The Exposure Triangle Is Misleading And Unnecessary.

This seems like another thread where the argument is over what is the definition of the word "exposure" in photography.

The definition of "exposure" is illuminance--time product.

Illuminance is controlled by two things - scene illumination and f-stop.
Time is time - shutter speed.

Therefore, by definition, digital ISO settings is not a factor in "exposure", at least directly.

Indirectly, setting your ISO adjusts the way the camera's meter chooses to set the exposure, if you are using an auto mode or the in-camera meter for a manual setting of exposure. The same is true of a hand-held light meter. So, indirectly, the setting of ISO can affect the exposure chosen, but that doesn't mean ISO is part of exposure. As shown above, it's not, by definition.

Too much to respond to all at once :-)

Langford doesn't use the phrase "exposure triangle" in his book, as I already mentioned. In the section Iliah referred to, in my edition I think Langford lists 8 different things to consider (including the illuminant). Elsewhere (I forgot to mark the page), somewhere about using a light meter he simplifies all this by stressing the importance of shutter speed, aperture and film speed (maybe because that is all you can set on most meters).

But to my mind, the concept of using those 3 parameters is basic introductory stuff, just to get you started. A fancy version of "Sunny 16". It doesn't stand up to intense scrutiny but it isn't meant to. It's just a starter metaphor. Perhaps the problem is that other people in internet world have run with it and blown it up beyond all reasonable application?

I don't think the idea was to provide a literal statement of photographic physics. If you are using film, you set a film speed, point your meter and the meter suggests combinations of shutter speed and aperture that will yield a well exposed image. You can use any of the combinations (until you hit reciprocity failure) and expect the same level of exposure. If you vary one of them and wish to retain the original level of exposure, you have to vary one of the other parameters. All three operate in lockstep. To me, it makes perfect sense as a quick and dirty rule of thumb metaphor for beginners.

The problems start with digital because the ISO knob isn't exactly the same as film speed. What the ISO knob does in practical terms as it isn't film speed is the adjustment I need to make to my out of date mental model.

Could you please provide a quote? I don't see "key parameter" occurring in my 7 and 9 editions.

Mine is the 1977 edition. I've lost the page but I agree he never said "key parameter" - that was my paraphrase.

EDIT: Ok, I found this: "On the face of it, measuring correct exposure might be thought to be (a) measuring total light reflected by the subject towards the camera (b) setting this reading against the manufacturer's film speed on the calculator dial, to find the required shutter setting at each aperture."

This is part of a longer section that considers more complicated things, but this is the essence of the exposure triangle idea, I think.

Because film processing is (supposed to be) fixed. Further, "well-exposed" means within latitude, and thus for the same negative film and processing one can vary exposure rather significantly and still get a "well-exposed" shot.

Processing isn't fixed with digital, especially when one shoots raw.

Interestingly, in some regards negative film is more forgiving than digital, partly because there is no forced clipping with the film.

Ninth edition is from 2010, still no mention of exposure triangle.

See quote in my response earlier. Anyway, I feel there is a danger in getting a bit lawyerly in this debate, which I feel would be missing the point. I don't know the historical origin of this phrase, but my feeling has always been it is something people interested in photography were taught about 5 mins after the Sunny 16 rule. This may be a complete fabrication of my sub conscious but that is how I feel about it. I also feel that for film it was a perfectly reasonable beginners aid, similar in status to Sunny 16. Looking at the web references to it now, I'm shocked to see it is now rolled out by just about everybody as if it were something found in the CMB. A bit overkill for a little convenience rule of thumb. But that's the web for you, copy what someone else said.

Getting back to what is the point from my perspective, now that we agree the ISO knob is not setting film speed and I'd better stop pretending that is what is does, what do I do with it?

Still fighting on too many fronts here, can't keep up!

I don't understand what you mean here. If one pair of aperture/shutter settings is to = a corresponding pair, that assumes film speed remains constant. But reciprocity causes effective changes to film speed at longer shutter speeds so the relationship that held for fast shutter speeds no longer holds for long shutter speeds. What's wrong with that?

I'm engaging in this thread because I want to correct my previous incorrect assumptions about what the ISO knob is for. I don't want to get into acrimonious debate or have to start backtracking and all that lark and I'm beginning to feel tense because it feels like I'm being asked to defend things I don't believe.

Here is the first sentence of my original Langford post for the avoidance of doubt:

"I have consulted my edition of Michael Langford's classic college text book "Basic Photography". He makes no mention of the term "exposure triangle". "

Here is the ISO perspective, reference number ISO 6:1993:
"Black-and-white films will generally provide excellent results in several
different developers and processing conditions. At the same time, it is
realized that the speed of a film depends on the process used. Therefore,
this International Standard specifies a method for determining the photo-
graphic speed of film/process combinations. This means a particular film
may have several I S 0 speeds associated with it depending on the pro-
cesses used. For this reason, it is important that manufacturers indicate
the processing conditions for which ISO speed values are quoted.

This International Standard recognizes that black-and-white films do not
generally have a unique speed if several different processes are rec-
ommended. This conflicts with the tradition of associating a specific speed
value with a particular product. In the future, the process used for deter-
mining speed values should be unequivocally described to avoid misinter-
pretation. Since users often do not know how these films will be
processed, manufacturers have an obligation to provide a speed value for
this situation which will ensure good results. Usually they will take ad-
vantage of the overexposure tonal latitude of the film and give it a con-
servative speed value to protect users from underexposure effects in case
the film is put through a process which yields low speed.

It is recognized that the speed at which a film can be exposed is de-
pendent on the extent of development, scene luminance range, subject
matter, printing paper, etc. This International Standard specifies that
film/process speed is determined when the film is processed to obtain a
specified contrast level. The relative ISO speed ranking of various films in
different process systems will generally differ. The I S 0 speeds will pro-
vide correct exposures for average scenes with exposure meters con-
forming to ISO 2720 or ISO 2721 when the film is processed as specified
in this International Standard.

For the purposes of this International Standard, the
following definitions apply.
3.1 exposure, H: The time integral of illuminance on
the film, measured in lux seconds.
Amounts of exposure are often expressed in logarith-
mic terms (log,&).
3.2 speed: A quantitative measure of the response
of the photographic material to radiant energy for the
specified conditions of exposure, processing and im-
age measurement.

5 Test method
5.1 Principle
Samples are exposed and processed as specified be-
low. Density measurements are obtained from the
resultant images to produce a sensitometric curve
from which values are taken and used to determine
ISO speed."

Ok, always good to standardise terms.

So what effect on the "density" of digital images does adjusting the ISO knob have? I have always pretended that the ISO works similarly to film speed, in that if I change ISO by one stop, the sensor behaves as if it had become one stop faster. As I now have learnt that isn't true, what does it do?

EDIT: to be more correct, this is a fairly new thing for me. Until recently, I was terrified of using anything other than base ISO because I was frightened of a drop in image quality. The result was in dark conditions I constantly used too low a shutter speed and my images were ruined by camera shake. About a couple of years ago I woke up to the existence of auto-iso. And now I get grainy but sharp images. That seems a reasonable trade off. But should I be using the ISO setting in a different way eg deliberately underexposing then pulling up exposure in post?

OMG, they said "overexposure" and "underexposure"! 😱

It causes a push somewhere downstream, either "intensifying" raw, or "intensifying" the conversion (push in a developer), or both.

It's about latitude, yes.