 stops of dynamic range or 15, exposure is still exposure. Even if you want to have a very conservative exposure (such as when using S-LOG) zebras are still the easiest and most accurate method of setting the exposure unless you want to drag around a waveform monitor. Zebras are like audio VU meters that show you the intensity of light instead of sound, but that doesn't mean you have to run them up to the red line all the time. With audio, sometimes you may want to be conservative and adjust your levels to consistently peak at a lower level to avoid any chance of clipping. Well, you can use zebras to do the same thing for exposure. Conservative, but consistent.

However, I do agree that it is quite all right for some subjects, such as light bulbs and specular highlights to clip. Zebras should be used to judge whites, not highlights.


Here's a guideline I use for exposing my subjects relative to middle grey being 50 (REC709 and Cine 3):

White or light skin: IRE 30-40
Light dark skin IRE 40-50
Very dark skin IRE 50-60

Can you please explain this approach? It doesn't make sense to me. You are exposing dark skin at a higher level than white skin? That makes no sense at all. Why would you expose any dark object at a higher IRE than a ligher object in the same lighting? Also, what if you have a light and a dark skinned person in the same shot? What are you going to do then? How are you ever going to get the white skin at 30-40 and the dark skin at 50-60. Impossible. This approach makes no sense at all.

On the other hand, if I correctly expose for white or a 18% gray card (whether judging by zebras, waveform, or light meter), then it doesn't matter what shade someone's skin is, because the exposure is going to be correct and everything else will fall into place. For example, If f/2.8 is correct for a white person, then it is going to be correct for a black person. Will the black person look darker? Yup, because they are darker!! You don't want to make everyone look the same. I would never adjust the exposure based on skin tone. This is just basic 101 incident light reading that has been around for more than a hundred years. Judging exposure via a reflected reading from any surface (skin in this case) of unknown luminance is a bad habit. Bright white, 18% gray, or a chip chart are the only conistent surfaces that I can trust.
 
Last edited:
The problem is by using zebras on a grey card, just how much of an area should be covered in stripes to know exactly if its at the right level or not? Simply changing the angle of the grey card to the camera will change the amount of stipes shown. My previous post was mainly to illustrate the inconsistencies you would end up with from shot to shot if you continually exposed for highlights using zebras.

As for using zebra stripes on a grey card, with the F3 why wouldn't you use the far more accurate built-in marker function (page 45/46 of manual) to show you the brightness level of what the center of the camera is pointed at? It gives you the same data as a waveform monitor except will only show you the value of dead camera center, not the whole scene.

That brings me to one last point tying into the original point of this thread; if you adjust your lighting so you can get 50 on your grey card (using whatever method makes you happy like zebra, marker, waveform) wouldn't it make sense that once you get 50 you take a meter reading with your light meter so you can use your light meter from now on to get 50 in your camera?

If you set your light meter to ISO800, 1/48th, 24fps and your F3 is at 0db, 1/48th, 24fps, REC709, without any ND filters engaged and you take a reading, this value, for example say 'F4' is exactly how much light you need to get middle grey exposed to 50 in your camera.
 
There is no one way to expose for video. Judgement is the most important factor. I find lightmeters unreliable becaus if I change my set-up the response will be different, but all tools whether zebra, waveform, histogram light meter, monitor etc have to be used with discretion. Using 100% zebras with cinegammas can really screw you up though.
 
I guess I'll weigh in on this as I'm prepping extensively for a documentary, an do a lot of interviews.
What I've found (and I use every available tool-Wave form, Vectorscope, Grey Card, light meter...) is that once you decide where you want flesh tones to fall, that value applies to ALL flesh tones.

So if you want a Caucasian to expose at 45 IRE, you will also want the darker skit tones at 45 IRE. Of course that means opening up your lens to compensate.
This is easy when only one person is in the frame, and obviously a challenge to balance when you have a light person next to a darker person but that's nothing new.

I shoot the BET awards every year, it works for me : )
 
Bill, you might satisfied by the results you get by trying to make everyone's skin tones look the same, but in your example, I guarantee you that either the darker or lighter skin tones are not exposed correctly if you try to put them both at the same IRE. Different people have different skintones, and in my experience, the exposure should not try to equalize them out and make everyone look the same. Maybe the difference is close enough that it doesn't matter very much, but nevertheless, they both can't be correct.

For example, If I setup an interview and expose it perfectly for a white person, and then that person gets up and is replaced by a black person, my exposure will remain exactly the same. Nothing has changed with the intensity of the lighting, so why would I change my exposure? I might adjust the lighting in other ways to fit the individual, but the exposure would be the same. Here's another way of looking at it. If both the white and dark person were on camera together, and I exposed the camera perfectly, I would not change the exposure if one of the two people left the frame.

That's the beauty of basing the exposire on an INCIDENT light reading (which is basically what zebras are doing if you use them properly on white or 18% gray) -- because you don't have to compensate for how you perceive the subject matter. If the light meter says the correct exposure is f/4 when the camera "l1/48th, 24fps and your F3 is at 0db, 1/48th, 24fps, REC709, without any ND filters engaged" then that is the correct exposure regardless of the subject matter.

Quote:
"White or light skin: IRE 30-40, Light dark skin IRE 40-50, Very dark skin IRE 50-60"
Starcentral, you avoided answering my questions about why you expose higher for dark skin than white skin. I'm asking you again to please explain this unusual approach. Why would you expose a dark person 20 IRE brighter than a white person?
 
Last edited:
Doug, F4 might be the correct exposure for 18% middle grey (50IRE) but when you shoot really fair skin you might chose to drop it down a bit and if you shoot darker people you might bring it up a little. If you have a mixed shot with both, then obviously you can't do much but a blend between the two. Some DOP's will actually light darker people up to a stop more within the same shot to balance out. It doesn't mean you will make a dark person look white (one stop is pretty subtle) you will just get a nicer image by bringing up what would normally appear darker on camera.

As I specifically indicated in my previous post the ranges I listed were just a guideline that I use for my own style of shooting. If you want to bump all the ranges up by 10, 20 or even 30 go ahead. Some people want faces sitting at 60-70 IRE based on what they are shooting ie. more for TV broadcast. I mostly shoot for a cinematic look and style so rarely would I go over 50 on a white face. But for example a black guy in black clothes against a light backgroud guaranteed I'm going over 50 to bring up those darker levels, or will use Cine 4.

Regardless of the baseline level you chose for your liking, it makes perfect sense depending on the circumstance to adjust your mid level up or down slightly based on skin type. A difference of 20 might be a little extreme between a fair skinned person and dark person but that is why I specified a range, you might go as low as 10 or as high as 20. Either way I find that lighting for different skin levels will result in a more pleasing and evened out image.

This thread is about using light meters with your F3: for people who want to learn how to match your meter to your camera there is an excellent video on AbelCine's website that shows how you can do this. Even if you don't care to calibrate your light meter to your camera as I have suggested doing, remember that a light meter is still VERY handy for checking your lighting ratios in your scene and even better can be done without any subjects, or a camera on set.
 
Thanks for your input Doug. I thonk StarCentral got the gist of what I was saying.

For example, if I were shooting Nicole Kidman I would meter 18% grey but would probably close iris a little for proper exposure.
Then if Don Cheadle sat in I would open up from 18% grey. But in both instances skin tones would remain in the same ire vicinity because a waveform merely reads the photons
coming in to a camera and tells me how many electrons I need to compensate.
 
Uh, I'm though here. Clearly my points are not being understood and I'm not going to continue to repeat the same point over and over again.
 
Last edited:
No way I'm handing a tape off to a producer where I don't ride the iris depending on the person who sits in the chair.
 
And if you ride the iris, not only will the background change, you've also lifted the black levels and the shadows won't be as deep as they should be. Blacks that should living around 0 will now be at 10 or something else depending how how much the iris has been changed. Now you've got a muddy picture.
 
I use my Light Meter set to the ISO the camera displays to determine the brightest, darkest area that I want details. If both exposure fall into the latitude of the camera then I set my t-stop in the middle and adjust once for PP detting. After that all adjustments are done at the grading stage.
 
There are a lot of scenarios to consider, perhaps thats why none of us can agree.

MOST people who shoot I would guess worry about the look of the talent, so riding the iris is a quick way to compensate for different levels of skin - yes it affects your background, but we're talking 5-10 IRE, one stop or less. If that's a concern to you then adjust your key instead - yes that works too.

What DougJ has suggested however is just balancing your camera 50IRE to a grey card using zebra stripes, which to me suggests never adjusting your iris or light depending on color skin you are shooting. I plainly disagree with that but shoot how you want, I certainly don't care. DougJ also says that using zebras on a white card or grey card is useful, I think with the F3 you are better off using the marker function - OR just whipping out your light meter and making sure you get F4 or f5.6 (whatever you've precalibrated your F3 as to get middle grey). This lets you muck around with lighting before you get to a set and use a $20,000 camera as your light meter. Once you are ready to shoot you can make final tweaks to your iris, lighting, whatever you want. Take another meter read and WOW now you know exactly how much light was going on your subject incase you need to recreate the same lighting later on, perhaps another day.

It's a common technique and not like I invented it. I used to work in a tv studio and it wasn't uncommon to see our producer of all people climbing a ladder to adjust the lighting to a certain brightness level on the meter so he could make it consistent each time. How are you going to measure the backlighting on some guys head with a grey card and $50,000 studio cam on wheels?

Anyway, in the end hopefully we can appreciate there are a variety of ways to gauge your exposure. To each his own.
 
Traveled all day for a shoot and just checked in here.

Again, StarCentral has echoed my sentiments exactly.

Ride the iris, adjust your key, either way, the point is the same. You're going to compensate for skin tone.
What you're never going to do is say your meter read a grey card at a certain exposure and walk away.

Back to the OP, I use a Spectra Professional IV-A and the brightness (spot meter) in camera in combination with an 18% grey card.
As you can infer from the posts this thread generated there's many ways to approach the subject so perform a few tests and see what works for you.
 
Last edited:
And if you ride the iris, not only will the background change, you've also lifted the black levels and the shadows won't be as deep as they should be. Blacks that should living around 0 will now be at 10 or something else depending how how much the iris has been changed. Now you've got a muddy picture.

Doug, I appreciate your point, and you're correct. Since this is a public forum I want to point out that I work with a small crew and we're always tweaking the values on the set. If I say I ride my iris, you can be sure one of my electricians is refining the background and shadow falloff, etc, which as you pointed out, is also a function of setting exposure.
 
20120319_102045.jpg
Today's interview: taken with my cell phone off a Panasonic BT-LH900.
F3 with Zeiss Super Speeds.

Olive skin-tone at about 55 IRE.
 
Back
Top