Jump to content


Photo

Understanding Characteristic Curve


  • Please log in to reply
13 replies to this topic

#1 CHE

CHE

    New

  • Basic Members
  • Pip
  • 8 posts
  • Other

Posted 21 October 2005 - 01:11 PM

Let me see if I have got this straight....

If I were to take an incident reading on my main subject and get a read-out of a 1.4 f-stop...what is this telling me? Is it telling me that given the lighting conditions, film speed, and frame rate, I have just found middle gray? If I were to expose for that f-stop, would I be exposing near the toe and basically killing all the shadows while creating a huge range for the highlights? Follow up...I keep the same film speed and frame rate but introduce a couple of 1Ks. I take an incident reading on my subject on now get a F-16. Am I now creating a huge range for the shadows and little to no room for the highlights?

Finally, does middle gray have to fall in the middle of the characteristic curve and can it be located in nearer the toe or shoulder as mentioned above? Any help would be greatly appreciated. Thanks in advance.

curve_2.JPG curve_2.JPG

Attached Images

  • curve.JPG

Edited by CHE, 21 October 2005 - 01:17 PM.

  • 0

#2 Laurent Andrieux

Laurent Andrieux
  • Sustaining Members
  • 1527 posts
  • Cinematographer
  • France

Posted 21 October 2005 - 02:45 PM

If you are a student, you should wait for sensitometry courses..

If not... Let's try to explain a bit...

If you shoot a 18 % grey card, and expose it correctly, its density will be on the 11th point of the curve, let's say, in the middle of it.

If the object you shoot is not just a plain 18 % grey card (a real object, with blacks, whites and greys...) it will then be correctly reproduced : a black being black, a white being white, and all grey tones in the beetween.

If you underexpose, the values will fall in the toe, if you overexpose, they will go up in the shoulder.

You see, it's not that much a mistery. The purpose with photographic film is to reproduce as well as possible what we see with our eyes, since you use it correctly ie expose it correctly, according to its sensitivity. So if you see something's black or white or grey, and expose correctly, it will be black or white or grey.. That is all...

Did that help ?
  • 0

#3 John Pytlak RIP

John Pytlak RIP

    (deceased)

  • Sustaining Members
  • 3499 posts
  • Industry Rep
  • Rochester, NY 14650-1922

Posted 21 October 2005 - 07:37 PM

There's a pretty good sensitometry tutorial on the Kodak website:

http://www.kodak.com...structure.shtml

Plus alot of other good technical information:

http://www.kodak.com...=0.1.4.11&lc=en

And training:

http://www.kodak.com...d=0.1.4.9&lc=en
  • 0

#4 L K Keerthi Basu

L K Keerthi Basu
  • Basic Members
  • PipPipPip
  • 128 posts
  • Cinematographer
  • india

Posted 23 October 2005 - 10:38 AM

To learn it deeply please go through "Practical Sensitometry" by Wheeler and "Motion picture cameraman's handbook" by Russel Campbell

L.K.Keerthibasu
  • 0

#5 John Pytlak RIP

John Pytlak RIP

    (deceased)

  • Sustaining Members
  • 3499 posts
  • Industry Rep
  • Rochester, NY 14650-1922

Posted 23 October 2005 - 05:00 PM

To learn it deeply please go through "Practical Sensitometry" by Wheeler and "Motion picture cameraman's handbook" by Russel Campbell

L.K.Keerthibasu


Also the book by Todd and Zakia of RIT.
  • 0

#6 dbledwn11

dbledwn11
  • Guests

Posted 23 October 2005 - 06:00 PM

just wondering if incident and reflected light meters both work on the 18% grey principle.

I know that with a reflected light meter the reading always assumes that what you are pointing at is 18% grey and therefore reproduces it on film as that same grey, which is why you have to compensate by figuring it out in your head. for example. white skin reflects 36% light therefore you go down by one stop to produce white not gray.

obviously if i'm wrong here please correct me, but if not can someone explain if this is the same for incident readings. thank you.


these were the two bits from laurent's post that threw me - just in case that helps:

"If the object you shoot is not just a plain 18 % grey card (a real object, with blacks, whites and greys...) it will then be correctly reproduced : a black being black, a white being white, and all grey tones in the beetween"

"So if you see something's black or white or grey, and expose correctly, it will be black or white or grey"
  • 0

#7 Laurent Andrieux

Laurent Andrieux
  • Sustaining Members
  • 1527 posts
  • Cinematographer
  • France

Posted 23 October 2005 - 06:58 PM

just wondering if incident and reflected light meters both work on the 18% grey principle.

Le'ts look at it another way...

the reading always assumes that what you are pointing at is 18% grey

It's not exactly like that (though it's true) just follow me a minute. You are supposed to have a good reading when you point at the scene you want to photograph. Not everybody has a 18 % grey card in its pocket. Built-in meters work the same way. The thing is that since film is supposed to assume that a 2.5 % reflectance has to be black with details and 80 % reflectance to be white with details, the "average" (which is a geometrical mean, not arithmetical) is, mathematically, 18%. One usually considers, by eye, usually and actually, that this grey is "half way" beetween black and white. So the meters are calibrated on that value.

An incident light meter integrates the illuminations coming on it and from that value, give you a f-stop and time as well but doesn't depend on what you're shooting.

So that if ever the average reflectance of your subject is not 18 %, metering with a reflected light meter will give you a mean value so that your object's contrats fits the best in the curve (but therefore, a dark subject will be lighten up and vice versa), and reading with an incident light meter will let the values were they are supposed to be on the curve (a 18 % by position 11) but you may loose details if your subject is especially dark or light. The values given by the both methods will be the same if you read the rflected meter on a 18 %grey.

This is why one as to consider both "object's contrast" (depending only on its different reflectance, whatever light comes onto it) and "subject's contrast" (that includes the light and can be metered in reflectance, with a spotmeter better). Both being the same only if the light is totally uniform and everywhere the same.

I already posted once or twice a formula that helps understanding these things, which is actually the "meter calibration setting" formula, relating the different parameters that play a role in these stories :

N²/t = 2^EV = I.Siso/245 = L.S.ISO/12.5

Where N is the f-stop, t the exposure time, I the subject illumination in LUX (assuming 1 fc is 10.7 Lux, for incident light), L the brightness in cd/m² or nit. EV the exposure value you can find on meters.

Look at this, very interesting : http://www.schorsch....kbase/glossary/

Just thinking... Don't forget that a reflected light meter also integrates value, it integrates reflectance values. Only a spotmeter gives you a precise reading (making that if you read something that is not 18 % reflectance, will give you a value you have to "interpolate", like the skin that is 2/3 of a stop over KL)

Did that help ?
  • 0

#8 Charles MacDonald

Charles MacDonald
  • Sustaining Members
  • 1157 posts
  • Other
  • Stittsville Ontario Canada

Posted 23 October 2005 - 07:08 PM

just wondering if incident and reflected light meters both work on the 18% grey principle.

I know that with a reflected light meter the reading always assumes that what you are pointing at is 18% grey and therefore reproduces it on film as that same grey, which is why you have to compensate by figuring it out in your head. for example. white skin reflects 36% light therefore you go down by one stop to produce white not gray.

"So if you see something's black or white or grey, and expose correctly, it will be black or white or grey"


The reflected light meter is calibrated on the assumption - Which is true statisticaly that the average subject will refelct overall 18% grey. For quick shots, The Nickon scholl used to say aim the meter at the grass, and go from there. If you have ever had prints made from your stills at a one hour lab, you may have noticed the walls behind your subjects chage color with the clothing in the forground, as the printing machine also goes on the aumption that each still negative should add up to 18% grey.

If you aim the refected meter at an 18% grey card, you should get the exact same reading as an incident meter. and your average subject should fall right on the money.

If your subject is darker than average, say a head shot of a black actor in a navy suit, or lighter, say a swedish blond in a wedding dress. You may want to adjust the exposure to be sure that you capture everything.

If you use a refected meter for these examples, one with show that you need more exposure, and one will show you need less, even though your light has not changed. It is up to your "artistic decison" to decide if you want to shift the exposure under these conditions. If you use an incident meter, or course it will read the same in there cases as for an average shot. Again you may want to fudge the exposure to get more of the range onto the negative, so that you can pull more detail out when you print it. (or these days scan it). The fudging in this case would go the other way as with the refected meter the meter has compensated for the subject "all the way" and you may want to tone it down, while the incident meter has not canged at all for the suject and so you have to add whatever compensation you think you need/want all by yourself.

Again the bride, the refected meter would show MORE light, and call for LESS exposure, but you would probaly add a bit to compensate for the meter. other wise a one-light dialy would show the dress as Grey. The incident meter would not have changed for the subject so you might ask for a bit less so that the highlights don't get over exposed. - and you can see detail in the lace.

You normaly see pro photographers and cinematographers using the incident meter, as it gives one less level of subjective interpretation.

The clasic reference is probaly the books by Ansel Adams.
  • 0

#9 dbledwn11

dbledwn11
  • Guests

Posted 24 October 2005 - 02:15 PM

both posts from laurent and charles contain a lot of useful information, but which is kinda going over my head. i'm very grateful for the time and effort put into these explanations but i wonder if someone would be kind enough to simplify even further:-)

i hope this doesn't make me sound like a complete dunce.

might there be a way of setting out some more specific examples of when and how a dop decides to use either a reflected or incident meter. certain situations seem to call upon one type over the other.

Charles said: "You normaly see pro photographers and cinematographers using the incident meter"

would most people agree with this statement. why???

also might laurent (or anyone for that matter) clarify this idea of subject and object contrast. i can't quite make the distinction although i get the sense from Charles' follow-up post that the contrast between background and foreground action (in terms of colour and brightness) is an important issue in relation to metering principles (i.e. choice of meter type)

hope that makes sense and thank you very much for all the help so far and for any left to come.

Will.
  • 0

#10 Laurent Andrieux

Laurent Andrieux
  • Sustaining Members
  • 1527 posts
  • Cinematographer
  • France

Posted 24 October 2005 - 03:13 PM

Charles said: "You normaly see pro photographers and cinematographers using the incident meter"

would most people agree with this statement. why???


Yes they would. Nobody almost uses reflected light meters, but spotmeters. that allow to meter the brightness of a subjet, including its reflectance and the incident light coming upon it, at the same time.

If you consider the illumination coming upon it, E, it's reflectance, R, then it's brightness L is given by the Lambert's law : R.E = PI.L (where PI = 3.1415927...)

You see, a dark object, in the broad light can look brighter than a white object in the shade... THat's the whole diffrence beetween object's contrast and subject's contrast.

The only way to meter these subjects is the reflected spotmeter.

Dig it ? ;)
  • 0

#11 dbledwn11

dbledwn11
  • Guests

Posted 26 October 2005 - 06:33 PM

a late response to both charles' and laurent's posts, but just wanted to let you know that with a bit of further research and your very informative posts i am just starting to get a grip on the whole incident/reflected meter issue. thank you again. however, i cannot guarantee that there will no longer be any questions about this topic as i strive to find out even more about this fascinating subject:-)

Regards. Will.
  • 0

#12 Laurent Andrieux

Laurent Andrieux
  • Sustaining Members
  • 1527 posts
  • Cinematographer
  • France

Posted 26 October 2005 - 07:30 PM

You're welcome. I only wo,der what Che, the former poster, thinks about it...
  • 0

#13 Fran Kuhn

Fran Kuhn
  • Sustaining Members
  • 352 posts
  • Other

Posted 02 November 2005 - 02:23 AM

Let me see if I have got this straight....

If I were to take an incident reading on my main subject and get a read-out of a 1.4 f-stop...what is this telling me? Is it telling me that given the lighting conditions, film speed, and frame rate, I have just found middle gray?

curve_2.JPG curve_2.JPG

I'm not sure I understand the little graph attachment, but if you use your incident meter as explained above, you have, in fact, found the infamous middle gray (also known as 18-percent gray or Zone V). But that doesn't mean your SUBJECT will be middle gray on film. Follow this if you have a minute. I think it might help.

One of the simplest ways I've heard this whole, mysterious, 18-percent gray metering thing explained was by Zone System advocate Fred Picker. He used the example of black-and-white photos of two horses: one horse was white and the other was black. Old Fred simply pointed his trusty spot meter at the white horse and exposed according to the reading straight from the meter. He then used the same spot meter to take a reading of the black horse in the exact same light, and he again used the straight reading from the meter and exposed accordingly. Printing the developed negatives without exposure manipulation showed he had not a white horse and a black horse, but two identical 18-percent gray horses. If you care to experiment, a densitometer reading of such negatives will prove it.

The reflective meter reading simply tells you how to place your subject right at a middle exposure value. Things in the shot that are lighter than your metered subject will be lighter on film, things that are darker than your metered subject will be darker, right up to the limits of the film to record highlight and shadow detail. This simple example shows why reflective readings must be interpreted to provide the results you want.

So how do you interpret reflective readings? Look at the case of the two horses. Since you know that the spot meter is going to try and turn the white horse into a middle gray horse, simply take your reading and open the lens about two stops from what the meter says. Opening the lens' aperture that much provides enough additional exposure to go from middle gray, or Zone V in Zone System parlance, to about Zone VII, or what most observers would interpret as something that looks like white. Same thing in reverse for the black pony: take your spot reading and close down a couple of stops from what the meter tells you and you now have a nice, solid Zone III, something closer to black. Of course, this all depends on the film?s latitude, but close enough for this explanation. And while there are some specific differences, this theory applies to color film in most cases.

Now, if you had stood in front of those same two horses with your little incident dome and pointed it at the camera, the incident reading would, in theory, give you a correct setting for both horses without a lot of interpretation. So why would anyone use a spot meter? Well, since a reflective reading tells you the exact amount of light bouncing back at the camera to strike the film, it gives you a lot more useable information than a simple incident reading.

Spot meter readings used in conjunction with a working knowledge of the film?s (or video's) latitude can give advance warning of problem areas in your frame. If you rely exclusively on incident readings taken at the main subject?s position, that subject may be properly exposed, but when the film comes back, you may spot things your eye missed in the heat of the moment. Maybe important detail is concealed in dark areas in the background, or maybe a table lamp is excessively bright and draws your eye away from the main subject. A few moments spent sweeping the spot meter over the scene will often allow you to fine-tune your lighting and sometimes help avert a total disaster.

I have to admit I often use both meters on the same setup. I use the incident meter to get the lights roughed-out and the spot meter for fine-tuning.

Anyway, hope this helps.
  • 0

#14 John Pytlak RIP

John Pytlak RIP

    (deceased)

  • Sustaining Members
  • 3499 posts
  • Industry Rep
  • Rochester, NY 14650-1922

Posted 02 November 2005 - 10:14 AM

Again, the Kodak website has a good tutorial:

http://www.kodak.com...eP.shtml#curves
  • 0


Metropolis Post

FJS International, LLC

Tai Audio

Ritter Battery

Paralinx LLC

Technodolly

Willys Widgets

Abel Cine

Aerial Filmworks

Opal

Visual Products

Media Blackout - Custom Cables and AKS

rebotnix Technologies

Glidecam

The Slider

Rig Wheels Passport

CineTape

Gamma Ray Digital Inc

Broadcast Solutions Inc

Wooden Camera

CineLab

Tai Audio

Opal

The Slider

Media Blackout - Custom Cables and AKS

Paralinx LLC

Wooden Camera

rebotnix Technologies

Broadcast Solutions Inc

CineLab

Rig Wheels Passport

Glidecam

Technodolly

Willys Widgets

Gamma Ray Digital Inc

Abel Cine

Aerial Filmworks

Ritter Battery

FJS International, LLC

CineTape

Visual Products

Metropolis Post