Jump to content


Photo

Light meter with digital?


  • Please log in to reply
25 replies to this topic

#1 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 19 June 2017 - 03:08 AM

I've been using the waveform in the viewfinder to set exposure. But I want to get to using a meter. I know many say they are not good for digital. But I still want to give it a go.

 

I have heard the rule of thumb to put caucasian skin at 70 IRE, darker complexion at 60 IRE and African American skin at 50 IRE.

 

But when it comes to an incident meter the reading is based on 18% gray, right? So that does not mean this is the correct reading for skin. So what is a good workflow for using the meter to set your key light etc? Thanks.

 

 


  • 1


Support Cinematography.com and buy gear using our Amazon links!
PANASONIC LUMIX GH5 Body 4K Mirrorless Camera, 20.3 Megapixels, Dual I.S. 2.0, 4K 422 10-bit, Full Size HDMI Out, 3 Inch Touch LCD, DC-GH5KBODY (USA Black)

#2 Brenton Lee

Brenton Lee
  • Basic Members
  • PipPip
  • 82 posts
  • Student
  • Melbourne, Australia

Posted 19 June 2017 - 04:06 AM

I'm a new jack so don't totally have my head around using IRE as a 'rule' but if you use the 'zone system' or a rough derivative of it, the 18% grey gives you a good starting point to how your scene is lit.

 

If you held a grey card in the same light as your subject and spot metered it, you know the measurement you get is around the middle of the 'system'. Skin tone / reflectivity can relate back to that ... a darker skin person could be similar, a lighter skin person would be a stop higher etc. 

 

As I said, I'm still learning the craft but look forward to any input the more experienced guys have to give. 


  • 0

#3 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 19 June 2017 - 04:13 AM

What I mean is, the reading you get in the meter is not really your T-stop right? It's just your starting point. So for example, if I'm shooting ISO 800 with a 180° shutter and my meter says T4. Putting my lens at T4 does not guarantee I have the right t-stop for the shot.


Edited by Michael Maier, 19 June 2017 - 04:14 AM.

  • 0

#4 Igor Trajkovski

Igor Trajkovski
  • Basic Members
  • PipPipPipPip
  • 277 posts
  • Other
  • Macedonia

Posted 19 June 2017 - 05:22 AM

If you take a reading with an incident meter for the sun,
any object or subject lit by it will be exposed correctly.

Pale, caucasian or black skin, foliage, sand, concrete etc.

will be properly exposed and show their true tonality.

Same with artificial key light.  You measure the light and
everything you place at the point of metering will be properly exposed.

That's in theory.

In practice, your starting point is the ISO, f-stop and shutter
as indicated on your camera and meter, and

there is a good possibility the image can be slightly over or under.

Then you compensate by offseting your ISO.

If you camera gives you proper(*) exposure at f2.8, 1/50, 800 ISO,
and your meter shows f2.8, 1/50 but on 640 ISO,
you'll know always to decrease the ISO by 1/3rd stop on the meter
from your camera setting.

(*) Proper - what you like on the monitor, the subjective method,
or by using gray card, or other charts with black, gray and white patches and placing them properly on the waveform.


...

Other factors to consider while you test this out:

 

- Metering toward the light source or metering toward the lens.
- Types of light affecting the metering. For example, daylight might give proper exposure, tungsten lower.
- Lens apertures behave differently then their indicated values.

 


  • 0

#5 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 19 June 2017 - 05:33 PM

So you are basically saying the meter gives you your key t-stop?

 

So if I set a key light up for an actor and the meter gives me T2.8 for that light, I can just set my lens to T2.8 and shoot ? Since a meter gives you 18% gray and I have always heard skin is around 1 stop above 18% gray, I didn't think the meter reading was a straight stop for the key.


  • 0

#6 Igor Trajkovski

Igor Trajkovski
  • Basic Members
  • PipPipPipPip
  • 277 posts
  • Other
  • Macedonia

Posted 19 June 2017 - 07:09 PM

Yes, you set the aperture indicated by your Incident meter.
That's the idea of incident metering.

See this example with white, gray and black plates exposed by incident meter:

Sekonic - The benefits of using light metering - Incident Metering

...

The 18% thing and +1 for skin tone is relevant in the realm of reflected, spot metering.

You point your spot meter at a white wall and expose by the readings it'll appear ~18% gray.
At 18% graycard it'll appear ~18% gray.
At a black cat, you'll get a ~18% gray cat.

So the spotmeter says "Point me at any surface and i'll tell you the exposure to make it appear mid gray".

The spot metered Caucasian skin will render mid gray.

It is lighter,  as the saying goes, by one whole stop.
You open up by stop from the reading and that's you proper exposure for that theoretical Caucasian skin.

 


Edited by Igor Trajkovski, 19 June 2017 - 07:10 PM.

  • 0

#7 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19490 posts
  • Cinematographer
  • Los Angeles

Posted 19 June 2017 - 07:18 PM

So you are basically saying the meter gives you your key t-stop?

 

So if I set a key light up for an actor and the meter gives me T2.8 for that light, I can just set my lens to T2.8 and shoot ? Since a meter gives you 18% gray and I have always heard skin is around 1 stop above 18% gray, I didn't think the meter reading was a straight stop for the key.

 

 

He was talking about an incident meter reading, you are thinking of a reflective meter reading (as what a spot meter gives you.) Incident meters measure the amount of light hitting the meter and therefore don't care what the actual subject is, whether it is a black bowling ball or a white baseball.  A reflective meter reads the amount of light reflected off of a subject and gives you a meter reading that would make your subject be 18% gray in terms of reflectance, so you have to decide how much brighter or darker to expose the subject.

 

ALL meter readings are just information that you factor in when placing your exposure, there are many situations where a face is brighter or darker than "normal" exposure.


  • 1

#8 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 20 June 2017 - 02:40 AM

Sorry for the confusion. I'm using an incident meter. I guess I confused the info with the spot meter or thought it applied to both. Thanks for clarifying guys.

 

But what David said is exactly what prompted me to ask the question. David, when you said "ALL meter readings are just information that you factor in when placing your exposure, there are many situations where a face is brighter or darker than "normal" exposure." This is why I was thinking the reading the incident meter gives me won't be exactly my lens T-stop for my key. It seems I still have other things to factor in to decide on my t-stop for my key? Or do you mean the t-stop reading from the meter will always be the correct or optimal t-stop for my key, and only if I want the face to be darker or brighter to fit the mood I will have something else to factor in? So as long as I want a correctly exposed face the t-stop from the meter will always give me that? Maybe correctly exposed is not the right word. Maybe normally exposed is a better term.

 

With a waveform if I'm setting a key light for a face and want the key side to be normally exposed, I will just get the key light to get me 65-70IRE and I'm done. Then based on the key and the contrast ratio I want I will set my fill etc. I will just eyeball my monitor till the contrast looks good and what I want and then I check the IRE and judge if it's enough contrast or too much difference.

 

If I want the face to be darkly lit, like for a scene which is supposed to be at night inside a dimly lit bedroom, I will instead set the key to give me less than 70IRE for the key. Then proceed to set fill etc if needed based on the key light.

 

With a meter it's easy enough to check contrast ratios. Actually easier than with the waveform in my opinion. My problem is how to arrive at the first reference point, the key.

 

So if I'm understanding this right and the t-top from the incident meter is indeed my key light t-stop, does that mean whatever t-stop the meter gives me will also give me 70IRE in the waveform? Somehow it doesn't seem that would be the case. Because the waveform measures the actual face. The incident measures the light. So regardless if it's Caucasian skin or African American skin the incident meter will always give me the same reading. Which is what made me think the straight reading from the incident meter can't be my final reading and there is still something else I have to factor in.


Edited by Michael Maier, 20 June 2017 - 02:49 AM.

  • 0

#9 Igor Trajkovski

Igor Trajkovski
  • Basic Members
  • PipPipPipPip
  • 277 posts
  • Other
  • Macedonia

Posted 20 June 2017 - 04:24 AM

You got it right by "...reading from the meter will always be the correct or optimal t-stop for my key, and only if I want the face to be darker or brighter to fit the mood I will have something else to factor in".

...

Make a shot with 2 heads.

One Caucasian skin the other African American.
How will you expose on a waveform?

Probably as you said, putting the Caucasian at 70IRE.
Where would the dark skinned person fall on the IRE?
Definitely lower.

Under that same amount of light you have properly exposed
light and dark skin.

Take an INCIDENT reading.
(If the meter readings don't match with your camera settings,
adjust the meter ISO until you got the same or near stop as in your camera.
Note the ISO differences)
 

 

Start from scratch, but measure with you meter first
(taking into account the possible differences in ISO between meter and camera).

Fire up your waveform, you should see the IRE's as before.
70 and lower for the dark skin.
 

...

Under the Sun, the white persons skin is white, the dark is dark.
They are lit by the same amount of light.

...

 

I have the impression, you might have exposed the dark complexions a bit hotter,
to appear lighter on screen and now be uncomfortable with the result of incident meter
making them darker.  The above relationship of IRE's in the two shot will remain when you
meter with the dome.
 


  • 0

#10 Guy Holt

Guy Holt
  • Basic Members
  • PipPipPipPip
  • 532 posts
  • Gaffer
  • Boston

Posted 20 June 2017 - 07:00 PM

... when it comes to an incident meter the reading is based on 18% gray, right? So that does not mean this is the correct reading for skin. So what is a good workflow for using the meter to set your key light etc? 

 

 

This is a complicated subject. The key tone, or mid gray because it appears in the middle of a photographic gray scale, is the common reference point used by light meters, lab processors, and transfer colorists.  A lab processor knows what density his processed emulsion should be to render mid gray and so he is able to adjust his processing accordingly. A transfer colorist, knows what gain will reproduce the luminance of mid gray so he is able to set his gain. Simply put it is the reference point to which we peg all other values on the characteristic curve of the film or digital format we are using. Meters are likewise calibrated for mid gray. For example there are basically two types of meters: incident and spot.  Incident meters read the light falling on your subject. Spot meters read the light reflecting back from your subject.  An incident reading gives you an exposure that after normal processing would render an 18% gray card as 18% gray (a specific density of the film) had you held it in front of the camera in the same light.  Incident meters enable you to peg the key tone (18% gray) in this fashion even though there may not be a mid-tone in your scene. There does not have to be a mid gray tone (commonly called the “key” tone) in the scene to expose for.

 

A spot meter is then typically used to take reflective readings to see how other objects will expose relative to the key tone that was pegged with the incident meter.  The thing to remember about spot meters is that they want to expose everything as mid gray.  For instance, if you expose a black piece of paper with the reading of a spot meter it will appear as mid gray after normal processing (not pushed or pulled) – likewise for a white piece of paper. But, if you place an incident meter down on the black piece of paper and expose for the incident reading the black paper will be black, and the white paper will be white, after normal processing because you exposed for the key tone by using the incident reading and thereby pegged the other values (white and black) relative to it. If you don’t have an incident meter, but want to peg the key tone under your subject's key light, an old trick is to take a spot meter reading of the palm of your hand under the key light and open up one full stop.  This will give you a close approximation because the average Caucasian flesh tone is one stop more reflective than 18% gray.

 

Metering_characteristic_curve.jpg

 

(The “Characteristic Curve” of a high contrast B&W Reversal emulsion. The object of exposure is to place the contrast range of the scene on the straight line portion of the curve so that the different luminances of objects in the scene are reproduced accurately on the film. Mid gray being the common reference point.)

 

You use the reading from the spot meter to be sure that the object you are metering will be within the exposure range (characteristic curve) of the film stock you are using. If the stock has a nine stop range (five stops over before detail burns out, and four stops under before detail blocks up), and your reading of a dark object is six stops under your key tone, it will not be rendered on the film after it is processed normal (to reproduce mid gray as mid gray).

 

Metering_Contrast_Exceeds.jpg

(The contrast range of this scene exceeds the film’s exposure range, so when the image is exposed for the light outside the arch (Left Image), detail is lost in the archway. Likewise, if the image is exposed to hold detail in the archway (Right Image), detail is lost outside the arch. Pegging the key tone centers the contrast range of the scene on the straight line portion of the curve such that some detail is lost outside and inside the arch way, but the luminance values of most of the scene are rendered accurately.)

 

Since in this situation the contrast range of the scene is beyond the exposure range of the film you have two choices. 1) You can open up and expose for the shadows (over exposing the key tone and blowing out your highlights more in the process) and print down to make mid gray mid gray again. In the end you have the detail in the shadows you want, but in the process you have lost detail in the highlights.  Why?  Because the contrast range of the scene was beyond the exposure range of the film, and you exposed for shadow detail, you burned out the highlights (no detail) so it is not there when you print down to mid gray.  You can't bring it back. Is that bad - not necessarily. It's just another "look."

 

Metering_Scene_Contrast.jpg

 

(The contrast range of this exterior exceeds the exposure range (flat line portion of the “Characteristic Curve”) so shadow details, in his hair and the black felt under the MacBeth Chip Chart, that fall on the bottom of the ”toe” of the curve “block up” (max density) and detail in that area of the frame is lost. Likewise, highlights like the specular sun on the windshield and the white towel, that fall on the upper end of the “shoulder” of the curve “burn out” (min density) ) and detail in that area of the frame is lost. The object of lighting is to compress the contrast range of the scene so that it fits on the straight line portion of the curve so that the different luminances of objects in the scene are reproduced accurately on the film.)

 

Or, 2) you can throw some light into the shadows to bring the reflective value of the dark object within the exposure range of the film (onto the straight portion of its’ characteristic curve) without changing the exposure of the key tone value (mid gray) or blowing out the highlights.  In this fashion you fit the contrast range of your scene into the exposure range of the film emulsion you are using. Of course this is only the starting point. From this “correct” exposure a DP will further manipulate the relationship of the contrast range of a scene to the exposure range of the film stock to create a desired effect. This is old school film exposure theory, but it is a good conceptual framework for exposing digital video, especially now that you can record "raw" and apply "looks" to the raw data.

 

Metering_Challange.jpg

(Post in this thread the problems in this image and how to fix them)

 

A fun exercise is to shoot a frame with just available light and then think about how you can improve upon it through lighting. For example, identify the problems in the image above and then list how to fix them through lighting and wardrobe.

 

Guy Holt, Gaffer, ScreenLight & Grip, Lighting & Grip Sales and Rentals in Boston


  • 0

#11 Guy Holt

Guy Holt
  • Basic Members
  • PipPipPipPip
  • 532 posts
  • Gaffer
  • Boston

Posted 20 June 2017 - 08:14 PM

 

This is a complicated subject. 

 

 

To help make sense of my post above let's take the scene below from “Miller’s Crossing” as an example.

 

Millers_Crossing_Example.jpg

 

It is a common fallacy that dark scene’s like this are “underexposed.” This scene is not underexposed, but rather the reflective values of the objects in the scene are carefully balanced (placed on the film’s characteristic curve) relative to the key tone by lighting so that most of the scene remains dark but serves up the full contrast range the film emulsion is capable of. In other words, even though the scene is correctly exposed, nothing in the scene is “correctly” exposed. The flesh tones are underexposed and the lampshade is over exposed in order to create the mood of the scene.

 

In an instance like this, the DP would not use a meter (incident or spot) to find the exposure of the key tone; rather, he would choose the exposure of the key tone from the outset  - say T5.6 for deep focus. And, having balanced the elements of the scene to that exposure using either his incident or spot meter, he will “lock it in” for lab timers or transfer colorists, by giving them the key tone (by properly exposing a chip chart with an 18% gray patch) as a reference at the head of the scene. Without providing the key tone, a timer or colorist will not know how dark the shadows should be or how bright the highlights should be because there is no other reference value at full exposure by which to calibrate the brightness of the scene.

 

Here are a few techniques, all of which are evident in this shot, that one could use to balance the lighting to create a dark scene without under exposing it.

 

1) Edge light objects in frame. Use reverse keys for talent and underexpose flesh tones by at least two stops or more. As long as you define the contours of your subject with subtle underexposed edges, don’t be afraid to let your talent fall off into black. There is a scene beautifully lit by James Merifield in the “Deep Blue Sea” of Rachel Wiesz and Harry Hadden-Paton standing in a dark alley way. They are back light by a practical at the end of the alley. Their contours are defined by the rims motivated by the practical, but otherwise their flesh tones fall off to complete shadows. James Merifield probably used a spot meter and negative fill to make sure that their flesh-tone would fall off the emulsion’s characteristic curve and reproduce as a pure silhouette. Sonnenfeld probably did the same in this scene to assure large parts of the frame had minimal detail.

 

2) I personally believe you should always have a hot spot in a frame – a practical in the scene or something in the deep background.  You can shift your overall exposure in the camera or in post to create a dark scene, but without a hot spot reference in the frame it will lack contrast and look underexposed.  A hot spot in the frame serves as a reference point and creates contrast. Practicals should be close to clipping and appear to be the source of light in a scene.

 

3) Don’t try to light your talent with only practical’s because they will blow out – the hot spot in your scene has to look natural. Not only is supplemental lighting required to light your talent, but you must also treat the practicals to make them look realistic. I find that practical lamps never look convincing unless one treats the lampshade as well as boost the bulb wattage. That is because if you stop down to keep the shade from burning out, the output of the practical, on the table it sits on or the wall its on, looks rather anemic. I find you get a more realistic look if you boost the wattage of the bulb and line the inside of the shade with ND gel. It is a delicate balance to obtain.

 

You can obtain this delicate balance without a monitor, by using the old school method with incident and spot meters and a selection of practical bulbs including PH 211, 212, and 213 bulbs. Years ago Walter Lassaley, BSC, instructed me to balance practical’s such that an incident reading of the direct output one foot away from the bulb is one stop over exposure. I have found that rule of thumb gives a realistic output to the practical - the light emitted downward onto the table top and upward onto the wall or ceiling is realistic. After establishing the practical’s output using an incident meter, you then use a spot meter to determine how dense an ND gel is needed to line the inside of the shade to place the brightness value of the shade on the characteristic curve of the emulsion so that it does not too hot and without detail.

 

4) Define the edges of your frame with a little detail. As long as you define the edges of your frame with a little detail, as Sonnenfeld does here, you can leave most of it black without it looking under exposed.

 

5) Soft sources like China Balls and Kinos are the wrong kind of fixtures for this kind of scene. You will need fixtures that you can easily control because you will need to cut them off large parts of your set. It will be hard to keep china balls and Kino Flos from spilling light all over the place and filling shadow areas that you want to keep dark. Fresnels with light diffusion inside the doors, cut with flags and nets, will give you the control you need. Spot meter readings of objects on the edge of the frame, like the upholstered chair on the left, will tell you if they are within the exposure range (characteristic curve) of the film. If they are not, use a little light to bring out detail that will define the edges of the frame as Sonnenfield has done here with the chair.

 

Guy Holt, Gaffer, ScreenLight & Grip, Lighting and Grip Equipment Sales and Rentals in Boston. 


  • 0

#12 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 21 June 2017 - 05:05 AM

You got it right by "...reading from the meter will always be the correct or optimal t-stop for my key, and only if I want the face to be darker or brighter to fit the mood I will have something else to factor in".

...

Make a shot with 2 heads.

One Caucasian skin the other African American.
How will you expose on a waveform?

Probably as you said, putting the Caucasian at 70IRE.
Where would the dark skinned person fall on the IRE?
Definitely lower.

Under that same amount of light you have properly exposed
light and dark skin.

Take an INCIDENT reading.
(If the meter readings don't match with your camera settings,
adjust the meter ISO until you got the same or near stop as in your camera.
Note the ISO differences)
 

 

Start from scratch, but measure with you meter first
(taking into account the possible differences in ISO between meter and camera).

Fire up your waveform, you should see the IRE's as before.
70 and lower for the dark skin.
 

...

Under the Sun, the white persons skin is white, the dark is dark.
They are lit by the same amount of light.

...

 

I have the impression, you might have exposed the dark complexions a bit hotter,
to appear lighter on screen and now be uncomfortable with the result of incident meter
making them darker.  The above relationship of IRE's in the two shot will remain when you
meter with the dome.
 

 

Yes, I find it that sometimes you need to wiggle a bit to get the skin to look as you see it in real life. If I just take the straight t-stop reading from the incident meter, especially darker skin it sometimes looks darker than it looks in real life.
 


  • 0

#13 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 21 June 2017 - 05:08 AM

Thanks for the detailed reply.

 

 

 If you don’t have an incident meter, but want to peg the key tone under your subject's key light, an old trick is to take a spot meter reading of the palm of your hand under the key light and open up one full stop.  This will give you a close approximation because the average Caucasian flesh tone is one stop more reflective than 18% gray.

 

But if the skin is one stop more reflective, it means it reflects twice as much light as 18% gray. So shouldn't you close the lens one stop instead of opening it? It seems to me if skin reflects more light than mid gray it is already overexposed if you leave it at the same stop, let alone if you open the lens even more. Or what am I missing?

 

 

 


You use the reading from the spot meter to be sure that the object you are metering will be within the exposure range (characteristic curve) of the film stock you are using. If the stock has a nine stop range (five stops over before detail burns out, and four stops under before detail blocks up), and your reading of a dark object is six stops under your key tone, it will not be rendered on the film after it is processed normal (to reproduce mid gray as mid gray).

 

Metering_Contrast_Exceeds.jpg

(The contrast range of this scene exceeds the film’s exposure range, so when the image is exposed for the light outside the arch (Left Image), detail is lost in the archway. Likewise, if the image is exposed to hold detail in the archway (Right Image), detail is lost outside the arch. Pegging the key tone centers the contrast range of the scene on the straight line portion of the curve such that some detail is lost outside and inside the arch way, but the luminance values of most of the scene are rendered accurately.)

It seems the spot meter has basically the same function as a waveform and if you have a waveform a spot meter is not needed. Or is there a need for one even if you have a waveform on set?

 

And exposing by incident meter, meaning taking the direct T-stop reading and setting your lens to it, since it exposes to 18% gray, it will expose to the mid tones and will average the exposure. Since it sets your exposure in the middle it gets you a compromises between exposing for the shadows and exposing for the highlights. Did I understand correctly?


  • 0

#14 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 21 June 2017 - 05:35 AM

 

 

To help make sense of my post above let's take the scene below from “Miller’s Crossing” as an example.

 

Millers_Crossing_Example.jpg

 

It is a common fallacy that dark scene’s like this are “underexposed.” This scene is not underexposed, but rather the reflective values of the objects in the scene are carefully balanced (placed on the film’s characteristic curve) relative to the key tone by lighting so that most of the scene remains dark but serves up the full contrast range the film emulsion is capable of. In other words, even though the scene is correctly exposed, nothing in the scene is “correctly” exposed. The flesh tones are underexposed and the lampshade is over exposed in order to create the mood of the scene.

 

In an instance like this, the DP would not use a meter (incident or spot) to find the exposure of the key tone; rather, he would choose the exposure of the key tone from the outset  - say T5.6 for deep focus. And, having balanced the elements of the scene to that exposure using either his incident or spot meter, he will “lock it in” for lab timers or transfer colorists, by giving them the key tone (by properly exposing a chip chart with an 18% gray patch) as a reference at the head of the scene. Without providing the key tone, a timer or colorist will not know how dark the shadows should be or how bright the highlights should be because there is no other reference value at full exposure by which to calibrate the brightness of the scene.

 

Here are a few techniques, all of which are evident in this shot, that one could use to balance the lighting to create a dark scene without under exposing it.

 

1) Edge light objects in frame. Use reverse keys for talent and underexpose flesh tones by at least two stops or more. As long as you define the contours of your subject with subtle underexposed edges, don’t be afraid to let your talent fall off into black. There is a scene beautifully lit by James Merifield in the “Deep Blue Sea” of Rachel Wiesz and Harry Hadden-Paton standing in a dark alley way. They are back light by a practical at the end of the alley. Their contours are defined by the rims motivated by the practical, but otherwise their flesh tones fall off to complete shadows. James Merifield probably used a spot meter and negative fill to make sure that their flesh-tone would fall off the emulsion’s characteristic curve and reproduce as a pure silhouette. Sonnenfeld probably did the same in this scene to assure large parts of the frame had minimal detail.

 

2) I personally believe you should always have a hot spot in a frame – a practical in the scene or something in the deep background.  You can shift your overall exposure in the camera or in post to create a dark scene, but without a hot spot reference in the frame it will lack contrast and look underexposed.  A hot spot in the frame serves as a reference point and creates contrast. Practicals should be close to clipping and appear to be the source of light in a scene.

 

3) Don’t try to light your talent with only practical’s because they will blow out – the hot spot in your scene has to look natural. Not only is supplemental lighting required to light your talent, but you must also treat the practicals to make them look realistic. I find that practical lamps never look convincing unless one treats the lampshade as well as boost the bulb wattage. That is because if you stop down to keep the shade from burning out, the output of the practical, on the table it sits on or the wall its on, looks rather anemic. I find you get a more realistic look if you boost the wattage of the bulb and line the inside of the shade with ND gel. It is a delicate balance to obtain.

 

You can obtain this delicate balance without a monitor, by using the old school method with incident and spot meters and a selection of practical bulbs including PH 211, 212, and 213 bulbs. Years ago Walter Lassaley, BSC, instructed me to balance practical’s such that an incident reading of the direct output one foot away from the bulb is one stop over exposure. I have found that rule of thumb gives a realistic output to the practical - the light emitted downward onto the table top and upward onto the wall or ceiling is realistic. After establishing the practical’s output using an incident meter, you then use a spot meter to determine how dense an ND gel is needed to line the inside of the shade to place the brightness value of the shade on the characteristic curve of the emulsion so that it does not too hot and without detail.

 

4) Define the edges of your frame with a little detail. As long as you define the edges of your frame with a little detail, as Sonnenfeld does here, you can leave most of it black without it looking under exposed.

 

5) Soft sources like China Balls and Kinos are the wrong kind of fixtures for this kind of scene. You will need fixtures that you can easily control because you will need to cut them off large parts of your set. It will be hard to keep china balls and Kino Flos from spilling light all over the place and filling shadow areas that you want to keep dark. Fresnels with light diffusion inside the doors, cut with flags and nets, will give you the control you need. Spot meter readings of objects on the edge of the frame, like the upholstered chair on the left, will tell you if they are within the exposure range (characteristic curve) of the film. If they are not, use a little light to bring out detail that will define the edges of the frame as Sonnenfield has done here with the chair.

 

Guy Holt, Gaffer, ScreenLight & Grip, Lighting and Grip Equipment Sales and Rentals in Boston. 

I guess this is the best part of digital's WYSIWYG. Looking at a calibrated monitor with a waveform a shot like the above is much easier to set up than in film where you can't see it and have to figure all the values.


  • 0

#15 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 21 June 2017 - 05:38 AM

Years ago Walter Lassaley, BSC, instructed me to balance practical’s such that an incident reading of the direct output one foot away from the bulb is one stop over exposure. I have found that rule of thumb gives a realistic output to the practical - the light emitted downward onto the table top and upward onto the wall or ceiling is realistic. After establishing the practical’s output using an incident meter, you then use a spot meter to determine how dense an ND gel is needed to line the inside of the shade to place the brightness value of the shade on the characteristic curve of the emulsion so that it does not too hot and without detail.

Very interesting rule of thumb. If we were talking about a bare bulb practical, making it difficult to gel, would you see it worth going through the trouble of figuring the t-stop the bulb would read one stop over exposed from a foot way and starting from this t-stop as a reference to your key, just to be able to keep the rule of thumb in effect? Or in this case would you abandon the rule of thumb for this one shot?


  • 0

#16 Robin R Probyn

Robin R Probyn
  • Basic Members
  • PipPipPipPip
  • 1520 posts
  • Cinematographer
  • Tokyo

Posted 21 June 2017 - 09:08 AM

Ive sort of come down this road the other way.. I gave my meters away when I started shooting  100% digital .. in this day with amazing OLED EVF (ok not cheap but compared to the old color ones pretty amazing ).. with false color,WFM,multiple color zebras.. all sorts of stuff in the EVF and very good monitors available now.. packed with every exposure tool in the universe .. or even just judging by eye off either these days.. Im not sure you need meters anymore.. unless you already have them and have worked that way for many years.. 


  • 0

#17 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19490 posts
  • Cinematographer
  • Los Angeles

Posted 21 June 2017 - 10:33 AM

Thanks for the detailed reply.

 

 

But if the skin is one stop more reflective, it means it reflects twice as much light as 18% gray. So shouldn't you close the lens one stop instead of opening it? It seems to me if skin reflects more light than mid gray it is already overexposed if you leave it at the same stop, let alone if you open the lens even more. Or what am I missing?

 

 

It seems the spot meter has basically the same function as a waveform and if you have a waveform a spot meter is not needed. Or is there a need for one even if you have a waveform on set?

 

And exposing by incident meter, meaning taking the direct T-stop reading and setting your lens to it, since it exposes to 18% gray, it will expose to the mid tones and will average the exposure. Since it sets your exposure in the middle it gets you a compromises between exposing for the shadows and exposing for the highlights. Did I understand correctly?

 

 

No, the reflective meter is reading the brighter skin tone and will give you a reading so that it will be 18% grey in reflectance if you use that reading, so you have to compensate and open up the f-stop to make it look brighter than 18% grey. You can think of it this way, if the skin is twice as bright, the reflective meter is "compensating" for that extra brightness by "stopping down" to get it to be 18% grey -- you're the one that has to compensate for that by opening up to get it to be lighter than 18%.  If the skin tone is normally 1-stop brighter than 18% grey and you take your spot meter reading and then close down 1-stop on the lens, now the skin tone will be 2-stops darker than its normal appearance.

 

The incident meter dome doesn't average exposures between shadow and key unless you aim the dome so that it gets both lights hitting it -- often you point the meter at the key and read the light coming from that direction (if you have a traditional 3/4 frontal key and frontal fill, then the bright side of the face is lit by the combination of the key and the fill, so pointing the dome towards the key will give you an accurate reading, you don't need to shadow the fill light off of the dome.  However, if you want to meter the fill alone, you'd want to use your hand to shadow the key off of the dome.

 

The incident meter isn't averaging the brightnesses on a subject, it's just telling you how much light is hitting the dome -- it doesn't care if the object is black, grey, or white.  So by using an incident meter, you know the exposure for the light hitting the space regardless of who or what moves through the space -- you know that the light hitting that desk against the office wall is, let's say, f/2.8 and when the actor walks into that light and sits down at the desk, they will also be "normally" exposed at f/2.8.

 

All meter readings are just information you take into account -- the actual choice in exposure will be a creative decision.  If you shoot a scene of someone standing outside in full sun, you may want it to "feel" sunny and hot and set your exposure to be a 1/2-stop brighter than what the incident meter tells you.  And in shooting in backlight, where the sun is mainly just a halo edge around objects, you will usually read the shadow side since that's the majority of the shot and decide how dark they should be to "feel" like shadows, 1-stop under for the shadows would be very conservative and your backlit highlights would be very bright (probably too bright for some video cameras), 2-stops under would still have shadow detail but definitely "feel" like you were looking at shadows, 3-stops under would be quite moody, almost like exposing for the highlights instead.

 

One good practice is to imagine or actually shoot a 360 degree shot that circles the subject -- think about how the exposure you choose would look as you went around from frontally lit to fully backlit.  Think about how it would look if you didn't adjust the f-stop during the circle and how it would look if you did, etc.  That will give you a good sense of how bright to let highlights go and how dark to let shadows go.


  • 0

#18 Michael Maier

Michael Maier
  • Basic Members
  • PipPipPip
  • 215 posts
  • Camera Operator

Posted 21 June 2017 - 02:20 PM

 

 

No, the reflective meter is reading the brighter skin tone and will give you a reading so that it will be 18% grey in reflectance if you use that reading, so you have to compensate and open up the f-stop to make it look brighter than 18% grey. You can think of it this way, if the skin is twice as bright, the reflective meter is "compensating" for that extra brightness by "stopping down" to get it to be 18% grey -- you're the one that has to compensate for that by opening up to get it to be lighter than 18%.  If the skin tone is normally 1-stop brighter than 18% grey and you take your spot meter reading and then close down 1-stop on the lens, now the skin tone will be 2-stops darker than its normal appearance.

 

The incident meter dome doesn't average exposures between shadow and key unless you aim the dome so that it gets both lights hitting it -- often you point the meter at the key and read the light coming from that direction (if you have a traditional 3/4 frontal key and frontal fill, then the bright side of the face is lit by the combination of the key and the fill, so pointing the dome towards the key will give you an accurate reading, you don't need to shadow the fill light off of the dome.  However, if you want to meter the fill alone, you'd want to use your hand to shadow the key off of the dome.

 

The incident meter isn't averaging the brightnesses on a subject, it's just telling you how much light is hitting the dome -- it doesn't care if the object is black, grey, or white.  So by using an incident meter, you know the exposure for the light hitting the space regardless of who or what moves through the space -- you know that the light hitting that desk against the office wall is, let's say, f/2.8 and when the actor walks into that light and sits down at the desk, they will also be "normally" exposed at f/2.8.

 

All meter readings are just information you take into account -- the actual choice in exposure will be a creative decision.  If you shoot a scene of someone standing outside in full sun, you may want it to "feel" sunny and hot and set your exposure to be a 1/2-stop brighter than what the incident meter tells you.  And in shooting in backlight, where the sun is mainly just a halo edge around objects, you will usually read the shadow side since that's the majority of the shot and decide how dark they should be to "feel" like shadows, 1-stop under for the shadows would be very conservative and your backlit highlights would be very bright (probably too bright for some video cameras), 2-stops under would still have shadow detail but definitely "feel" like you were looking at shadows, 3-stops under would be quite moody, almost like exposing for the highlights instead.

 

One good practice is to imagine or actually shoot a 360 degree shot that circles the subject -- think about how the exposure you choose would look as you went around from frontally lit to fully backlit.  Think about how it would look if you didn't adjust the f-stop during the circle and how it would look if you did, etc.  That will give you a good sense of how bright to let highlights go and how dark to let shadows go.

I understand now why you need to open the iris to correctly expose the skin with the spot meter. Thanks for clarifying.

 

So you prefer pointing the incident meter to the light instead of to the lens?

 

I understand I will have to judge and nudge the exposure to match the mood I'm trying to achieve. I'm just trying to understand how to get my first base number and how to interpret it. Which now I have understood that the incident meter gives me 18% gray, which is always the normal exposure. So all skins should look correct under it. And if I want it to look darker or brighter to match a scene which shouldn't be normally exposed, I have to adjust and either overexpose a little or underexpose a little. I think this seems to be clear now.

 

Although like I said, sometimes going by the incident meter reading, some skin complexions look darker than in real life even when other skin complexions, sometimes even in the same shot look like they normally do in real life. This I'm still trying to understand the reason for.


  • 0

#19 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19490 posts
  • Cinematographer
  • Los Angeles

Posted 21 June 2017 - 03:41 PM

The incident meter will preserve the brightness differences between faces of different tonalities, but the contrast of the film stock or the video gamma being used, the dynamic range of the camera, etc. will also have an effect on tonality, some things may go darker or lighter than what you see with your eyes, or when comparing different formats, cameras, processes.

 

I prefer pointing the incident meter dome at the source rather than at the camera -- I'd rather know that level the key is and then decide how I want to expose the key, I don't want the meter to do some sort of averaging between two sources.


  • 0

#20 Guy Holt

Guy Holt
  • Basic Members
  • PipPipPipPip
  • 532 posts
  • Gaffer
  • Boston

Posted 21 June 2017 - 07:29 PM

 Im not sure you need meters anymore.. unless you already have them and have worked that way for many years.. 

 

Without meters you are tethered to the camera and the camera can be a real bottleneck when it comes to each department of a production (Electric, Grip, Camera, Wardrobe, Make Up, Set Dec, etc.) having to accomplish what they have to accomplish before the camera can role again.  The value of meters, and knowing how to use them, is that they provide you the information you need to light a scene in your mind's eye and then translate that to reality. Rob Draper, when he was teaching at the Maine Film Workshops, used to call it the Zen of Cinematography and tell this story about how he works.

 

He said he would always  plot the lighting for a scene in advance on paper – specifying every detail down to the FC candle output from each instrument.  This allowed him to pass off to his Gaffer all the details he needed when he arrived on set.  Rob would then go off to craft service to get a cup of coffee.  By the time he finished saying good morning to everyone (client relations are very important for a DP), and got back to the set, the lights were starting to come up. With a cup of coffee in hand, he possessed a clarity of mind that enabled him to now take the lighting to the next level. He found time within time to address the finer nuances of shading and color: the Zen of cinematography. This is what you lose when you tether yourself to the camera.

 

The old school method was that the DP would choose the camera stop, which would establish the Key Tone - say T5.6 for deep focus.  Having chosen his exposure he can then calculate how many Foot Candles (FC) he needs on different elements of the scene.

 

To figure out how many FC you need for exposure, all you need to know is that it takes 100 FC to get an exposure of 2.8 with an ISO 100 film with a 180 degree shutter at 24 FPS (1/50th of a second shutter speed.)  If your digital camera is 2 stops faster than an ISO 100 film, you will need 100 FC to get a stop of 5.6. Once you know how many FC you need for exposure you can simply calculate how many FC will give you the effect you see in your mind’s eye. Of course, it helps to have done a lighting test of what effect over and underexposing a subject will give.

 

Such a lighting test for talent (you may also want to do one for key props or sets) would consist of testing in a systematic fashion the effect of Key, Fill, Back Light, Kickers, Hair Lights, and Liners that are over and under exposure. For example, to test the effect of your key light on flesh tones, set your exposure with two doubles and a single in your key light. Then remove them a half stop at a time (without changing your camera exposure setting or exposure of the chip chart), and systematically note on a slate in the frame what you are doing. Once you have removed all the scrims, your flesh tone will be two and a half stops over exposed (since you have not changed the camera setting.)  

 

Put all the scrims back in and now, using single and double nets, systematically under expose the flesh tone in half stop increments (remember rotating a net relative to the light source will make it "fatter" or "thinner", which will enable you to "dial in" the exact level you want from the light.) If you want to play on the lower register continue to under expose the flesh tone until it becomes a pure silhouette. Do the same for Fill, Back Light, Kickers, Hair Lights, and Liners in isolation and in specific combinations that you plan to use them. Having systematically tested each light, you can now see the effect that different ratios of each has on the scene and can even use the test as a reference on set when lighting the scene.

 

An example of this type of pre-visualization would be say you are shooting a couple conversing at a bar. After working through in your minds eye that you want a low-key look with selective (shallow) focus you might settle on a stop of 2.8. Say the script calls for the guy to be somewhat mysterious and distant and the women to be very open and receptive, then you may choose to keep him in deep shadow with just enough of a liner to separate him from the subdued background of the bar. This type of lighting on him could be motivated by a practical fixture you establish behind him, which would be consistent with the more frontal key you want for her, since you would want to light her more frontally so that her character is clearly apparent, but not him to retain some mystery to his character.

 

Having roughed out your style and light placement you can begin to set your levels and balance your lights based on lighting tests you have shot over the years. For instance, if your camera is two stops faster than an ISO 100 film, you will need 24 FC to properly expose your key tone (mid gray) at a T Stop of 2.8. 24 FC would then give you a “properly” exposed flesh tone on her. But this is a bar with subdued lighting, so you don’t want full exposure on her. You liked the feel of a half key (1 stop under) in your lighting tests so you would light her with 12 FC from a high frontal key. Again, because the scene takes place in the subdued lighting of a bar, you don’t want to over fill her. Going back to your lighting tests you like the look and feel of an 4:1 key to fill ratio so you would give no more than 3FC of fill light. You need to separate her from the dark background of the bar and so you might give her a backlight of 6 FC because that's what looked appropriate in the lighting tests to separate her hair color from a dark background without looking over-lit.

 

You would want to make sure you flag her backlight off him since you want to play him in near silhouette and so have to keep any frontal light on him to under 1 FC because four stops under exposure was a near silhouette with just the right amount of detail in the lighting test. For the liner to separate him from the dark background of the bar you will need a fairly strong fixture capable of delivering 48FC from directly behind him since your lighting tests established you need to be at least a stop over exposure for the liner to read.  Once you have figured out how many FC you need for the effect (a liner in this case) you can figure out which lights will give you that using the photo-metrics that manufacturers provide on their websites, or you can download Arri’s handy photometric calculator (be wary of the photo-metrics given for LED lights.) With a little experience you begin to develop a feel what light will give you what you need in different situations.

 

You wouldn’t want to try to use the practical fixture that you are flying in behind him to motivate this lighting scheme as the source for the liner on him because, first of all it’s placement in the shot may not be far enough around his back to serve as a liner. But, also to deliver 48 FC on him, it would be screaming hot in the shot.  For this reason it is better to use a separate light to light your talent and treat the practical so that it looks realistic in the shot. I find that practical lamps never look convincing unless one treats the lampshade as well as boost the bulb wattage. Unless it is completely opaque, you typically need to treat the shade to keep it from burning out (remember stopping down to keep it from blowing out will throw off the balance you have set with your other lights)  You can put a lower wattage lamp in it,  but then the output of the practical on the bar will look rather anemic. I find you get a more realistic look if you boost the wattage of the bulb and line the inside of the shade with ND gel. It is a delicate balance to achieve.

 

You can achieve this balance without a monitor, by using the old school method with incident and spot meters and a selection of practical bulbs including PH 211, 212, and 213 bulbs. Years ago Walter Lassaley, BSC, instructed me to balance practical’s such that an incident reading of the direct output one foot away from the bulb is one stop over exposure which in this case would be 48 FC. I have found that rule of thumb gives a realistic output to the practical. After establishing the practical’s output using an incident meter, you then use a spot meter to determine how dense an ND gel is needed to line the inside of the glass shade.

 

You can do all of this pre-visualization, setting of levels, and balancing based upon a location scout, blocking with stand-ins, and your lighting tests. In other words, almost everything can be worked out ahead of time so that when you arrive on set you know exactly what you need to do. This is especially helpful on low budget projects since, generally the time spent with minimal crew in scouting and blocking with stand-ins, is considerably less than the time wasted working these things out on set with a large crew and principle talent.

 

Guy Holt, Gaffer, ScreenLight & Grip, Lighting Rental & Sales in Boston


  • 0



CineTape

The Slider

Paralinx LLC

Glidecam

Metropolis Post

FJS International, LLC

Rig Wheels Passport

Tai Audio

Quantum Music Works

CineLab

Visual Products

Technodolly

Ritter Battery

Gamma Ray Digital Inc

rebotnix Technologies

Aerial Filmworks

Abel Cine

Broadcast Solutions Inc

Media Blackout - Custom Cables and AKS

Willys Widgets

Lenser

Technodolly

Visual Products

CineTape

Quantum Music Works

Glidecam

Media Blackout - Custom Cables and AKS

Willys Widgets

Ritter Battery

rebotnix Technologies

Gamma Ray Digital Inc

FJS International, LLC

Broadcast Solutions Inc

Abel Cine

Metropolis Post

The Slider

Lenser

CineLab

Paralinx LLC

Aerial Filmworks

Rig Wheels Passport

Tai Audio