Talk:F-number

N=f/D is unreliable as a definition of f-number and most often only an approximation. For instance N is not equal to f/D in the typically shown thin lens diagram, as explained by Kingslake: "It is a common error to suppose that the ratio D/2f is actually equal to tan(theta), and not sin(theta)"^[1]. So assuming f/stop = f/D maybe fine as a rule of thumb in the field, but fails when put to the test. For instance it may induce people to assume that N can be made arbitrarily small simply by making lens diameter arbitrarily large. Or that one could get an arbitrarily low equivalent f/stop by mounting a reference lens on an arbitrarily small sensor. Both of which are physically impossible for a practical photographic lens.

N is on the other hand always equal to 1/[2sin(theta)] in air for a practical photographic lens that produces reasonable images, per Kingslake and others^[2]. Knowing the precise definition of N allows a photographer to decode advanced optics formulas and to answer questions like "what's the smallest f-number possible for a reasonable photographic lens?" (Obvious answer: minimum N = 0.5 in air). That's not possible with the article as it stands.

Therefore I would suggest specifying in the initial definition that N = f/D is a useful approximation that works fine most of the time in the field. But I would also add a section with the precise definition of f-number N = 1/[2nsin(theta')], with n the index of refraction of the medium between the lens and the image plane. That's air in interchangeable lens systems, so n=1. It should then be apparent that N = 1/2NA exactly, with NA = Numerical Aperture.

Can page admins make the relative changes? Jack Hogan (talk) 16:54, 3 March 2016 (UTC)

Well, we should check the appropriate references, but I have always seen the f-number defined as N = f/D. There is no contradiction with N = 1/(2NA) because, for any practical lens, D/f = 2 sin(θ).

The above formula obviously doesn't hold for the thin-lens diagram, but a thin lens is not a practical photographic lens, it's just a very simplified representation which is consistent with the small-angle approximation. So, in the small-angle limit, sin(θ) ≈ tan(θ) and the thin lens diagram can be trusted. For fast lenses, N = f/D = 1/(2 sin(θ)) still holds as per the Abbe sine condition, but the thin-lens (which implies D/f = 2 tan(θ)) is wrong.

— Edgar.bonet (talk) 09:44, 4 March 2016 (UTC)

Addendum: Here is a derivation of f/D = 1/(2 sin(θ)) based on the conservation of etendue. And here is a serious comic discussing why you can't make a lens arbitrarily fast. — Edgar.bonet (talk) 09:55, 4 March 2016 (UTC)

Seems to me that f/numbers were defined when lenses were small enough, and focal lengths long enough, that the difference between sin(θ) and tan(θ) wasn't significant. Now that you mention it, I am not sure how it is actually done in the lenses where it is significant. Well, the amount of light collected should be proportional to the area of the lens. That is, the solid angle as viewed by the subject. Even more, note that the ratio does not correctly follow the light delivered to the image when focused closer than infinity. Except for extreme close-ups, it is usually close enough, and those doing close-ups know to correct it. Even more, with through-the-lens metering, it mostly doesn't matter. Gah4 (talk) 07:36, 20 July 2016 (UTC)

Numerical_aperture#Numerical_aperture_versus_f-number explains tan(θ) vs. sin(θ). Gah4 (talk) 16:18, 13 September 2018 (UTC)

On a camera, the aperture setting is usually adjusted in discrete steps, known as f-stops. This was usually true from the rangefinder days through the manual focus SLR days. Though there are detents at full stops, the aperture changes continuously. The later electronic SLRs, and continuing into digital SLRs, normally allow steps of 1/2 or 1/3 stop. The settings are electronic, and not continuous. Seems like we should at least take out the usually. Gah4 (talk) 07:55, 20 July 2016 (UTC)

Thanks, page useful. Will read in detail, but at this point just trying to help daughter into using the loaned camera. THe split image showing effect of f-number - great. Below a just as useful image with caption. "Shallow focus with a wide open lens" - if this just said (high f-number) or (low f-number) it would be great, much easier to grasp by those reading the page to learn. I totally, understand that if I already knew this all, read the page thoroughly, it would be redundant infromation, but that's why I'm reading it. As a wonderfully informative page, it would be even easier with this consistency of comment in the caption as well. — Preceding unsigned comment added by 122.60.105.241 (talk) 23:53, 25 August 2018 (UTC)

As the comments above indicate, photographers pretty much never say high or low, or increase, or decrease, the f-number. Expressions like wide open, open up, or stop down are common, and everyone very quickly learns that open up is a smaller number written on the lens, and stop down is a larger number. Even more, though, photographers pretty much always write it in the f/(number) fraction form, such that increasing (number) decreases the fraction. There just isn't enough room for the f/ on the lens barrel. Gah4 (talk) 16:26, 13 September 2018 (UTC)

Can someone rewrite this section? I understand many aspects of photography, but I did not understand this section. Could someone please write it in layperson language, clarify what is being said, and expand it a bit to make it more useful? because it looks like it has the potential of being a very useful section! Misty MH (talk) 03:43, 10 November 2018 (UTC)

Doesn't look so bad to me. f/number is defined as the focal length divided by the (effective) lens diameter. (Entrance pupil in more complicated lenses.) That gives the appropriate amount of light when the lens is focused (and the object is at) infinity (or close enough). But as you focus closer, and the lens moves farther from the image plane, the light gathering ability reduces. It is, pretty much, based on the new lens to image plane distance. Most of the time it can be ignored, but with extension tubes or bellows, the lens can be much farther away, and the correction gets more important. That is why it is commonly the bellows correction factor. Gah4 (talk) 04:31, 10 November 2018 (UTC)

It doesn't look so bad because I edited it between when he posted and when you replied.--Srleffler (talk) 16:08, 10 November 2018 (UTC)

Regarding the correction for objects not at infinity, with TTL (through the lens) metering systems, this will automatically be corrected. Should this be mentioned? Gah4 (talk) 04:33, 10 November 2018 (UTC)

The article says Most old cameras had a continuously variable aperture scale. I suppose this depends on how old you mean by old. Many box cameras from the early 1900's have a sliding metal tab, or rotating metal disk, with holes in it. The continuous aperture might have been from about the 1930's or so. Many cheaper cameras only offer one aperture setting. With electronic aperture control, the continuous scale went away. Gah4 (talk) 04:46, 27 March 2019 (UTC)

Actually, even modern electronically controlled lenses have an essentially continuously variable aperture setting. I don't recall the details, but I seem to recall the increments are tiny; it's just the camera UI that quantizes them coarsely. But more generally, you're right, there has always been a mix. But for cameras with iris diaphragms, they were usually continuous. Dicklyon (talk) 04:34, 12 August 2019 (UTC)

@Rkinch: Regarding this edit, please provide a page reference in the book cited, and a quote. I do not believe the new definition you have put in the article is correct.--Srleffler (talk) 03:48, 12 August 2019 (UTC)

Nevermind. From Google books snippets it's clear that Smith defines "clear aperture" as a synonym for "entrance pupil". Since we have an article on the latter and it is the more common term in optical design, I'm reverting your change.--Srleffler (talk) 03:52, 12 August 2019 (UTC)

From p. 183 in Smith's book (emphasis added):

...energy collected from a small area of the object is directly proportional to the area of the clear aperture, or entrance pupil, of the lens. At the image, the illumination (power per unit area) is inversely proportional to the image area over which this object is spread. Now the aperture area is proportional to the square of the pupil diameter, and the image area is proportional to the square of the image distances, or focal length (f). Thus, the square of the ratio of these two dimensions is a measure of the relative illumination produced in the image.

The ratio of the focal length to the clear aperture of a lens system is called the relative aperture, f-number, or "speed" of the system, and...

It's not wrong to define "clear aperture" as a synonym for "entrance pupil" as Smith has done, but if you want to do that you have to define the term. Linking it as "clear aperture", as in the edit I reverted, is simply wrong. The entrance pupil diameter ("clear aperture") is related to the physical aperture diameter, but they are not the same thing. The f-number is defined in terms of the former, not the latter.--Srleffler (talk) 04:19, 12 August 2019 (UTC)

"Clear aperture" is the essential concept, the sound and logical foundation, and why Smith uses that term, as should the WP article. "Entrance pupil" is a more complex concept dealing with multi-element designs and the theory of stops. "Entrance pupil" is not the concept underlying relative aperture, although of course it does determine the *calculation* in multiple elements. The two terms are not equivalent (aperture != pupil, else why would we have redundant terms), although the evaluation may be. We don't call it "relative pupil". In terms of an encyclopedic approach, the intro should state the concept in its essential terms (aperture vs focal length), and then the detailed section can refine how the pupil may differ from the physical aperture. An intelligent but non-expert reader is not going to understand "entrance pupil", but "clear aperture" is immediately understandable. Intro'ing with "entrance pupil" is going to lose most readers at the start. The article is a mess of confusion because it's trying to speak in terms of photography some places and geometric optics in others, and this pupil/aperture conflict is one case of that confusion. You and I as editors may understand the difference between a pupil and an aperture, but we cannot assume that expertise in whoever is reading the intro. The biggest fault in this article is that it correctly defines the f/number as a property of an ideal lens (circular apertures, zero aberrations), and then proceeds to indoctrinate the idea that real lenses have a perfectly precise, one-dimensional constant f-number, which they don't. This is similar to the confusion of resolution with MTF, likewise a reduction of the truth. Richard J Kinch (talk) 22:14, 13 August 2019 (UTC)

Something like "clear aperture (entrance pupil) diameter" might be a good compromise? Dicklyon (talk) 18:59, 12 August 2019 (UTC)

Yes, that would reserve the distinction and avoid the reduction, which could be explained later in detail. Perhaps qualify entrance pupil with the word "typically". Richard J Kinch (talk) 22:14, 13 August 2019 (UTC)

I don't understand what you mean by "typically". I am interpreting clear aperture as a synonym for entrance pupil, and as distinct from aperture. If you are interpreting it differently, what definition are you using? I am starting to regret taking Dicklyon's suggestion—my concern from the beginning was that clear aperture is an ambiguous term that is not yet defined in the article or elsewhere on Wikipedia, while entrance pupil is clearly and specifically defined. You write above that "clear aperture" is "the essential concept, the sound and logical foundation". Before it can be a sound foundation for anything, it needs a clear definition.--Srleffler (talk) 00:50, 14 August 2019 (UTC)

Clear aperture and entrance pupil are not synonyms! When Smith (Modern Optical Engineering, 2nd ed, p142) says, "... clear aperture, or entrance pupil, ..." he is distinguishing the two terms, not equating them. Smith explains clear aperture as the clear diameter of a [circular] lens or diaphragm (p133). That is, whatever central-circular area of a lens or stop which is not obstructed by hardware, defects, imperfections, etc. This is the essential factor in the definition of relative aperture in terms of a simple lens. In the case of a multielement system with refraction ahead of the aperture stop, then the image of the aperture stop (the entrance pupil), not the clear aperture, governs. So the definition is conditional, but still stated in terms of the essential case. At that point it becomes a philosophical quibble on whether a clear aperture is a degenerate case of a pupil, or a pupil is a glorified case of an aperture, but however you resolve that non-problem, an abstract ratio must be founded on real factors, and that means only clear aperture truly defines relative aperture, with the "or" proviso about the abstract image (pupil) in the multielement condition. If the WP article doesn't define CA, then it should, instead of referring to the EP that may not apply and is too advanced a concept for an introduction. This involves the theory of stops (on which Smith has a chapter), and I can testify that expert photographers and even optical engineers don't get that delicacy of a theory. Smith's textbook is a genuine authority, but it's not perfect on this subject: I once spent hours puzzling over his treatment of "Field lenses and relay systems" described by "CA", which doesn't appear in the index, until I recognized he was abbreviating "clear aperture", also not in the index. By "typically" I mean in the case of camera lenses, which have entrance pupils differing from the clear aperture, since the article seems to be about camera lens f-numbers and not the theoretical optics. Richard J Kinch (talk) 02:51, 14 August 2019 (UTC)

If they aren't synonyms, then defining f-number in terms of clear aperture is simply wrong. You're defining a general property of an optical system in a way that is meaningful only for the case of a singlet lens. F-number is widely used for multi-element lens systems. Defining it in a way that doesn't work for such systems makes no sense. The f-number of a camera lens is the ratio of its focal length to the diameter of its entrance pupil. You mention "referring to the EP that may not apply", but the situation is exactly contrary to that: EP applies to any optical system, while by your definition CA only applies to one special case.

The entrance pupil of any optical system (including a singlet lens) is the clear aperture of that system in the sense that it is the largest diameter at which rays entering the system are able to pass through. It is only in this sense that one can say that clear aperture is used to define f-number. --Srleffler (talk) 03:39, 14 August 2019 (UTC)

Some books do treat them as synonyms (not a use I was familiar with, I admit); , . Is it just in simple telescopes that they talk this way? I don't know. Dicklyon (talk) 06:50, 14 August 2019 (UTC)

The terms are not synonyms because they can't be freely interchanged. Not every clear aperture is an entrance pupil. That they measure as equivalent in some systems (which is what your refs are stating, as a treatment, not a generalization), does not mean that the terms are synonymous. The term clear aperture defines relative aperture, because that's the cross-section of a parallel beam transmitted by the system, not any object in the system itself. Another confusion is that these terms are all referring to the abstract ray geometry, not physical items as is often wrongly assumed. Beware reification. Same reason for "iris" (real object) vs "pupil" (an abstract property). You can't take a lens apart and point to the pupil, clear aperture, or focal length; these are not physical objects, but abstract properties. A donut is a physical object, but the outside diameter or hole in the middle is not. Likewise temperature or heat in thermodynamics, not physical objects (although once misunderstood to be). Richard J Kinch (talk) 02:20, 16 August 2019 (UTC)

Note that MIL-STD-150A defines relative aperture and f-number in terms of the effective aperture: "The effective aperture of a photographic objective for distant objects, for a given setting of the diaphragm, is an opening equivalent to a right section of the largest beam of parallel light from an axial object point that is transmitted by the lens. It is usually circular, or approximately so, and is specified by its diameter. If the section is not circular, the effective diameter shall be the diameter of a circle having the same equivalent area." This solves several definition troubles in the WP article: f-number is defined for distant objects on axis, what's the f-number for non-circular pupils, how it applies to a telecentric lens that has no real entrance pupil, etc. The article should explain how f-number modifies off-axis or at finite conjugates, and specify efl instead of "focal length". The article is also completely backwards on notation: f-numbers are a decimal number N, like on a camera lens; the hooked-f form f/N indicates the inverse relative aperture, not f-number. And 1:N is also seen on lenses. Richard J Kinch (talk) 08:14, 15 August 2019 (UTC)

That's a good definition. The "effective aperture" they describe is exactly the entrance pupil, although maybe it's the case that as you move off-axis the "effective aperture" is a partially occluded entrance pupil. It would be interesting to find a source that makes such a distinction. Dicklyon (talk) 16:11, 16 August 2019 (UTC)

The effective aperture is strictly defined by an axial point source at infinity (parallel rays). Off-axis sources have an aperture that is effective, but that is not the formal characteristic whose singular value is termed "effective aperture". Richard J Kinch (talk) 08:18, 20 August 2019 (UTC)

I think Dick is right; for a given diaphragm (aperture stop) setting, the diameter of the largest possible axial bundle of parallel rays is just the diameter of the entrance pupil. --Srleffler (talk) 00:33, 24 August 2019 (UTC)

As for the f/N notation, we've been over that many times. It's sometimes a mathematical expression for diameter, and sometimes just a notation convention for f-number. More often the latter. Dicklyon (talk) 16:13, 16 August 2019 (UTC)

The ATM machine reads the PIN number with a camera lens with f-number f/8. Richard J Kinch (talk) 08:18, 20 August 2019 (UTC)

A recent change removes photography as the use of f/number. As well as I know, optics other than photography, uses Numerical Aperture, symbolized as N.A. In most cases, N.A. is twice the denominator of the f/number. (That is, if you think of it as a fraction.) Is there a WP:RS that shows it widely used in optics, other than photography, or I suppose when photographic lenses are used in non-photographic applications. Gah4 (talk) 01:22, 9 August 2020 (UTC)

Any of the first three references in the article suffice. Note that it's not a "recent change", but rather a partial revert. The implication that f number is only used in photography was inserted yesterday. --Srleffler (talk) 16:41, 9 August 2020 (UTC)

More specifically, this place:

″A 100 mm focal length f/4 lens has an entrance pupil diameter of 25 mm. A 100 mm focal length f/2 lens has an entrance pupil diameter of 50 mm. Since the area varies as the square of the pupil diameter, the amount of light admitted by the f/2 lens is four times that of the f/4 lens. To obtain the same photographic exposure, the exposure time must be reduced by a factor of four.

A 200 mm focal length f/4 lens has an entrance pupil diameter of 50 mm. The 200 mm lens's entrance pupil has four times the area of the 100 mm lens's entrance pupil, and thus collects four times as much light from each object in the lens's field of view. But compared to the 100 mm lens, the 200 mm lens projects an image of each object twice as high and twice as wide, covering four times the area, and so both lens produce the same illuminance at the focal plane when imaging a scene of a given luminance.″

The excerpt's second paragraph doesn't mention what f-number of the compared 100mm focal length lens is? Is it from the previous example of the first paragraph? If yes, is it f/2 or f/4 100 mm lens? Be more specific and detailed about such details, please.

"The 200 mm lens's entrance pupil has four times the area of the 100 mm lens's entrance pupil, and thus collects four times as much light from each object in the lens's field of view" - why? Expand on that with a couple of words more (by including relevant information in the article, not here).

Also, going back to the same example: "Since the area varies as the square of the pupil diameter" - why? This step occurred abruptly after the preceding one without any transition explaining how the square was come up with. Scrutinizer798 (talk) 01:13, 23 September 2020 (UTC) Scrutinizer798 (talk) 01:03, 23 September 2020 (UTC)

I fixed the ambiguity. It was the 100 mm f/4 lens that was intended.

The fact that the area of a circle is proportional to the square of the diameter is pretty basic, but I added a footnote that references Area of a circle for readers who are unfamiliar with it. The fact that an opening with four times as much area collects four times as much light is pretty obvious. I don't think it really needs further explanation. --Srleffler (talk) 03:02, 24 September 2020 (UTC)

The article currently reads, "Modern electronically controlled interchangeable lenses, such as those used for SLR cameras, have f-stops specified internally in 1⁄8-stop increments, so the cameras' 1⁄3-stop settings are approximated by the nearest 1⁄8-stop setting in the lens." I'd really love to see a citation for this! It seems really unlikely, given that aperture is often controlled by a solenoid on cheap cameras, and a stepper motor on more expensive ones.71.63.160.210 (talk) 09:01, 27 December 2021 (UTC)

Stepper motors often step smaller than the actual needed step size, so I wouldn't be surprised if they do 1/8. You should have a WP:RS to put it in the article, though. Otherwise 3/8 is close enough to 1/3. Gah4 (talk) 11:43, 27 December 2021 (UTC)

I wonder if that's something I put in there, years ago. It is something we discovered when I was chief scientist at Foveon, and using Canon lenses. We wanted to do 1/3-stop increments, but the lenses did 1/8 stop increments. I doubt I could find an RS for that though. Dicklyon (talk) 02:46, 3 December 2022 (UTC)

Sure enough, that was me, in this edit in 2006. Foveon and Sigma had each independently reversed-engineered the Canon EF lens interface, and found that apertures were set by integers representing eighths of stops. But sorry, I don't know that it was ever published, and I don't know that it's applicable to other interfaces such as Nikon's. Dicklyon (talk) 02:49, 3 December 2022 (UTC)

Coincidentally, as an irrelevant aside, today I donated to the Computer History Museum a 1999 Foveon Studio Camera ("model 9000" or "Foveon I"), with Canon lens, in completely working condition. The control, UI, strorage, etc. are all on a Dell Inspiron 7000 series laptop running Windows 98, as part of the camera mounted on the tripod (as here). This was the first camera we sold, to Sheldon of Los Altos, in 1999; I recently bought it back for next to nothing. It's got FoveonCam and FoveonLab 2.7 running nicely (our last release before we pivoted to the X3 sensor technology). It even has portrait session photos on it still, from their last uses in 2009. After a decade of Foveon, they moved to a modern DSLR. Dicklyon (talk) 03:25, 3 December 2022 (UTC)

3/8 is pretty close to 1/3, probably within uncertainty elsewhere in the system. Gah4 (talk) 08:02, 3 December 2022 (UTC)

Right, but then some of the increments are just 2/8. Close enough. Dicklyon (talk) 02:26, 4 December 2022 (UTC)

Yes, but one is one side, and the other the other side, so the difference from an actual 1/3 is pretty small. 1/3-3/8 = 1/24. So each is within about 4% of the marked value. Gah4 (talk) 05:31, 6 December 2022 (UTC)

We are in agreement. Perhaps, however, this is just about Canon lenses (and others that emulate them); that's all I know. Dicklyon (talk) 06:57, 6 December 2022 (UTC)

Actually, I just checked the old FoveonCam 2.7 user manual, which shows that we supported aperture settings in 1/8 stop increments, denoted like f/5.6 + 1/8 (that's the example showing in the manual); that was more sensible than trying to make up f-numbers in 1/8-stop increments when the conventional numbers are not that accurate. Dicklyon (talk) 01:55, 6 December 2022 (UTC)

(I know essentially nothing about photography, and I think this article does an astonishingly good job of explaining its complex topic.)

In a couple of places, it is pointed out that certain assumptions are only strictly valid when the subject of the photograph is infinitely far away, and therefore those assumptions must be modified when the subject is close to the lens. But how close is "close"? In other words, at what distance from the lens does it become necessary to compensate for the closeness of the subject? I imagine it might be proportional to certain characteristics of each lens. TooManyFingers (talk) 16:14, 19 September 2023 (UTC)

Maybe the article should explain this. The correction factor is 1+M, where M is the magnification. Image size/object size. It is usual, often enough, to use whole stops. (Slide films usually need to do better.) You would then round when 1+M was about 1.2, or M of 0.2. This is pretty much what everyone calls close-up photography. Among others, there can be a shadow of the camera or lens on the object. Most cameras now have through-the-lens metering, so the correction is included. Gah4 (talk) 23:40, 19 September 2023 (UTC)

As per this actual article, an f-number is the ratio of the focal length to diameter of the aperture. It has only a correlative relationship to light gathering ability, hence the need for T-stops to account for actual light transmission through a lens as many factors affect the light gathering ability of an optical system.

The introductory sentence "An f-number is a measure of the light-gathering ability of an optical system such as a camera lens." should be re-worded to something along the lines of "An f-number is the ratio of the focal length to the aperture of an optical system such as a camera lens. However, while this often correlates to and is used as a short-hand to express the light-gathering ability of the optical system, it doesn't actually account for light transmission factors, thus T-stops are used where transmission factors need to be accounted for."

I'm open to suggestions of how to word this, but the existing line is simply objectively wrong, and contradicts the rest of the article.

Nabeel_co (talk) 06:55, 16 July 2025 (UTC)

Your version is way less readable. The intro needs to introduce the topic in a simple way, for a reader who knows nothing about it. I also disagree with your assertion. F-number is indeed a measure of the light-gathering ability of a lens. It is a relative measure, not an absolute one. It is a perfectly good measure for comparing the effect of different aperture settings on a single lens, or for comparing different lenses of similar construction. It is widely used (and not just in photography) for both of these purposes. With modern optical coatings the difference between f-number and t-stop is small enough to be negligible for most purposes. Do you really care if your lens is f/1.4 or T/1.48? --Srleffler (talk) 11:53, 16 July 2025 (UTC)

Indeed. The F-number is a reasonably good measure of the light-gathering ability of a lens. Good enough for most photographers. The T-stop is a better alternative for when you need ultimate accuracy. —Edgar.bonet (talk) 18:59, 16 July 2025 (UTC)

For sure, but saying F-number is a measurement of light is explicitly wrong. It gets used that way, but that's not what it actually is. Nabeel_co (talk) 20:58, 16 July 2025 (UTC)

The article does not say that f-number is a measurement of light. It says that it is a measure of the light-gathering ability of a lens. These are not quite the same thing. Perhaps some of your misunderstanding is just semantic? --Srleffler (talk) 04:34, 18 July 2025 (UTC)

Sure, I'm open to other suggestions of how to word it. The core problem is how it's written right now is totally incorrect, contradicts the rest of the article, and misrepresents how F-numbers work. It can't stay how it is, because it's wrong. Nabeel_co (talk) 20:59, 16 July 2025 (UTC)

It's not wrong. The f-number is a measure of the light-gathering ability of a lens. It's not the perfect measure of this ability, but it's nevertheless the most used one.

Starting the article with “An f-number is the ratio of the focal length to the aperture of an optical system” would presumably be a better choice for an article intended for people who already know the notions of focal length and aperture. For an article intended for a general audience, the first sentence should summarize the most important idea to take home.

If you have to explain the notion of f-number, in a very short sentence, to someone who has no background in optics, then the first sentence of the article, as currently written, is a good choice. It gives just enough information to let the reader understand most of the usages of the term. If instead you use your alternative, then you would have to explain what is a focal length (and then a principal point, and a focus) and what you mean by “aperture” (no, it's not the diameter of the diaphragm, it's about the entrance pupil)... Note that readers who want to go down this rabbit hole need only reach the second sentence of the article (and the links therein).

I propose:

1. Changing the beginning of the second sentence to read “It is defined as the ratio of...”, just to make clear this is the formal definition.
2. Adding a caveat at the end of the first paragraph stating that the f-number is only an approximate measure of light gathering ability, and the T-number (internal link) should be used when a better accuracy is needed.

—Edgar.bonet (talk) 08:23, 17 July 2025 (UTC)

I agree with Edgar.--Srleffler (talk) 04:34, 18 July 2025 (UTC)

I just implemented this proposition, although I wrote item 2 as a separate paragraph, also mentioning the effect of the focusing distance, and linking to the “working f-number”. While doing so, I removed the sentence “The relative aperture indicates how much light can pass through the lens at a given focal length.” which has two issues:

1. it wrongly implies that this only works at a given focal length
2. it is redundant with the first sentence of the article

—Edgar.bonet (talk) 12:41, 19 July 2025 (UTC)

Looks good.--Srleffler (talk) 15:02, 19 July 2025 (UTC)

A cursory read of the section “Camera equation (G#)” seems to imply that G# is some kind of measure of the light-gathering power of the lens. However, something seems wrong with the provided equation, as it goes to a non-zero limit when the aperture gets fully closed (N → ∞).

Turns out this is not a problem with the formula, which may well be correct. The problem is that the quantity G# has nothing to do with the light-gathering power of the lens. In fact, the article says nothing about the meaning or usefulness of this quantity, which seems like an arbitrarily-defined ratio. What is G# good for? Why would anyone want to use it? Unless someone has a compelling argument for keeping this section, I will remove it.

The correct measure of the light-gathering power of a lens would be the ratio of the illuminance I reaching the camera sensor to the luminance L of the scene in front of the lens:

$${\displaystyle {\frac {I}{L}}={\frac {\pi \tau }{4N_{w}^{2}}}}$$

where τ is the transmittance of the lens and N_w the working f-number. I would like to add this to the article, in order to give some background on why the f-number is an approximate measure of light-gathering power, and why the working f-number and the T-stop are defined the way they are. However, I am lacking a reference. Does anyone here have a reference where the formula above (or anything equivalent) is provided?

—Edgar.bonet (talk) 20:09, 19 July 2025 (UTC)

I pulled the section out and copied it below, for reference. The text accurately follows the cited reliable source, but other online sources agree with you.^[1][2] The source does say that the equation is only valid for a "traditional camera with an unobscured aperture", so the failure for the case where the aperture diameter goes to zero doesn't invalidate it.

What this seems to be good for is if you want to work backward to determine from the captured image what the brightness (radiance) of the original object was. Useful for radiometry, which was the subject of the cited paper.--Srleffler (talk) 05:25, 20 July 2025 (UTC)

Removed text

The camera equation, or G#, is the ratio of the radiance reaching the camera sensor to the irradiance on the focal plane of the camera lens:^[3]

$${\displaystyle G\#={\frac {1+4N^{2}}{\tau \pi }}\,,}$$

where τ is the transmission coefficient of the lens, and the units are in inverse steradians (sr⁻¹).

The introduction says:

The f-number is dimensionless and is usually expressed using a lower-case hooked f with the format ƒ / N, where N is the f-number.

The diagram in the introduction (to the right of the text in my browser) uses f (not hooked). This inconsistency reduces the article's credibility - something stated in the first paragraph of the article is immediately contradicted by the article itself. I suggest either getting a new diagram that uses a hooked f, or changing the introduction text. 2601:600:877F:60B0:C60:2230:4B62:C026 (talk) 17:49, 20 October 2025 (UTC)

This appears to be some kind of weird svg rendering issue. The original uploaded image does have hooked f's. When the image is rendered at different resolutions a different font is used. I'm not sure why.--Srleffler (talk) 05:17, 21 October 2025 (UTC)

I think I have it fixed. I can't adjust the original image (lack of permissions) but I uploaded a new version that renders correctly (at least for me).--Srleffler (talk) 06:22, 21 October 2025 (UTC)

Nice. I see the new diagram rendering as hooked in firefox. The only problem I see is that the hook should be below the bottom part of the numbers so that the f lines up with the numbers. I think. 2601:600:877F:60B0:1420:A1C4:39C3:C63F (talk) 06:50, 21 October 2025 (UTC)

I have now fixed the original diagram and restored it to the article. If you still see the unhooked f's, hold shift and click the "refresh" button on your browser to refresh the cache. It should now display as a hooked italic f for everyone.--Srleffler (talk) 04:59, 22 October 2025 (UTC)

It does display hooked after shift refresh. The f is also now up in the air and is pretty distracting to me. Is that how it's typically written? This alignment differs from what's in the introduction text, at least in my browser (Firefox). 2601:600:877F:60B0:84B7:4B0C:FCF2:B4E4 (talk) 07:19, 22 October 2025 (UTC)

I agree that the vertical alignment of the f is off in the image, but that is not something I can fix. --Srleffler (talk) 05:29, 24 October 2025 (UTC)

What's stopping the fixing? If it's unicode then I can see about finding a string that works. If it's something else then I don't know. 2601:600:877F:60B0:578:FD86:1FE0:6101 (talk) 03:27, 27 October 2025 (UTC)

It's an svg image file. I don't know the format well enough to tweak the character placement in the image. I could probably figure it out, but I don't have time right now, and that is best done by someone more familiar with the format.--Srleffler (talk) 03:46, 27 October 2025 (UTC)