ISO is not a part of exposure.

[Daniel Browning]

What is this post about?

The purpose of this post is to explain the correct meaning of the word "exposure" and dispel the common misconception that ISO is part of the definition of exposure.

Why post about definitions?

I think it is valuable to explain and understand the correct meaning of the word because it is central to photography itself. When a novice thinks the word "zoom" means a long focal length, I think it's valuable to explain the true meaning. Exposure is even more important than zoom.

What is the real definition of exposure?

Exposure is the total amount of light per unit area that falls on the sensor.

This is not a matter of opinion or individual perspective. It's concretely defined in authoritative resources like "Photography", 9th Edition, by London and Upton. Most photography books get the definition correct as well, though I have found several that do not. Even regular dictionaries have the correct definition. There are many magazines and web sites that do not use the word correctly.

What is the difference?

Compare two photos taken with the same scene luminance:

Photo A:

f/2.8
ISO 100
1/125


Photo B:

f/11
ISO 1600
1/125


Both photos have the exact same brightness, histogram, and meter position; therefore, they both qualify as the same "exposure" according to the incorrect and colloquial definition of exposure. But in reality, the exposure is different.

What about ISO?

Changing the ISO setting on a camera does not affect the total amount of light per area; therefore it does not change the exposure. ISO is the combined sensitivity of the sensor, electronics, raw conversion, and post processing. Since none of these affect the total amount of light per area, they are not a part of the definition of the word exposure: ISO is completely and wholely separate.

But isn't ISO a part of the exposure triangle?

Yes, it is a part of the triangle, but that does not make it a part of exposure itself. Cameras with flash and ND settings affect exposure even though they are not a part of the triangle.

But doesn't ISO affect the camera's meter and histogram?

Yes, but that doesn't mean it is exposure. As photographers, the brightness of the image is very important to us. The histogram and meter are two tools that help us get the brightness we desire. In order to get the correct brightness, they must factor in the choice of ISO. That allows the tools to indicate what exposure to use to get the desired brightness.

How can you use ISO if it's not a part of exposure?

There are three examples below. The first sentence uses the incorrect colloquial definition of exposure ("[colloq.]"). The second sentence uses the correct definition.


My image is too dark. I am already at the maximum f-number, shutter, and amount of light, so I am going to increase the exposure [colloq.] by changing ISO 100 to ISO 200.
My image is too dark. I am already at the maximum f-number, shutter, and amount of light, so I am going to increase the brightness by changing ISO 100 to ISO 200.



My image is too noisy. I am using ISO 1600, 24mm f/2.8 lens wide open, and the slowest shutter speed I can manage. I am going to spend an extra $1,400 on a 24mm f/1.4 lens to use wide open. I will keep exposure [colloq.] the same by reducing ISO from 1600 to 400.
My image is too noisy. I am using ISO 1600, 24mm f/2.8 lens wide open, and the slowest shutter speed I can manage. I am going to spend an extra $1,400 on a 24mm f/1.4 lens to use wide open. I will keep brightness the same by reducing ISO from 1600 to 400, but I will increase the exposure by two stops thanks to f/1.4.



My shutter speed is too slow. If I increase it, the image will get too dark, I can't compensate because I'm already at the maximum f-number and amount of light, so I'm going to increase the exposure [colloq.] by changing ISO 200 to ISO 400.
My shutter speed is too slow. If I increase it, the image will get too dark, I can't compensate because I'm already at the maximum f-number and amount of light, so I'm going to increase the brightness by changing ISO 200 to ISO 400.


I hope those examples help illustrate how it is possible to use ISO, exposure, and brightness together without including ISO in the definition of exposure.

How does this relate to photography?

Light is an important part of photography. The word photography itself comes from the Greek and means "drawing with light". The correct definition of exposure links it directly with light itself, while the many misconceptions about exposure move this important connection to light.

Why do we need a word that is related to light?

There are many reasons why we need a word that is related to light. One is that photographers spend lots of money trying to get more light (larger sensors, faster lenses, flashes, etc.), and they spend lots of time getting the exact right amount of light. One of the biggest reasons for this is that the amount of light is a vital factor that affects noise, color depth, dynamic range, and more. Another reason is that raw photographers can change brightness easily in post, but the amount of light cannot be changed.

What if you don't care about light?

There certainly are some photographers that do not care about light. For them, all that matters is how bright a photograph is. They feel that ISO and f-number are equally important, because they both affect the brightness of the image by the same amount and they don't care about noise or any of that other stuff. That is fine for them if that's what they want to believe, but I still think it's valuable for them to understand and use the correct definition. There are plenty of words for them to use to talk about the ideas that are important to them, such as "brightness", "darkness", and "apparent exposure". They can use those words correctly rather than using "exposure" incorrectly.

For example, there may be some photographers that do not care about the difference between "zoom" and a long focal length. The only zoom lenses they use are ones with long focal lengths, so the difference is unimportant to them. However, I think that it's still valuable for them to know the correct definition, even if they don't know or care about zoom lenses which do not have long focal lengths (such as wide angle zooms).

How did things get this way?

I think the root cause of the widespread misconception about ISO and exposure is film. Back in the days of film, the only convenient way to increase brightness was to increase exposure. (Sure, there were some inconvenient ways, such as changing the film, "pushing" chemically during processing, etc., but most people just used more exposure instead.)

This lead to many film shooters to think that exposure was the same as brightness. It wasn't really, of course, but with film the difference wasn't important. You could use it incorrectly and no one would know because there is no difference. If the only method you have is exposure, then the word "exposure" might as well mean the same thing "brightness".

But digital is different. Now it is very easy to "push" (in camera or in post). Unfortunately, many people
still have the film-era misconception.

Can you sum up?

It is important to understand what exposure is: the total amount of light per unit area.

[KrishnenduKes]

Amazing write up. Thanks for sharing.

Disussion Approved

[shutterbug]

Thanks for the detailed write up Dan! Going through it presently.

[hitanshu]

Very interesting. But you have totally confused me now.

Need to read it again to fully understand how it should impact my shooting :)

[Aryan]

That is SOME information. Thanks for sharing this! :)

[powerslave]

I think about it in terms of my dad's Nikon FE. The film remains at a constant ISO. What you see through the light meter ir what you get on the film.

Leave it to the digital age to make stuff complicated!

Nice writeup!

EDIT: Please put me on the confused list.

[robin234]

excellent write up

ISO only increases the sensitivity to capture more light :),but its not exposure .

[KrishnenduKes]

I finally read the whole thing. Very interesting article. And very precise. Summed up in the last line. ISO has got nothing to do with Exposure. Thanks for sharing. Now we would like a similar thing about focal lengths and zooms!

[Daniel Browning]

Thanks.


ISO only increases the sensitivity to capture more light :)

Actually, in digital cameras it does not. The amount of light that is captured depends only on exposure; no matter if you set the camera to ISO 100 or ISO 100,000. All that happens when you change the ISO setting is the camera boosts the brightness. Sometimes the brightness is boosted in the electronics (analog domain), other times in software (digital domain), but either way, the underlying sensitivity to light is the same.

[robin234]

Thanks.



Actually, in digital cameras it does not. The amount of light that is captured depends only on exposure; no matter if you set the camera to ISO 100 or ISO 100,000. All that happens when you change the ISO setting is the camera boosts the brightness. Sometimes the brightness is boosted in the electronics (analog domain), other times in software (digital domain), but either way, the underlying sensitivity to light is the same.

well,if thats the case ,then we should rename it to some other term as ISO is misnomer in this case .

I'm not sure ,but only your article states ISO as "Brightness" in digital photography .Never read this in some-other article/book .Its good to know little bit of technical details ,but at the end of the day ,picture is all that matter

[Bibudesh]

Very nice writeup Daniel, thanks for sharing.


Thanks.
Actually, in digital cameras it does not. The amount of light that is captured depends only on exposure; no matter if you set the camera to ISO 100 or ISO 100,000. All that happens when you change the ISO setting is the camera boosts the brightness. Sometimes the brightness is boosted in the electronics (analog domain), other times in software (digital domain), but either way, the underlying sensitivity to light is the same.

The noises created at higher ISOs is the software limitation of camera or the analog domain limitation? Taking only the DX format sensors for an e.g.- two camera with similar sensor gives different level of noises at same ISOs.

I need to understand what is the factor which decides the price of two camera with nearly same specs but differing performance at higher ISO.

Also, when we have separate sliders for Exposure and Brightness in any raw handling software. What is the main difference? What does it actually mean by saying that RAW files contains info about extra stops, I mean practically how this works?

Also, (this question may need a new thread), do u have any experience in sensor freezing to handle noise in long shutters? If yes then can u please throw some light on that?

EDIT: Apologies, a D40/D80 user always has issues with Exposure (read inaccurate metering LOL)

[Daniel Browning]

The noises created at higher ISOs is the software limitation of camera or the analog domain limitation?


Neither. The primary cause is reduced exposure. The reduced exposure increases photon shot noise in any given tonal level as well as decreasing the distance to read noise dominated tonal levels.

As for the difference between analog/digital ISO, one important distinction is that analog gain sometimes has less read noise (in an absolute sense) than digital gain (especially with CMOS). Another way to put it is that at high analog gain, less noise (relative to a fixed exposure) is added by the electronics in the camera. So that means there is a benefit to analog gain over digital gain, but only when there is highlight headroom to spare and when read noise improves.



I need to understand what is the factor which decides the price of two camera with nearly same specs but differing performance at higher ISO.


Sorry, I don't think I understand this question.


Also, when we have separate sliders for Exposure and Brightness in any raw handling software. What is the main difference?


It's completely arbitrary. In raw conversion software, "exposure" and "brightness" can mean just about anything (and often do). For example, in Adobe, the exposure control applies a highlight guessing algorithm, while the brightness control applies nonlinear EC to preserve highlights but does not use the highlight guessing algorithm. In other programs (e.g. RawTherapee 3.0 alpha), the "Exposure" slider is the nonlinear EC control.


What does it actually mean by saying that RAW files contains info about extra stops, I mean practically how this works?


It means that the default JPEG engine in the camera throws away information (highlights and shadows) needlessly for a more pleasing image (higher contrast). By converting the raw file, you can create a different jpeg that does not throw away the information, but uses it.


Also, (this question may need a new thread), do u have any experience in sensor freezing to handle noise in long shutters? If yes then can u please throw some light on that?


Yeah. The only type of photography I can think of that benefits from active sensor cooling is astrophotography. It reduces thermal noise, which increases linearly with exposure duration and ambient temperature. You can have your DSLR serviced by a third party to add a cooling unit, or build your own Peltier system.

[Deltaone]

Why o why are you confusing something which is essentially both simple and logical?

Yes, your essential definition is correct. But you're extrapolating it into something else and confusing things beyond where they should be confused.

"In photography, exposure is the total amount of light allowed to fall on the photographic medium"

"In photographic jargon, an exposure generally refers to a single shutter cycle. "

http://en.wikipedia.org/wiki/Exposure_%28photography%29

Read that, particularly the difference between radiometric and photometric exposure. You are talking about photometric exposure, which is weighted vis a vis the human eye, in which case sensitivity of the medium/sensor doesnt matter.

However, we are talking about the process of taking photographs. We take a photo to get an effect we wish to acheive, the "accetable" settings for any given scene depend on what the photographer wishes to acheive in that particular scene. In that case, you actually have to look at radiometric exposure not photometric.

Now it gets even more complicated when you come to digital.

"For digital cameras, an exposure index (EI) rating—commonly called ISO setting—is specified by the manufacturer such that the sRGB image files produced by the camera will have a lightness similar to what would be obtained with film of the same EI rating at the same exposure."

If you get really technical about this, it can get confusing,which is why i prefer to use the simpler and rather more practical methods i did in the other thread. (Not that i'm any great expert mind you... ).

[Daniel Browning]

Why o why are you confusing something which is essentially both simple and logical?


For every important concept, there is a description that is both simple and logical, but totally wrong
--H. L. Mencken.

Redefining "exposure" to have the same meaning as "brightness" might make it more simple and logical to you, but it's still wrong. We already have plenty of words for "exposure + ISO": darkness, brightness, etc. If you don't want to discuss the amount of light, then use the word brightness instead of the word exposure. If you do care about the amount of light, then exposure is the word you need and it should not be watered down by adding things that are not related to light.


http://en.wikipedia.org/wiki/Exposure_%28photography%29

Read that, particularly the difference between radiometric and photometric exposure. You are talking about photometric exposure, which is weighted vis a vis the human eye, in which case sensitivity of the medium/sensor doesnt matter.


No. I am not talking about either of those. In any case, neither of those include anything like the meaning of ISO either. You misunderstood the portion where it says "adjusted to account only for light that reacts with the photo-sensitive surface". That does not mean sensitivity in the same way as ISO. It only means relative spectral sensitivity -- i.e. constraining the total amount of light from the entire EM spectrum to just the relevant portion (e.g. visible, IR, etc.).


However, we are talking about the process of taking photographs. We take a photo to get an effect we wish to acheive, the "accetable" settings for any given scene depend on what the photographer wishes to acheive in that particular scene. In that case, you actually have to look at radiometric exposure not photometric.


No. In the first place, the difference between the two is irrelevant for normal photography, since we're only interested in visible light and the cameras only respond to visible light. (UV/IR photographers and radio astrophotographers are the ones who care about that stuff.)

Second, it's quite possible to discuss the acceptable settings for any given scene without using the word exposure incorrectly. For example, instead of saying "I think the photo is too dark, fix it by increasing the ISO, which will increase the exposure", you can say this: "I think the photo is too dark, fix it by increasing the ISO, which will increase the brightness."


Now it gets even more complicated when you come to digital.


Not at all. The meaning of the word exposure is the exact same between digital and film.


"For digital cameras, an exposure index (EI) rating—commonly called ISO setting—is specified by the manufacturer such that the sRGB image files produced by the camera will have a lightness similar to what would be obtained with film of the same EI rating at the same exposure."

If you get really technical about this, it can get confusing, which is why i prefer to use the simpler and rather more practical methods i did in the other thread.

There is nothing simple or practical about using the word exposure incorrectly.

[Xavier]

Brilliantly written, Daniel :)

[synn]

I am simply not in the mood to repost everything I have posted in this thread and the ones before it or even post web links, but I'll just say I oppose Browning's stance on the matter. The reasons are quite clear.

[Daniel Browning]

I am simply not in the mood to repost everything I have posted in this thread


For clarification, I think what synn means by "this thread" is the following:

Another controversial topic : High ISO = less noise?

I'll just say I oppose Browning's stance on the matter. The reasons are quite clear.

No problem. Since you're not in the mood to repost, I'll respond to your comments in the "High ISO = less noise" thread.

[synn]

I think you misunderstood.

I'm not reposting the things I've posted every other month just because another thread pops up. Not in this thread, not in any other thread. The posts are all still there and anyone looking for a counter point can refer to them.

That's about it.

[Deltaone]

So, basically you're saying, there is only one possible definition for a word and anyone else who says anything else is absolutely wrong, misunderstood, or just doesn't get it?

Did you actually read the wikipedia article? Really read it i mean? Mind you, you wouldn't have to try so hard at selective understanding if it weren't for the fact that you are missing the forest for the damn trees.

"No. I am not talking about either of those. In any case, neither of those include anything like the meaning of ISO either. You misunderstood the portion where it says "adjusted to account only for light that reacts with the photo-sensitive surface". That does not mean sensitivity in the same way as ISO. It only means relative spectral sensitivity -- i.e. constraining the total amount of light from the entire EM spectrum to just the relevant portion (e.g. visible, IR, etc.)."

Constraining the total amount of light from the EM spectrum to just the relevant portion? Kinda obvious that one, aint it? Why the heck would anyone be concerned about the irrelevant portion anyways? The relevant portion of light which reacts to the surface....... Kinda important the reaction part.

Photography or the capturing of a picture, when film was used is done by chemical reaction. I hope you have no problem with that statement. As such which chemicals react in which way to what light and to what intensity is as much relevant as anything else. Time is as much a factor of the equation as anything else.

Now time works in a very simple way, in this particular case. How long is a film exposed to get a desirable result (i use the word desirable because not everyone sees the same thing in the same scene, which is the whole point of photography in the first place). You control that by means of shutter speed. Now the amount of light is controlled by the aperture. True, only these two matter when the medium of sensitivity remain the same. However, if i were to use a more sensitive medium, then the other two have to be adjusted accordingly. It isn't called a photographic triangle for the fun of it. It isn't because somebody thought three points are more fun than two.

Again, you're going too much into the definition and missing the point altogether. Again, that is why i included two examples of both "exposure" and "an exposure". Please do remember, the whole purpose of a language is to convey ideas and as such, no language is complete or perfect in itself. WHich is why there are such things as double meanings and phrasings and a whole lot of other things.

Now, personally i don't care what you call something. To paraphrase, a rose by any other name is still a rose. However, you seem to be laboring under the misapprehension that the sensitivity of the medium doesn't matter. Which is just plain old wrong.

Yes, only two things actually control the amount of light which falls on the sensor. The aperture and the shutter speed. Unfortunately, its the sensor/medium which actually does the work of recording the image in question and if you actually think that it doesn't matter, i have no idea of how in the world you arrive at such a conclusion......

Ok, something i just remembered. You seem convinced that only two things affect photography and that iso/film speed is just brightness. Ok, lets go with your assumption.

I have for you a simple thought experiment. I take a photo at 9 aperture and 1/100 SS. Let us assume that for that particular scene, that produces a good exposure, no clipped highlights etc.

Now i take the same settings, same scene, ramp up the iso to 1600, i get blown out highlight or just a white frame. You say, that is because the camera ramped up the noise. Now mind you, since the scene, time, and other settings are the same, it is only possible that the camera did this after the shot was captured, because laws of physics don't allow the camera to produce light inside the sensor to increase the brightness.(unless it boosted the sensitivity, but lets not go there for now).

Now, the same experiment, but i use film this time. Two cams, running 100 and 1600 ASA film. Same settings. I get a good shot in one, a blown out/white frame in the other. Now since it is film, there is no in-cam post processing. How exactly did the brightness as you call it increase given the amount of light falling on the medium is the same?

[Daniel Browning]

So, basically you're saying, there is only one possible definition for a word and anyone else who says anything else is absolutely wrong, misunderstood, or just doesn't get it?


No. If you want to make up your own definition for words, that's fine for you, but don't be surprised if other people correct you.


Did you actually read the wikipedia article? Really read it i mean?


Yes.


Mind you, you wouldn't have to try so hard at selective understanding if it weren't for the fact that you are missing the forest for the damn trees.


You are mistaken.


Constraining the total amount of light from the EM spectrum to just the relevant portion? Kinda obvious that one, aint it? Why the heck would anyone be concerned about the irrelevant portion anyways? The relevant portion of light which reacts to the surface....... Kinda important the reaction part.


Whether you assume photometric or radiometric exposure has no effect on this thread: you can pick whichever one you want and ISO is still not a part of exposure. Again, ISO sensitivity is not the same as spectral sensitivity.


True, only these two matter when the medium of sensitivity remain the same. However, if i were to use a more sensitive medium, then the other two have to be adjusted accordingly.


No. First of all, there is no digital camera that has ajustable sensitivity. The amount of light detected by the camera is always the exact same no matter the ISO setting. It is only the analog gain or digital gain that is adjustable.

In any case, even if you did use a more sensitive medium, then you would have to make no adjustment to exposure if you adjusted your developer accordingly. This is a common case with digital photography so it's not correct to say that exposure has to be adjusted.

For example, if you were using Vision3 250T film stock developed with +1 push, then switched to Vision3 500T film stock with no push, there is no adjustment to exposure necesssary. I have to use a film example here because, again, no digital camera has adjustible sensitivity -- the ISO only affects the developer (gain).


It isn't called a photographic triangle for the fun of it. It isn't because somebody thought three points are more fun than two.


Three points of the triangle affect brightness, but only two of them affect exposure. If you change ISO, the exposure is still the same, but the brightness is different. In order to get the brightness back to where it was, you can change one of the other points of the triangle (which does affect exposure). But again, none of that means that changing ISO is a change in exposure. Only f-number and shutter speed change the exposure.


Again, you're going too much into the definition and missing the point altogether.


No.


Again, that is why i included two examples of both "exposure" and "an exposure".


Your statement merely points out that the difference between "b" and "c" below:

a. the act of presenting a photosensitive surface to rays of light.
b. the total amount of light received by a photosensitive surface or an area of such a surface, expressed as the product of the degree of illumination and the period of illumination.
c. the image resulting from the effects of light rays on a photosensitive surface.

Of course there are several different meanings of the word exposure in photography, but none of them conflict with eachother. This thread is about "b", and the common misconceptions about it.


However, you seem to be laboring under the misapprehension that the sensitivity of the medium doesn't matter.


No. You are laboring under the misapprehension that that changing ISO has any effect on the sensitivity of the digital sensors (it doesn't) and that a change in the sensitivity of the medium is a change in exposure (it isn't).


Yes, only two things actually control the amount of light which falls on the sensor. The aperture and the shutter speed.


That is incorrect. There are many factors that affect it, including lens transmissivity, flash settings, ND settings (for cameras that have them), and scene luminance.


Unfortunately, its the sensor/medium which actually does the work of recording the image in question and if you actually think that it doesn't matter, i have no idea of how in the world you arrive at such a conclusion......

I said nothing of the like. I did not say the sensor does not matter. I did not say that ISO does not matter. I said that ISO is not a part of exposure. It is a part of brightness.

[synn]

ISO affects brightness you say?

So I meter a daylight shot at 100 ISO, shoot it at 1600 ISO with the same aperture and SS, return home with a pure white frame, adjust the brightness in post and would get back the image that was supposed to be recorded at 100 ISO? No?

Conversely, I meter a night shot at 1600 ISO, shoot at 100 ISO with the same aperture and SS, return home with a pure black frame, adjust the brightness in post and would get back the image that was supposed to be recorded at 1600 ISO?

No again?

Oh, what a crazy world we live in!

[Deltaone]

I note that you have ignored my simplified question.

That said, so now your argument is, the definition of exposure involves only two things, iso does not affect it. Iso affects the brightness of the image and the exposure triangle is actually the brightness triangle by your definition.

So basically you're arguing that exposure the way we or I use it is wrong, what i'm talking about is the brightness of the image. But if i were to substitute the word brightness instead of exposure then my argument is correct? The sensitivity of the medium involved affects only the brightness? Is that your argument?

If that is so, its just semantics. And i can be just as adamant that i am right and show you umpteen number of examples as you can. Doesn't really matter. Which is why my earlier phrase of a rose by any other name....

That said, if you're actually arguing that the sensitivity of the medium doesn't matter, then you're flat out wrong and its not open to interpretation. That said, yes, what a digital sensor does is to actually increase the gain. And said gain is why you have such things as S/N ratios and quite a few other kind of noise. Which i why i said digital makes things a bit more complicated.

Now again, if you're just arguing semantics, it has no point in the discussion. But assuming the whole point of this thread was to prove the other thread wrong by giving me definitions? Er, you're going to have to do a whole lot better than that. So my original point still stands. "Exposure triangle."..... Sensitivity does matter. Read above example....

Now, if you will admit to the previous point and are still arguing about the exact word definition of the term exposure, this is going nowhere. It can go either way. There is a reason why most people tend to use simpler terms and definitions, complexity can confuse :P.

Which is especially the case when trying to explain the various kinds of noise in digital media. Which is what the other thread was going on about. IF you get kinda technical you could possibly come to the assumption that high iso produces less noise, which is just so laughably absurd to anyone who has actually taken a few photographs under different lighting, that it makes me want to beat my head against a wall each time this comes up in various forums.(Or to paraphrase, if i had a penny for each time, it would be a mountain of pennies by now :P ).

Try to understand this rather simple concept that i try to put across. There is a reason why we call it an exposure triangle. Because all three affect the final output, i.e. the photograph, (one assumes taking a photograph is why you're into this in the first place....). That is why we try to put it in simpler ways. Does understanding all the actual electronics inside a modern camera make you a better photographer? I doubt it. However, the exposure triangle is kind of a basic which you really can't do without. And as i keep saying, take the camera, go out, shoot, and you will understand the basics yourself. Which is something xavier seemed to keep missing in the other thread. Hence the long argument, which has lead to this thread which to be absolutely frank, if im right about the first part of this post, is quite useless because semantics never benefitted anyone, unless youre a professor of linguistics :P.

[Daniel Browning]

ISO affects brightness you say?

So I meter a daylight shot at 100 ISO, shoot it at 1600 ISO with the same aperture and SS, return home with a pure white frame, adjust the brightness in post and would get back the image that was supposed to be recorded at 100 ISO? No?


Why are you making things up? I never said anything like the idea that increasing brightness with ISO does not blow highlights, or that they could be recovered. Increasing brightness with exposure does the same exact thing. I'll break it down for you:


Increasing ISO increases brightness.
Increasing shutter duration increases brightness.
Both increase brightness.
Both can blow highlights.
Only one of them is an increase in exposure.

[Daniel Browning]

I note that you have ignored my simplified question.


Which question is that? If you are referring to the one that you added to your post in a later edit, the reason I did not respond to it is that I did not see it. I only responded to your original post, not the edited one. Now that you have brought it to my attention, I'll be happy to go back and respond to it after lunch.

[Daniel Browning]

I'm back from lunch and I'll answer your simple question now.


Ok, something i just remembered. You seem convinced that only two things affect photography


No. That is not what I said at all. I said that ISO is not included in the definition of exposure. I did not say that ISO is not included in photography.


I have for you a simple thought experiment. I take a photo at 9 aperture and 1/100 SS. Let us assume that for that particular scene, that produces a good exposure, no clipped highlights etc.

Now i take the same settings, same scene, ramp up the iso to 1600, i get blown out highlight or just a white frame.


I'm with you so far.


You say, that is because the camera ramped up the noise.


No. I said nothing of the kind. The point of this thread is that the blown highlights are caused by an increase ISO, not an increase in exposure.


Now mind you, since the scene, time, and other settings are the same, it is only possible that the camera did this after the shot was captured, because laws of physics don't allow the camera to produce light inside the sensor to increase the brightness.


Of course.


Now, the same experiment, but i use film this time. Two cams, running 100 and 1600 ASA film. Same settings. I get a good shot in one, a blown out/white frame in the other. Now since it is film, there is no in-cam post processing.


OK.


How exactly did the brightness as you call it increase given the amount of light falling on the medium is the same?

Because the film is more sensitive. You are asking the question as if you think something I said would is not in agreement with the premise, but you're wrong. Again, nowhere have I said that increasing ISO does not clip highlights. (In fact I've said the opposite). The fact that you think I have reveals your misconception that ISO is a part of exposure when in fact it is not.

[Daniel Browning]

That said, so now your argument is,


My position has not changed at all.


so now your argument is, the definition of exposure involves only two things


No, my position is that the definition of exposure involves only one thing, which I've stated repeatedly: the total amount of light per area.


iso does not affect it.


Yes, that is my position: if you change ISO by itself, you have not changed exposure at all. The only way to change exposure is to change the total amount of light per area.


Iso affects the brightness of the image and the exposure triangle is actually the brightness triangle by your definition.


There's nothing wrong with calling it the "exposure" triangle, especially for beginners who do not understand that ISO is not a part of exposure. They need things to be simple for them, so picking the ISO first and then helping them find the optimal exposure for that ISO is one way to make it easy for them to understand.


So basically you're arguing that exposure the way we or I use it is wrong, what i'm talking about is the brightness of the image.


Generally, yes. In most cases you mean "brightness of a given default raw conversion (tone curve and B/W clip level)", but "brightness" is a good way to shorten that.


But if i were to substitute the word brightness instead of exposure then my argument is correct?


Generally, yes.


The sensitivity of the medium involved affects only the brightness? Is that your argument?


Close. What I mean is this: people know that exposure affects brightness. They see that ISO affects brightness too, and so they think ISO is exposure. But it isn't, ISO is only brightness, not exposure. That's not to say ISO has no other effects: it has many effects, including read noise, pattern noise, quantization error, file size, histogram combing, etc.


If that is so, its just semantics.


It's not semantics at all. Would you call the 800mm f/5.6 a zoom lens? Or would you call the EF-S 10-22mm a long focal length? No. Both of those are the incorrect definitions. We have words like "zoom" because they have meanings. If everyone thought that "zoom" meant "long", then how would you communicate the concept formerly known as "zoom"? In the same way, it's important to understand the word exposure so that it is possible to communicate ideas related to the total amount of light per area.


And i can be just as adamant that i am right and show you umpteen number of examples as you can.


I'm sure you can find many examples of photographers who are mistaken about the definition of exposure. But how many authoritative references can you find? How many dictionaries and photography textbooks?


Doesn't really matter. Which is why my earlier phrase of a rose by any other name....


But there *is* no other word for the concept of "total light per area". Exposure is it. That's the word.


That said, if you're actually arguing that the sensitivity of the medium doesn't matter,


I'm not, and I never said that. Of course it still matters, it's just separate from exposure.


That said, yes, what a digital sensor does is to actually increase the gain. And said gain is why you have such things as S/N ratios and quite a few other kind of noise. Which i why i said digital makes things a bit more complicated.


Agreed.


Now again, if you're just arguing semantics, it has no point in the discussion.


Understanding the word exposure is not just semantics -- it's fundamental to photography itself. It's about light. Light is important to photography. When you pollute the word "exposure" by adding in things that are not related to light (such as ISO), then it will negatively affect your photography.


But assuming the whole point of this thread was to prove the other thread wrong by giving me definitions? Er, you're going to have to do a whole lot better than that.


No, the reason why I created this separate thread was to explain the meaning of exposure. I could not address the other more advanced topics until this relatively simple matter is put to rest.


So my original point still stands. "Exposure triangle."..... Sensitivity does matter.


Agreed. And my point still stands. ISO is not a part of exposure.


Now, if you will admit to the previous point and are still arguing about the exact word definition of the term exposure, this is going nowhere. It can go either way.


No, there is only one definition.


There is a reason why most people tend to use simpler terms and definitions, complexity can confuse :P.


I don't consider the definition of "exposure" to be complex at all. It's just the intensity and duration of light.


Which is especially the case when trying to explain the various kinds of noise in digital media. Which is what the other thread was going on about.


Then let's leave that in the other thread, shall we?


Try to understand this rather simple concept that i try to put across. There is a reason why we call it an exposure triangle.


Yes, there is a reason. But you don't understand what that reason is:


Because all three affect the final output, i.e. the photograph, (one assumes taking a photograph is why you're into this in the first place....).


Of course. And I agree with that. But you are trying to twist that into meaning that to change ISO is to change exposure, which is patently false.


Hence the long argument, which has lead to this thread which to be absolutely frank, if im right about the first part of this post, is quite useless because semantics never benefitted anyone, unless youre a professor of linguistics :P.

Again, it's not just semantics. Exposure is the *only* word we have that describes the total amount of light per area. Light is vitally important to photography. Therefore it is important to have a word for it and understand what that word means. When you fail to understand that word, it negatively affects your photography. The "high ISO = less noise" is just one example of that: in low light (fixed exposure), one should use the highest ISO possible (without clipping highlights) to reduce noise. There are other examples as well. If you're shooting a low contrast scene and you've increased exposure as much as possible, then it is better to increase ISO from 100 to 400 and decrease brightness in post by -2 EC then to leave it as ISO 100. That will result in more dynamic range, better color depth, and other benefits.

Those are just some examples of the benefits that come from understanding the word "exposure".

[synn]

Why are you making things up? I never said anything like the idea that increasing brightness with ISO does not blow highlights, or that they could be recovered. Increasing brightness with exposure does the same exact thing. I'll break it down for you:


Increasing ISO increases brightness.
Increasing shutter duration increases brightness.
Both increase brightness.
Both can blow highlights.
Only one of them is an increase in exposure.




What exactly did I make up? By your own definition, increasing/ decreasing exposure and SS results in the same end result, but apparently one is brightness and the other is erm... exposure? Two things look like ducks, walk like ducks and quack like ducks, but one of them is apparently a pigeon?

Like Delta said, you're going on and on about what is semantics. You go on and on about how textbooks, websites, Count blah and the martians agree with your definition, yet ignore the others that do not agree with it. Now imagine if you had put all that energy, patience and thirst for excelling at something into, I dunno... taking some pictures?

That really wold have been something, eh?

[Daniel Browning]

What exactly did I make up? By your own definition, increasing/ decreasing exposure and SS results in the same end result,


No, I did not say that at all. I said they both affect brightness.


but apparently one is brightness and the other is erm... exposure?


Yes.


Two things look like ducks, walk like ducks and quack like ducks, but one of them is apparently a pigeon?


That is a very poor characterization. The fact that you cannot tell the difference between real light entering the lens and electronic gain added in the camera is telling. If you could did the difference, you wouldn't say they both quack like ducks.


Like Delta said, you're going on and on about what is semantics.


No. Light is important. I amazed that I have to even explain that to you.


You go on and on about how textbooks, websites, Count blah and the martians agree with your definition, yet ignore the others that do not agree with it.


I do not ignore others who disagree. I explain to them how and why they are wrong.


Now imagine if you had put all that energy, patience and thirst for excelling at something into, I dunno... taking some pictures?

That really wold have been something, eh?


Imagine if you stopped spending all your time sticking your fingers in your ears say "nya nya nya" and actually realized that you were wrong and learned something new? That would really be something.

[synn]

Nyah, I prefer to keep on finger on the shutter release and the other around my lens and take some pictures that are worth my time.

[Vicky]

What is the real definition of exposure?

Exposure is the total amount of light per unit area that falls on the sensor.



Got it!

- Aperture & Shutter Speed are the only controls that can change the amount of light per unit area that falls on the sensor.
- A change of ISO cannot change the amount of light per unit area that falls on the sensor.
- Hence ISO is not a part of 'exposure'.

Thanks for the info and effort for the detailed writeup!

Did I learn something new today?
Yes. I learned the 'correct' definition of the word EXPOSURE.

Will this new info change my style of shooting or add to my photography?
Not really. But, I did learn something new today.. so, its still a good thing:)

[Daniel Browning]

- Aperture & Shutter Speed are the only controls that can change the amount of light per unit area that falls on the sensor.
- A change of ISO cannot change the amount of light per unit area that falls on the sensor.
- Hence ISO is not a part of 'exposure'.


You've got it! I would add that flash is the other common control that affects exposure. One of my cameras has three built-in ND positions -- that control also affects exposure.


Thanks for the info and effort for the detailed writeup!


You're very welcome.