Re: What is Equivalence

Thanks for the clarification.

Equivalence is basically a translation service, that describes AOV, DOF, and diffraction, based on what f-numbers, focal lengths, exposure indices, will give the same look on a common standard, usually a FF sensor (with the same exposure time). Part of the reason that it needs to exist as a concept because people are already erroneously drawing direct correlations between the looks and the numbers. The "looks" actually derive from the size of the entrance pupil and its proximity to the subject matter; not from f-ratios, exposure indices, or focal lengths. When two systems (or two different crops) are equivalent, then the pupil size is the same.

Thinking aloud...

Seems to me, "equivalence" is a bass-ackwards approach to describing the differences between formats. I guess I'd rather first understand the differences and their implications, before I spent brain cells figuring out how to make "equivalent" images between them.

Bear-of-little-brain thinking, so YMMV...............................

If I may expand on that a bit:

ISO = k/Hm where Hm = the mean exposure on the sensor or film** - and where:

k=10 for film and the Standard Output Sensitivity Method

k=78 for the older Saturation method

k= anything the manufacturer wants for the Recommended Exposure method;

Guess which is the most popular with Manufacturers these days, LOL. Not possible to hold their feet to the fire with REI, eh?

All of which meaning that setting the ISO on your cameras is a craps-shoot and can bring the real ISO off by a large margin, IMHO, I'll say up to +/- 30%.

** the amount of light that hits the sensor or film while shutter is open: Danno 2023

They need to know that it is wrong. For example, "f/2.8" is an aperture; it may be unknown, because we may not know what "f" is, but if the lens is 28mm, then the aperture is 28mm/2.8 = ~10mm. This 10mm, combined with subject distance, describes how any object renders geometrically; not "2.8". With the aperture or pupil being so fundamental to the geometry of imaging, conflation of the f-number with "aperture" hides the real root source of the geometry of imaging. It also hides the real subject-normalized noise.

It's not just rank beginners that seem to be confused; I meet experienced wildlife photographers all the time who are obsessed with f-numbers and ISOs (as well as full-sensor-image-level and pixel-level noise). I've been in many situations where many photographers were gathered around a rare bird that was skulking in deep shade, and heard chatter about sensor and lens choices; people announcing that they are switching from their 600/6.3 lens to their 200/4 lens for "less noise", without getting any closer to the subject, or switching from crop mode to FF mode, or were using an APS-C camera and wished that they had their FF for lower noise, even though the APS-C frame is wide enough with their lens, and they'd just crop more with the FF.

Many people are beginners for a very long time, and die beginners, because they never had a useful model, or were not fluent in the translation service of "equivalence". The person who switches from 600/6.3 to 200/4 should at the very least do some basic equivalence math: the subject will be 1/3 the height and width on the sensor, so crop equivalence will be for a 3x sensor, giving imaging like 600/12 on the full sensor; not 600/4. For me, it is even simpler; 94mm vs 50mm entrance pupils; no contest; 600/6.3 wins from the same subject distance. 9x as many pixels on subject, and almost 4x as many subject photons per millisecond, at a higher ISO, which gives less read noise relative to photon noise.

With respect, equivalence isn't a theory. It's a fact of light, optics, and the photographic process.

It's not unlike taxonomy and the identification of species based on sets of shared qualities. American bald eagles aren't theoretical. They're a species of bird sharing certain physical traits that exist in nature. Equivalence isn't theoretical. Photos made with different format cameras and sharing six qualities in common are equivalent.

You're overthinking it. Equivalence doesn't require two photos to be indistinguishable. If photos made with different format cameras share the following qualities, those photos are equivalent:

Perspective
Framing
Exposure Time
DOF/Diffraction/Total Light Captured
Lightness
Display Size

Notice, there's no mention of the photos being identical. No mention of them being indistinguishable. That's not a requirement.

There's no requirement that equivalence be of use to every photographer. It's available should one have a need.

A common scenario in which one can benefit from equivalence, is when considering the suitability of different format cameras to the kind of photography one does. In this situation, the question naturally arises, "How do I make this photo with those cameras?" equivalence answers that question.

How would you "use" it? When we say that a FF sensor with a 28/2.8 lens has certain imaging properties at a certain subject distance, we generally understand that this tells us nothing about pixel count, read noise, quantum efficiency, lens aberrations, etcetera, but understand that it does refer to something very significant. Why should we then reject equivalences that also ignore these things? It is the basic geometry that derives from distance and pupil size that equivalence addresses, which should normally be the major determinants of imaging. After that, you can think about those other things, too.

The question went OK until the Real World got mentioned - so often used as a counter to proven theory ...

Boost of what? You get 1 stop more raw headroom, and a doubling of post-gain read noise, at ISO 320 and above, than you would have had if they kept analog gain proportional to ISO just like ISOs below 320..

ISO calibration and raw midpoint calibration are two different things, and ISO calibration isn't applicable to raw data.
If two cameras have different raw midpoint calibrations the ratio needs to be taken into account when making some of the comparisons between those two cameras. But as a raw shooter I prefer comparisons at the same exposure or integrated exposure

I guess I'm more a "a 300mm lens fills the frame with less FOV on a FF sensor than on a APS-C sensor" sort of thinker, not especially worried about what would it take to make the same image with both sensors, either with that example or exposure or anything else. I use both FF and APS-C cameras interchangeably during steam locomotive stalking, kit-lens focal lengths on FF, tele on the APS-C, and when I pick up either camera/lens for a shot I'm concerned with what that particular camera/lens is going to do, not what the other might do in the same situation.

When I've made purchasing decisions, it's been first about dynamic range based on my personal experience. Mapping that and other considerations to a notion of equivalence just hasn't seemed useful.

Am I being too stupid-simple about this?

Are you sure about that?

Nor is there any evidence known to me that they are ... anybody have more than supposition?

Sorry, I'm struggling with that - what is "matrix metering" pardon my ignorance - any references?

By some factor ... and yes to the rest ...