Jump to content


Photo

Metering issues...help needed!


  • Please log in to reply
17 replies to this topic

#1 Mihai Nicolau

Mihai Nicolau
  • Basic Members
  • PipPip
  • 52 posts
  • Other
  • Bucharest, Romania

Posted 16 April 2008 - 03:20 PM

Hello. Today i went out to test the way my lightmeter (a Sekonic) works and after 100+ photos i'm more scared than ever because in a few days i'm going to shoot a project for college using a Kodak B/W emulsion (Kodak Plus-X rated at 80ASA outdoors). I understand the way meters are supposed to "see" and the difference between reflected and incident but after a full day of using the meter the theory just doesn't seem to add up.
I took with me a digital SLR - Canon 300D - and after measuring with the meter i dialed in the exposure on the SLR. Somebody told me that this is not relevant as DLSRs don't see as film does so maybe this was the problem, but in some situations i got a well exposed image with the measeurements taken by the meter so i don't think this is the issue here.

I'll post some photos to illustrate what i'm talking about. I only wanted to see how the meter "sees" so i apologise for the bad image quality :rolleyes:

1) I took an incindet reading of this scene.

Attached Images

  • incident_reading.jpg

  • 0

#2 Mihai Nicolau

Mihai Nicolau
  • Basic Members
  • PipPip
  • 52 posts
  • Other
  • Bucharest, Romania

Posted 16 April 2008 - 03:26 PM

Hello. Today i went out to test the way my lightmeter (a Sekonic) works and after 100+ photos i'm more scared than ever because in a few days i'm going to shoot a project for college using a Kodak B/W emulsion (Kodak Plus-X rated at 80ASA outdoors). I understand the way meters are supposed to "see" and the difference between reflected and incident but after a full day of using the meter the theory just doesn't seem to add up.
I took with me a digital SLR - Canon 300D - and after measuring with the meter i dialed in the exposure on the SLR. Somebody told me that this is not relevant as DLSRs don't see as film does so maybe this was the problem, but in some situations i got a well exposed image with the measeurements taken by the meter so i don't think this is the issue here.

I'll post some photos to illustrate what i'm talking about. I only wanted to see how the meter "sees" so i apologise for the bad image quality :rolleyes:

1) I took an incindet reading of this scene.


2) Then a reflected reading of the same scene (and it now is corectly exposed). If I measured in the incident way but with the sphere pointed towards the light strip and away from the camera lens (as for a reflected meter) the exposure was way better, though still an F stop or so lower than the reflected one.

Attached Images

  • reflected_reading.jpg

  • 0

#3 Mihai Nicolau

Mihai Nicolau
  • Basic Members
  • PipPip
  • 52 posts
  • Other
  • Bucharest, Romania

Posted 16 April 2008 - 03:36 PM

I had 2 more examples but it seems that i cannot upload more photos - they were the same thing, the incident reading was overexposed and the reflected one was ok. What am I doing wrong here ? Is the incident meter fooled by the backlight ? should i use only the reflected one in this case ? The problem with reflected is that my meter only has this 56angle lumisphere and i cannot know for sure for what part of the image it biases the exposure.

Thanks!
  • 0

#4 Michael Nash

Michael Nash
  • Sustaining Members
  • 3330 posts
  • Cinematographer
  • Pasadena, CA

Posted 16 April 2008 - 04:10 PM

An incident light meter reads the light falling on the subject, from the angle the dome is pointed toward. The meter is only measuring the light hitting the dome from that angle, at that position. It doesn't know or care what other subjects or light values are in your frame.

So in the first photo it looks as though the two women on the right are properly exposed, as though you took an incident reading at the sidewalk with the meter's dome pointed toward the camera. This is exactly the exposure I would expect from an incident reading at that position. Again, the meter doesn't know or care what other light values are in your frame, let alone how you wanted the scene to look.

screenshot2.jpeg

The reflected light meter in your camera, on the other hand, is reading ALL the light values it sees in the frame (depending on how the metering is weighted or biased), and averaging them together to find a mid point. It's trying to make all the values average out to middle gray, and since there are extreme values of light and shade in this frame you end up with an even distribution of values across the full tonal range.

You say the second photo is "correctly exposed", but what exactly is correctly exposed in the frame? The men and trees are almost silhouette, significantly underexposed. The sky is blown-out white, significantly overexposed. And the overall frame has an even balance of values. My point is that YOU have to determine what a proper exposure is; a light meter can only read light and tell you what will render middle gray.

Your incident meter is not being fooled by the backlight. It's telling you an accurate reading of the light falling on it. If anything it's the other way around -- YOU are being fooled (or influenced) by the backlight, deciding that a "proper" exposure is one that places the shaded side of the subjects darker than normal, in order to balance with the highlights in your frame.

This is the fundamental art of cinematography. You take a reading of the light, and then you decide how bright dark you want it to appear, and set your exposure accordingly.
  • 0

#5 Mihai Nicolau

Mihai Nicolau
  • Basic Members
  • PipPip
  • 52 posts
  • Other
  • Bucharest, Romania

Posted 16 April 2008 - 04:36 PM

Great! thanks, you opened my eyes :rolleyes:

But for me the first photo just has to many overblown details to be good. The second one has a more dramatic feeling to it because of the higher contrast.
The question that bothers me now is will my film stock (the Kodak black and white negative 100 ASA) catch more details from the first photo especially in the highlights as it has much more latitude than the sensor in my Canon SLR ? I don't know the current lattitude of my DLSR but i think its about 4 stops.

What is the best approach to a light situation as the one from the first photo ? To measure incident and then compensate 1-2 stops to get more details in the highlights with the consequence of losing some in the shadows or maybe using a reflected meter such as a spot meter to take readings from all of the scene and then bias my exposure as i see fit, or maybe lose the reflected metering ideea totally and just learn to deal with it the reflected way ?
  • 0

#6 Ed Moore

Ed Moore
  • Basic Members
  • PipPip
  • 54 posts
  • Cinematographer
  • UK

Posted 16 April 2008 - 04:52 PM

Hi Mihai,

Before you get too caught up in comparisons of the latitude of your film stock vs the DSLR, you should consider as Michael is suggesting, exactly *how* you want to expose the scene. The latitude of the particular imaging system you use will have an effect on the end image, but it will be noticeable only at the extreme ends of the image - i.e. the brightest and darkest parts of the frame.

You still need to decide where to 'place' the exposure in order to get the effect you want.

When you stood at that location in person, did you imagine an image like your second photo, with the people in silhouette and more detail in the areas in sunshine? Or did you imagine a scene with a strong backlight?

It's about having an image in your mind's eye which you then use your knowledge of exposure to commit to the film (or CCD as the case may be).

Do you have a spot meter on your Sekonic? If so, it might be worth playing with using that to select an area of the frame to be exposed as middle gray, or 18% reflectance, or Zone V. Or, if you want to expose caucasian skin 'normally', point your spot meter at some, take the reading from the meter and open the lens one stop from the reading (reason being caucasian skin is generally accepted to be Zone VI).

The zone system is fun. I recommend Blain Brown's Cinematography: Theory and Practice for a good introduction specific to cinematography.
  • 0

#7 Michael Nash

Michael Nash
  • Sustaining Members
  • 3330 posts
  • Cinematographer
  • Pasadena, CA

Posted 16 April 2008 - 06:24 PM

Yes, Ed's right. The best approach is to take the info that the meters are telling you, and interpret that info to give you the exposure you want to create. Sometimes you have meter readings to guide you; other times you have to make an educated guess and compensate the numbers in your head. Practice makes perfect!

Edit:
Also, latitude is the amount you can over- or under-expose the film and still recover a normal looking final image. Dynamic range is the range of light values that the system can capture. Your DSLR doesn't have 4 stops of latitude, in has 4 stops dynamic range.
  • 0

#8 Kevin Mastman

Kevin Mastman
  • Basic Members
  • PipPip
  • 23 posts
  • Student

Posted 16 April 2008 - 08:06 PM

I'm also a student and have also been confused about this for a while. It looks like in the second picture, the areas on the building and sidewalk that have sunlight hitting them are properly exposed. This would mean that if you took an incident reading at those locations, the sidewalk and areas not in shadow would be properly exposed and the shadows and highlights would fall into place relatively. Right?

When you take the incident reading and point it toward the camera, isn't that going to give you a reading resulting in an overexposed image because all the light is coming from behind the meter? The way I understand it is: its the same as taking a reading in the shadows because there is no sunlight hitting the meter and you are going to get an image more like the first one.

Am I correct/ if you were in this environment and wanted a shot to look like the second photo, what would be your process in reading the light?

Thanks,

Kevin
  • 0

#9 Christopher Arata

Christopher Arata
  • Basic Members
  • PipPip
  • 83 posts
  • 1st Assistant Camera
  • Los Angeles

Posted 16 April 2008 - 08:39 PM

You might want to try this...if your meter is both a incident & spot as a lot of Sekonic's are...

First take your Incident reading now people will argue this but there actually are many different ways to take a incident reading. I personally prefer to take it of whatever I am using as my "Key" light. Once I get that reading, say its a f/4 I will then switch to spot mode and find something that is of the same value of f/4 in the scene. Then I will press Avg and with my light meter still in spot mode I will go through the scene and see how much over and under certain parts are. From there I will adjust if needed the lighting but generally I come up with what I would like my lighting ratio to be before I start so I have a pretty good Idea of what I am trying to achieve.

All the best.

-Christopher Arata
  • 0

#10 Hal Smith

Hal Smith
  • Sustaining Members
  • 2280 posts
  • Cinematographer
  • OKC area

Posted 16 April 2008 - 09:50 PM

........... From there I will adjust if needed the lighting but generally I come up with what I would like my lighting ratio to be before I start so I have a pretty good Idea of what I am trying to achieve.

I think the first poster doesn't understand that any media has a given latitude (film) or dynamic range (digital). If the range of light values in a given scene are larger than the media being used can handle there will be areas that are too dark and/or areas that are blown out. Hence the necessity of lighting (fixtures or reflectors) to bring everything into a range that the media being used can handle. The first photo posted has this problem, if Mihai had added light on the figures on the sidewalk then (s)he (Sorry, I'm clueless as to name gender in Romanian!) could have stopped down or used some ND to bring everything within the dynamic range of the camera being used.
  • 0

#11 Christopher Arata

Christopher Arata
  • Basic Members
  • PipPip
  • 83 posts
  • 1st Assistant Camera
  • Los Angeles

Posted 16 April 2008 - 10:08 PM

I think the first poster doesn't understand that any media has a given latitude (film) or dynamic range (digital). If the range of light values in a given scene are larger than the media being used can handle there will be areas that are too dark and/or areas that are blown out. Hence the necessity of lighting (fixtures or reflectors) to bring everything into a range that the media being used can handle. The first photo posted has this problem, if Mihai had added light on the figures on the sidewalk then (s)he (Sorry, I'm clueless as to name gender in Romanian!) could have stopped down or used some ND to bring everything within the dynamic range of the camera being used.


Hal.

I'm glad that you brought this up in response to what I wrote. I actually had written something in there like that and then decided to take it out as I thought one would understand what I meant or was getting at. Looking back on it I should have left it in.

-Christopher Arata
  • 0

#12 Michael Nash

Michael Nash
  • Sustaining Members
  • 3330 posts
  • Cinematographer
  • Pasadena, CA

Posted 16 April 2008 - 10:46 PM

I'm also a student and have also been confused about this for a while. It looks like in the second picture, the areas on the building and sidewalk that have sunlight hitting them are properly exposed. This would mean that if you took an incident reading at those locations, the sidewalk and areas not in shadow would be properly exposed and the shadows and highlights would fall into place relatively. Right?


If you took an incident reading with the meter's dome pointed toward the sunlight, yes.

When you take the incident reading and point it toward the camera, isn't that going to give you a reading resulting in an overexposed image because all the light is coming from behind the meter? The way I understand it is: its the same as taking a reading in the shadows because there is no sunlight hitting the meter and you are going to get an image more like the first one.


Don't confuse the scene with the subject. Once again, an incident meter doesn't know and doesn't care what else is your frame. The meter only measures the light falling on the dome. If you take an incident reading at a subject that's in the shadows, it's going tell you the proper exposure for the subject that's in the shadows. If your scene includes areas in direct sunlight, you might choose to underexpose the shadowed subject to achieve a more balanced exposure throughout the frame. But if your frame doesn't include areas of direct sunlight, you might choose to go with what the meter tells you.

Case in point: This is an enlargement of the first photo, the one you say is an "overexposed image." Does this picture look like an overexposed image?
Posted Image

No, because it doesn't include the areas in direct sunlight. Don't confuse the readings of the subject with the reading of the scene.


http://www.cinematog...1208380103.jpeg

if you were in this environment and wanted a shot to look like the second photo, what would be your process in reading the light?


One of two ways. I'd either take an incident reading of the direct sunlight and an incident reading toward the camera and expose right in between the two readings; or, I'd spot meter a bright area (but not blown out sky), a dark area (but not pure black), and an area I wanted to appear middle gray, and I'd try to determine an exposure that would put all of those areas in the right place.
  • 0

#13 Michael Nash

Michael Nash
  • Sustaining Members
  • 3330 posts
  • Cinematographer
  • Pasadena, CA

Posted 16 April 2008 - 11:13 PM

I think the first poster doesn't understand that any media has a given latitude (film) or dynamic range (digital).


Just to clarify (not nitpick), it's not that "latitude" means film and "dynamic range" means digital. The terms describe two different concepts.

Latitude is the amount an image can be over- or under-exposed in camera and still brought back to an acceptable looking "normal" exposure.

Range is the range of brightness that the system can capture. In the strictest sense the term "dynamic range" comes from electronics and not film, so perhaps simply "range" is the better term.

They're two different things. If an imaging system can capture detail in medium-value subjects from 4 stops underexposed to 4 stops overexposed, it has 8 stops of range, but certainly not 8 stops of latitude. You could probably only underexpose the image about 2 stops before you'd start to lose too much shadow detail (and get increased grain), or two stops overexposed before you'd lose too much highlight detail. That means it has a latitude of + or - 2 stops.

So when people say things like "this camera has 8 stops of latitude," it's very misleading and incorrect. That would mean you could over- or under-expose the image by about 4 stops and still recover an acceptable looking image. And that simply ain't the case!
  • 0

#14 Saul Rodgar

Saul Rodgar
  • Basic Members
  • PipPipPipPip
  • 1682 posts
  • Cinematographer

Posted 16 April 2008 - 11:57 PM

Am I correct/ if you were in this environment and wanted a shot to look like the second photo, what would be your process in reading the light?

Thanks,

Kevin


If you wanted the picture to look like the second one, you would do what Mihai did: take a reflected reading standing from where the picture was taken. It is about figuring out what object you want to be middle gray. The light meter will tell you what the exposure for something to be middle gray is, but first you have to choose the object and you have to either point the light meter there or figure it out yourself in your head (compensate).

It looks like in the second picture, the areas on the building and sidewalk that have sunlight hitting them are properly exposed. This would mean that if you took an incident reading at those locations, the sidewalk and areas not in shadow would be properly exposed and the shadows and highlights would fall into place relatively. Right?


You are saying that certain areas are properly exposed, but that doesn't mean Miahi was trying to make it that way necessarily: they came out properly exposed but Mihai may have been going for the people walking by to be properly exposed, or the sky. Mihai broadly pointed the meter in reflective mode and took a picture. The meter averaged the light intensity and the sidewalk appeared well exposed, its value was the closest to middle gray within that picture. The sidewalk is correctly exposed within Mihai's picture, but are you going to use them as a way to expose your own image? So yeah, if you take their value as middle gray for your purposes, then your picture should more or less look like Mihai's second picture. If I wanted to choose the middle gray value of the image to be the sky, then I would have to measure, compensate and expose accordingly.

In other words, and in an overly simple way, look at the picture you want to take. Choose an object to represent your middle gray. Measure the light falling on it, or reflected from it. Compensate/ expose accordingly. Then your picture should come out with that object correctly exposed, with about 5 stops of detail in every direction, to white and black. The zone system is a little more complex, but still based on compensating/ exposing on what you choose (or not) to be the middle gray. But it is ALL about knowing what object you want to have as your middle gray value in the picture and compensating if need be to make it properly exposed.
  • 0

#15 Mihai Nicolau

Mihai Nicolau
  • Basic Members
  • PipPip
  • 52 posts
  • Other
  • Bucharest, Romania

Posted 17 April 2008 - 12:11 AM

THank you all for some great answers!

First of all I think there is a different way of interpreting these terms in my country since i was told that the latitude of film is the portion on a gamma curve on which the film densities rise proportionally with the amount of light the emulsion receives (the straight line between the toe and the shoulder of the curve). So in other words i think it would be the range of the film stock as you call it. Or maybe i've got this lesson wrong...i'll look into it <_<

My question was if i had shot the first photo with the same measurements from the meter, BUT on film stock and not on my DSLR what would the differences have been ? Could the film stock provide me with enough err...range (hope this is the right term to use now ;) ) to get more details in the highlights ?

PS. So if understand correctly from your example above if an imaging system has 8 stops range the latitude means that in the 2 upper stops from the total 8 stop range you can still expose and have details with the same going for the lower 2 stops right ?
  • 0

#16 Saul Rodgar

Saul Rodgar
  • Basic Members
  • PipPipPipPip
  • 1682 posts
  • Cinematographer

Posted 17 April 2008 - 12:26 AM

My question was if i had shot the first photo with the same measurements from the meter, BUT on film stock and not on my DSLR what would the differences have been ? Could the film stock provide me with enough err...range (hope this is the right term to use now ;) ) to get more details in the highlights ?


it'll be close but never identical, but it is a good way of getting an idea of what it'll look like on film. Reversal film is a little narrower on its latitude and dinamic range so that is going to influence the results as well.
  • 0

#17 Mihai Nicolau

Mihai Nicolau
  • Basic Members
  • PipPip
  • 52 posts
  • Other
  • Bucharest, Romania

Posted 17 April 2008 - 02:56 AM

So a film stock would not resoelve the highlight detail in that shot ? By "it'll be close" you mean that the film stock is near the range of my DSLR ? I'm getting a little confud sed here!
  • 0

#18 Michael Nash

Michael Nash
  • Sustaining Members
  • 3330 posts
  • Cinematographer
  • Pasadena, CA

Posted 17 April 2008 - 03:36 AM

My question was if i had shot the first photo with the same measurements from the meter, BUT on film stock and not on my DSLR what would the differences have been ? Could the film stock provide me with enough err...range (hope this is the right term to use now ;) ) to get more details in the highlights ?


Different film stocks have different characteristic curves and exposure ranges. You can look up the curve for the film you're using to get an idea how it may respond, but ultimately you have to test it to know for sure.

http://www.kodak.com...f002_0142ac.pdf

http://www.kodak.com...s...1.4.6&lc=en

PS. So if understand correctly from your example above if an imaging system has 8 stops range the latitude means that in the 2 upper stops from the total 8 stop range you can still expose and have details with the same going for the lower 2 stops right ?


With a hypothetical, linear 8-stop range you'll have detail in midtones up to about 4 stops over, but no highlight detail because the highlights would be beyond the range of the film. You'd have to keep your exposure limited to no more than 2 stops over if you wanted to capture highlight detail that's 2 stops brighter than your midtones. That's what's meant by "2 stops overexposure latitude."

http://www.kodak.com...s...0.4&lc=en#L

In the real world though, latitude is rarely linear. Color negative has much more overexposure latitude than underexposure, and video cameras and digital still cameras are just the opposite.
  • 0


Paralinx LLC

Ritter Battery

Abel Cine

FJS International, LLC

Visual Products

Technodolly

Broadcast Solutions Inc

Glidecam

Tai Audio

CineTape

rebotnix Technologies

CineLab

Opal

Aerial Filmworks

Wooden Camera

Gamma Ray Digital Inc

Rig Wheels Passport

The Slider

Willys Widgets

Metropolis Post

Media Blackout - Custom Cables and AKS

Rig Wheels Passport

rebotnix Technologies

Abel Cine

CineTape

Opal

Wooden Camera

Aerial Filmworks

Paralinx LLC

Metropolis Post

Glidecam

Ritter Battery

Technodolly

The Slider

Visual Products

Tai Audio

Media Blackout - Custom Cables and AKS

CineLab

Broadcast Solutions Inc

Gamma Ray Digital Inc

FJS International, LLC

Willys Widgets