dprevived.com/t/the-exposure-triangle-is-misleading-and-unnecessary/1602/post/14465/
April 10, 2023
638
-
-
It has a meaning in the photographic context, but is commonly misused.
-
Thanks, I'll re-read.
p.s. my last couple of replies seem to have lost some of my text.
-
1. That aperture, shutter speed and ISO are co-equal independent variables that together make up 'exposure';
Is this not a particular use of the word "exposure"? In my mind "exposure", as you use it here, refers to the total incoming light received by the sensor - which is determined by the brightness of the light and how long it is received i.e. the brightness of the subject as seen by the camera, the aperture used to restrict this and the time the shutter is open. But a lot of people might think of the exposure as being the end result ie how dark or light the picture ends up being. As in "this exposure is too dark, you need to use a wider aperture". In this latter case, this is also affected by how sensitive the sensor is, its "film speed" as you could replace "use a wider aperture" with "increase ISO". With this usage the exposure triangle concept seem legitimate.
2. When coupled with the idea that "exposure" is an amount of light - that ISO, like aperture and shutter speed governs an amount of light;
What is exposure if it not an amount of light?3. That is affects the amount of light by 'gain' - that is, the untutored beginner assumes - is that it creates more light, or -
4. That it makes the sensor 'more sensitive' so that somehow it sucks in more light (?);Don't understand what is meant by 3 and 4 - can you clarify
5. That the penalty for more sensitivity is that the 'amplification' causes noise.
Hardly surprising people think this - it is pretty much what every camera review says. Noise increases with rising ISO. I'm guessing that what you object to is the idea that the increased noise is the result of increased amplification rather the image being more underexposed and it is underexposure rather than amplification that causes the extra noise?6. As an alternate branch from 1, given that there is a 'correct exposure', that a certain amount of 'exposure' is 'needed' to make it correct, and back to 3.
I'm not sure what you are saying with this one. The way I think about "correct exposure" is that it is something that makes the picture look like you intended ie there is nothing wrong with a dark picture or a high key picture if that is how you want it to look. But I guess most people would think correct exposure is when you have good shadow detail, good highlight detail without clipping and the mid-tones are balanced to give a realistic looking bright image. How is this related to ET? -
I put exposure in quote marks, because this clearly isn't exposure, but if you put these three things in a triangle and call it the 'exposure triangle' then people very reasonably reach the conclusion that exposure is composed of those three things. It's wrong, but it's what the 'triangle' is telling them.
That is what it is (strictly, per unit area). So now, if you follow the triangle, you have these three things which together make an amount of light, therefore ISO must be an amount of light. What else can it be? Again, wrong, but the only way to consistently understand the triangle.
Well in (2) we reached the reasonable (but wrong) conclusion from the triangle that ISO is an amount of light. How can that be? It must mean that somehow ISO makes there be more light. That could either be by 'gain' - making more light, or by 'sensitivity', somehow sucking in more light. Again, wrong, but the only way of consistently understanding this triangle.
It is what they all say, and it's the result of confusing correlation with causation. Yes, noise appears to increase when you increase ISO, because increasing ISO sets a smaller target exposure. If you centre the meter, or let the camera do it for you, the result is a smaller exposure, which means fewer captured photons, which in turn means a lower shot-noise signal to noise ratio. That's the process, it is the aperture or shutter that caused there to be 'more noise' by way of lowering exposure, not anything to do with 'amplification'. Most triangles include this incorrect notion, that a higher ISO causes more noise. I'm also not keen on the idea that 'underexposure' causes noise. The amount of noise goes directly with the exposure, it matters not a jot whether you think it's 'under' or not.
In film days 'correct' exposure had nothing to do with how light or dark the picture looked, because that was set in the printing stage. It would affect the tonal range that you got. 'Correct exposure' was defined with respect to the characteristic curve (as Langford says). That doesn't make much sense with respect to the ISO control on a digital camera. But the key to my comment was the scary word 'correct'. There can be but one 'correct', and you 'need' that exposure for you to get it correct. The whole idea that you 'need' a specific amount of light is wrong headed. You 'need' as much as you choose, and that choice is based on your own priorities as a photographer. Or you look at it another way, if your exposure isn't 'correct', you can make it so just by using the ISO dial.
My intention was to show how the triangle leads to fallacious lines of thought if you don't have enough prior knowledge to know where it's misleading you.
-
It's a bit confusing for beginners that photon noise is directly proportional to the square of the exposure, but photon SNR is inversely proportional to the square of the exposure.
-
Indeed, and the kind of thing you don't really have to know unless someone has told you something to the contrary and then it has to be explained why that is wrong.
-
This is interesting. I don't think I'm conceptualising the ET as you are.
I think by "exposure" it means "the end result you get in terms of how dark or light your colour slide is, depends on 1) The amount of light your lens lets in controlled by the aperture setting 2) How long your shutter lets that light hit the film controlled by your shutter speed 3) How sensitive your film is (measured by its film speed rating). These three things are traditionally measured in units of doublings or halvings called 'stops' and you can change any one of these by any number of stops as long as you change another by the same number of stops in the opposite direction and get a result that is exactly as dark or light as the original".
I don't have any problem with this meaning in relation to film. Clearly, if films can be made that have different sensitivities, they won't all need the same amount of light to record an image. Therefore you can reduce the amount of incoming light with a more sensitive film and get the same density in your colour slide.
It may not work with digital sensors, but as we have already established I'm on shaky ground with my understanding of "sensor speed" and the effect of the ISO knob, so that will have to wait for improved understanding (that I'm hoping someone on this thread will provide me with. It'll take a while, not very fast :-) ).
I thought the misunderstanding about sensor speed was all about what analogue amplification does. A transistor appears to take in a small signal and output it as a big signal, so why can't the ISO knob take a small light signal and boost to a big one, therefore allowing the sensor to do more with less light? It might not be how sensors behave, but it seems intuitive based on analogies to experience with other forms of amplification, no? Is it that unreasonable to jump to that conclusion?
If the amplification doesn't increase the noise when boosting the signal as is commonly described, what does it do?
With colour slide film, the lightness or darkness of the slide is very much connected to how much light lands on the film.
-
"In film days 'correct' exposure had nothing to do with how light or dark the picture looked, because that was set in the printing stage. "
Im shooting film atm and i disagree with this statement. because alot of film photographers will set the exposure comp/asa settings if the camera model has a tendency to over or under expose the film eg: olympus trip 35 every one sets +1/3 stop exposure comp. (funny thing that... u use the asa to set exposure comp 🙄) yet apparently asa has nothing to do with exposure 🤨
-
Every somehow credible source explains it right, though, and only relates aperture and shutter speed. This here is personal agenda & crucification.
People are acting like Venn diagrams and other illustrations wouldn’t exist.
-
agree David, its everywhere and everyone is stupid just like us 😁lets even make the subject more stupid. ive watched the whole thread evolve and not one i mean not one has even mentioned "T" stop eg the olympus 45 1.2 is actually f1.8 . yet we are all so accurate with our knowledge 🙄
-
can someone please explain how i can reply in the middle of another persons post 🤨
-
Highlight the text you want to reply to. Then you should see word Quote in rectangle. Click on it.I
-
thats my understanding as well and your assumption is mostly correct.
-
hooray i stumbled on it by accident. ive been hitting the quote button first then highlighting it . instead of just highlighting it . which is kind of backwards 🤨maybe im a bit twisted 😵💫
-
i experimented by cutting and pasting the {quote="@] and {/quote] tags around the bits I was interrupting. It seemed to work. Although it has messed up this one! (Tried a different quote to hopefully fix it)
EDIT: There's an editor function! Yay, too used to coding by hand.
-
How are you processing and printing your film, Don?