Is there a "Resolution Triangle", like the "Exposure Triangle"?

[xpatUSA]

The question is asked because discussions, sometimes heated, often seem to involve just one variable or, at most, two. For example, a post might be entitled: "Which is better, 12MP or 24MP?" and we can guess the kind of discussion that would lead to! Or, elsewhere, I read yesterday that a 24MP micro-4/3" camera would not be diffraction-limited! Sometimes, in such discussions, contrast gets a mention but often not.

Lately, the thought occurs that any discussion on camera resolution is incomplete without due consideration being given to:

pixel pitch (not MP)
aperture setting
acceptable MTF (as output by the sensor to in-camera processing)

Why pitch and not MP? By example: for 12MP: a 1/2.3" sensor has a 1.5um pixel pitch; a FF sensor has an 8.5um pitch. Which could explain why you can't set your compact camera to f/22! (diffraction, etc).

Acceptable MTF must be included because, even with a perfect lens, there are many combinations of pitch and aperture that give less than 3% MTF which means invisible details in your image.

These three factors include all of the other factors that are so often brought into a discussion and that usually have the effect of confusing the issue.

This opinion matches the concept of the Exposure Triangle almost exactly. Just as it can be argued that ISO has nothing to do with exposure per se, so the introduction of "acceptable MTF" may raise an eyebrow or two.

To be pedantic, one should include lens quality as a consideration. How that could be done is open for debate - here if you like. There is a calculation for the effect on MTF (due to lens manufacturing tolerances expressed as a fraction of a light wavelength) but it is a local-only value which does not account for the dreaded sagittal and meridional variations.

Any thoughts?

[mknittle]

This is going to be interesting.

[jcuknz]

It may be basic but 'MTF'? What is MTF
dpreview used to provide a figure which related the sensor size to pixel total which I found interesting becuase the cameras I respected had quite low figures such as 25 [ Canon G's and FZ50's ] whereas the higher Mp cameras were up in the 40's ... of course DSLRs were down in the single figures, as was my Canon P&S s20. Not knowing anything about such tech stuff I wondered if that accounted for, or helped, me in those early days to make acceptable 15x12 inch prints from its 3.3Mp sensor.
After awhile dpreview stopped providing that and I wondered if it was too revealing .... or misleading?

Which could explain why you can't set your compact camera to f/22! (diffraction, etc).
Surely there is no need becuase the P&S's f/8 equates to the DSLRs f/22

[xpatUSA]

It may be basic but 'MTF'? What is MTF

https://www.cambridgeincolour.com/tu...resolution.htm as if you didn't know but I put a link in the OP, just in case.

dpreview used to provide a figure which related the sensor size to pixel total which I found interesting becuase the cameras I respected had quite low figures such as 25 [ Canon G's and FZ50's ] whereas the higher Mp cameras were up in the 40's ... of course DSLRs were down in the single figures, as was my Canon P&S s20. Not knowing anything about such tech stuff I wondered if that accounted for, or helped, me in those early days to make acceptable 15x12 inch prints from its 3.3Mp sensor.
After awhile dpreview stopped providing that and I wondered if it was too revealing .... or misleading?

Interesting indeed - do you remember what that relationship was? What was "size" - was it width, height, or area?

"Which could explain why you can't set your compact camera to f/22! (diffraction, etc)"

Surely there is no need becuase the P&S's f/8 equates to the DSLRs f/22

This response which talks only about apertures, proves my point I reckon.

[dc197]

it's helpful to define terms rather than making people hunt them down. This is taken form the link provided.

MTF: MODULATION TRANSFER FUNCTION
A Modulation Transfer Function (MTF) quantifies how well a subject's regional brightness variations are preserved when they pass through a camera lens.
https://www.cambridgeincolour.com/tu...resolution.htm

[GrahamS]

it's helpful to define terms rather than making people hunt them down. This is taken form the link provided.

MTF: MODULATION TRANSFER FUNCTION
A Modulation Transfer Function (MTF) quantifies how well a subject's regional brightness variations are preserved when they pass through a camera lens.
https://www.cambridgeincolour.com/tu...resolution.htm

That won't teach anyone anything. Try this: http://www.kenrockwell.com/tech/mtf.htm

[AB26]

That won't teach anyone anything. Try this: http://www.kenrockwell.com/tech/mtf.htm

If you'd like to learn how to make great photos, don't bother with these technical articles, instead read good books or take a local photo class. Your camera doesn't matter if you know what you're doing, and if you do know what you're doing, a better camera just makes it easier to get the results you demand.

Very interesting point.

[jcuknz]

I side with Andre here after going to the link and discovering what MFT is and it appears to me to have very little to do with photography as I know it which is recording light in a meaningful way to express an idea or whatever. Not to denigrate the geeks/boffins who understand and think about such things, without them to make the wonderful tools I have I wouldn't be able to do nearly as much as if I was mixing my own emulsion and spreading it over my sensor etc etc

The relationship is the area of the sensor to the number of pixels ... pixel size which I guess we all know is why the larger sensored cameras are better, in some respects, than smaller ones. The s20 is only a 3.3Mp camera and despite being a P&S had a reasonably large, for a P&S, sensor. The G's and the FZ both were moderate 10Mp cameras. It would seem to be a 'pixel pitch' judgement? But it does ignore improvements in sensor design if one is comparing sensors of different ages.

I almost added a comment about the practical difficulties of organising a reliable f/stop mechanism that small to the f/22 subject. A situation where step motor controlled 'Waterhouse Stops' might be the answer. [ lets ignore diffraction for a moment and sorry I am straying off the subject there as I often do ]

[xpatUSA]

I would like to try and explain the purpose of the OP because it is obvious that posters are not seeing the analogy with the exposure triangle. Instead, in this thread, experienced photographers are basically saying that a "resolution triangle" is unnecessary and that MTF is too technical.

But tutorials for beginners often make reference to the exposure triangle, which includes ISO - a subject not for the fainthearted (try reading the ISO publication or the thousands of posts on the subject). So, having taken a pic or two, I could make a similar negative comment about the exposure triangle that an understanding of ISO is not necessary for good photography.

By equating a "resolution triangle" with the exposure triangle I was hoping to show that the three aforementioned variables are necessary for a good understanding of resolution. Clearly, I failed in that endeavor.

[HenkB]

The endeavor was noble, I assure you. I am sorry it seemed to provoke resistance. I wonder if, for many, it seems somehow too removed from the act of taking a picture. As tool users we tend to achieve understanding of our tools through use as much as (or more than) understanding the tools before engaging in their use.

The exposure triangle comes into play directly for each and every exposure we make: it's unavoidable - either the shooter or camera (for most probably more often the camera, truth to tell) must manage it to get the desired exposure. On the other hand, at first sight, it looks as if such a consideration of resolution is the sort of thing one might engage in only when weighing the purchase of several competing lenses, much akin to the study involved in choosing a camera. One reader might observe that there are usually enough reviews scattered around the net that individual study on this scale is not entirely necessary. Another might relish the opportunity to cone down on the numbers.

On further reflection, being able to predict a lens' resolution is part and parcel of one leg of the exposure triangle. Understanding a lens' performance can have as profound effect on an image (i.e., choose an aperture) as an understanding of motion (i.e., choose an exposure speed). It's not just depth of field and bokeh. I suspect that is how the practice of bracketing arose - each bracket set is a discrete experiment.

The unvarnished truth is is probably that most shooters have internalized enough rules of thumb encapsulating such considerations that, in view of their results, seem reliable enough. Consequently consideration of the underlying principles at this resolution (sorry, I couldn't resist) can be seen as relatively remote from the task at hand. And, when out shooting, they might prove to be such. Once back at home, however, considering the result and wishing to have better small-scale detail in the image, this is just the sort of study that will help refine our field rules of thumb to yield more satisfactory results. That, and optimal shooting discipline in the field.

With your proposal of a resolution triangle I have little quibble. It would certainly give form to lens choice, especially if one had enormous numbers of lenses of overlapping focal lengths to choose from. However, in use, I imagine we would use the theoretical considerations to identify candidates and select our favorites from amongst those by using them and evaluating the results. A major difference would seem to be that, unlike the exposure triangle, at least one leg of the proposed resolution triangle is not easily manipulated - that is, our choice of camera fixes the pixel pitch to a static value. The exposure triangle is endlessly fascinating and vital since each leg is independently variable and the settings must be computed from the triangle's focus, the desired exposure. Desired, i.e., artistically optimal, not necessarily technically optimal, although a strong argument can be made that maximizing data puts one in the best position from which to make subsequent artistic decisions if they must be deferred at the time of capture.

[xpatUSA]

Thank you, Hendrik, for a well-reasoned response.

Indeed, the proposal was never intended to be a tool for shooting per se which would be most impracticable, as you clearly say. It was intended to be more of a aid to discussion - rather than yet another set of rules deemed necessary for the savvy photographer to learn.

However, as stated in Hendik's response, the exposure triangle has independent variables whereas the proposed resolution triangle does not. The term "desired MTF" is a bit of a stretch, even to the author's eye. In other words, speed, aperture and ISO result in exposure - but pitch, aperture and MTF do not really result in resolution, rather a change in spatial resolution results in a change to MTF. So, the analogy is back-asswards and fails anyway. :-(

Curiously enough, I do have an in-camera choice of pixel pitch, albeit limited. Sigma DSLR's have true pixel-binning capability i.e. right there on the sensor chip. Thus I can shoot at 9um or 18um pixel pitch.