The Half You Already Know

You already know how aperture, shutter and ISO trade against each other: give up a stop on one, take a stop back on another, brightness unchanged. This chapter does not repeat that.

It is about something else — each of those three numbers is doing two jobs at once. One is controlling brightness. The other is deciding what the photograph looks like. Most people get stuck because they only use the first job and leave the second to the camera.

So the real starting point of this whole site is one sentence:

This stop I am giving up — what am I buying with it, besides brightness?

What a stop actually is

Two minutes on the axiom of the system are worth it, because it explains why the numbers look so strange.

A stop always means the same thing: twice the light, or half of it. Wherever it happens.

• Shutter is obvious: 1/125 → 1/60 doubles the time. One stop.
• ISO is obvious: 400 → 800 doubles the amplification. One stop.
• Aperture appears to make no sense at all: 1.4, 2, 2.8, 4, 5.6, 8, 11, 16.

The origin of that sequence is simple. The f-number is focal length divided by the effective aperture diameter, and the light admitted depends on the area of that aperture. Area grows with the square of the diameter, so doubling the area means multiplying the diameter by the square root of two — about 1.414.

The sequence is just repeated multiplication by 1.414: 1 × 1.414 ≈ 1.4, again ≈ 2, again ≈ 2.8. It is not arbitrary. It is a geometric series.

Which also disposes of two recurring confusions:

• A bigger f-number means a smaller opening, because it is a ratio with the diameter underneath.
• The same f-number admits the same light at any focal length, because focal length has already been divided out. That is the entire reason f-numbers exist — otherwise every lens change would require recalculation.

One more consequence is worth naming because it saves a lot of confusion later: a third of a stop is a third of a stop everywhere too. Modern cameras move in thirds by default, which is why the aperture ring passes through 3.2 and 3.5 between 2.8 and 4, and why ISO offers 500 and 640 between 400 and 800. Those intermediate values are not a different kind of number; they are the same series subdivided.

The practical value is that all three numbers are priced in one currency. Want a stop more shutter speed? Open the aperture a stop and it is paid for. No table required.

Shutter: how much time you compress

A photograph is not an instant. It is an accumulation over a stretch of time, and shutter speed is how long that stretch is.

At 1/1000 s the slice is thin enough to hold a water droplet in the air. At 1 second the whole stretch is stacked together and the path that droplet travelled becomes one continuous white line. Same waterfall, but the two pictures describe different things: one says this instant, the other says water has been running through here.

So the shutter question is not “is this bright enough” but: the movement in this scene — do I want to keep it or erase it?

• To freeze: judge how fast the subject moves, then go above it. 1/125 holds a walking person and 1/250 is safer, a running child wants 1/500 or more, birds and ball sports head towards 1/1000.
• To draw it out: blur becomes obvious below 1/15, and beyond a second water turns to mist. Hand-held bottoms out around 1/60 (stabilisation buys a few stops); slower than that wants a tripod.

Two kinds of blur must be told apart, because their cures are opposite: subject blur (shutter too slow for the movement) and camera shake (hands not steady enough). With the first, only moving things blur while static ones stay sharp; with the second, everything blurs including walls and pavement. Stabilisation only cures the second — it moves the sensor to follow your tremor and can do nothing whatever about someone running through the frame. Chapter fifteen takes this further.

Aperture: how much of the frame is legible

Wider apertures (f/1.4, f/2) make the sharp zone thinner. This usually gets discussed as bokeh, but a more useful framing is: aperture is the tool you use to decide which things belong to this photograph.

At f/1.4 on a portrait, the signage, cables and passers-by dissolve into blocks of colour — still in frame, no longer participating. At f/11 they all come back and the viewer’s eye starts roaming. Which is right? It depends what you are saying.

Starting points:

What you want	Roughly	Side effect
Lift one person out of a busy background	f/1.4 – f/2.8	Focus must be exact; a small miss and the eyes are soft
Two or three people all sharp	f/4 – f/5.6	The safe range
Landscape or street, front to back	f/8 – f/11	Enough depth, and still inside the usable range
Smaller	f/16 and beyond	Diffraction begins softening everything; rarely worth it

Two things are widely misunderstood.

First, depth of field is a gradient, not a switch. No line separates sharp from unsharp; blur increases continuously in both directions from the plane of focus. The stated “depth of field” is only the zone where blur is too small to notice at a given viewing size — which is why a frame that looks sharp on a phone may not survive a full screen.

Second, the smallest aperture is not the sharpest one. Diffraction is present at every aperture — the smaller the hole, the more light spreads at its edges — and the only question is which stop it begins to dominate. On APS-C that threshold usually falls between f/8 and f/11, and past f/16 it is unmistakable.

As for the sharpest aperture, the convention is two or three stops down from wide open, but that depends where the lens starts: an f/1.4 prime peaks around f/2.8–f/4, while an f/2.8 zoom peaks around f/5.6–f/8. Rather than memorising a number, memorise this — when you want everything sharp, f/8 usually beats f/22, because the depth f/22 buys with diffraction costs more than the problem it solves.

ISO: what you pay for the other two

ISO makes no light. It amplifies a signal. So it is the only corner of the triangle that is purely a cost: raising it buys brightness with noise and dynamic range.

There is a causal story here that nearly everyone has backwards, and getting it right changes how you work.

Noise comes mostly from light, not from ISO. Light arrives as discrete photons at random times. Over one exposure, one pixel might collect 100 photons while its neighbour collects 105. That fluctuation is shot noise, it is physics, and no camera removes it.

The point is the ratio: the fewer photons collected, the larger the fluctuation looks. At 10,000 photons the fluctuation is around 100 — one percent. At 100 photons it is around 10 — ten percent.

So the correct statement is not “high ISO creates noise” but “there is already a lot of noise when there is little light, and high ISO merely amplifies it into visibility.”

Three consequences, all counter-intuitive:

1. Do not sacrifice shutter speed to avoid high ISO. A sharp frame at ISO 6400 beats a shaky one at ISO 400, every time. Noise is tolerable; blur is not.
2. The real fix is more light on the sensor — a wider aperture, a slower shutter, a brighter place, added light. Lowering ISO does not clean the picture up; it only darkens it.
3. Underexposing and lifting later is about the same as using high ISO. The same photons were collected either way, so the shot noise is identical; what remains is the camera’s own read noise, and whether that is added before or after amplification decides which side loses. On dual-gain bodies like the X-T5 and X-E4 the two are near-identical above the switching point — the property is called ISO invariance — while at low ISO, raising it at capture is cleaner than lifting afterwards.

On bodies of the X-T5 and X-E4 class, ISO 3200 needs no thought and 6400 is entirely usable at normal viewing sizes. Set the auto-ISO ceiling high and then forget it, and put your attention back on the two numbers that actually shape the picture.

Sunny 16, and why it still matters

One relic of the film era is worth carrying: on a clear sunny day, at f/16, the correct shutter speed is roughly one over the ISO. ISO 100 gives 1/100 at f/16 — and therefore 1/400 at f/8, 1/1600 at f/4, and so on down the sequence.

It is not there so you can meter by hand. It is there as a sanity check, and as the fastest way to internalise that the stops are one currency. Sunlight is close enough to constant that this rule held worldwide for decades, printed on the inside of every film box.

Its modern use is diagnostic: if you are outdoors in bright sun and the camera proposes 1/8000 at f/2 and ISO 3200, something is set wrong — most often an ISO left high from last night. Knowing roughly what daylight should read is how you catch that before the shoot rather than after it.

Why the meter guesses wrong

The meter’s default assumption is that the scene, averaged, should be middle grey. That holds for most scenes and fails reliably for two.

• Mostly bright scenes (snow, white walls, sky behind a subject): the camera thinks it is too bright, pulls down, and the snow turns grey. Add +1 to +2.
• Mostly dark scenes (night, dark backgrounds, dark clothing): the camera thinks it is too dark, pushes up, and the blacks turn grey while noise surfaces. Take −1 to −2.

So exposure compensation is not correcting the camera’s error. It is telling the camera that this scene was never middle grey.

Fujifilm bodies hold a real advantage here: the electronic viewfinder shows the image with exposure compensation applied, so you see the result before you press. What used to require a meter and experience is now simply displayed. The advantage only exists if you look at it — the habit of watching the frame change while turning the compensation dial is worth far more than memorising +1 and −2.

The shooting modes are asking which one you decide first

P, A, S and M get taught as a ladder of difficulty, as though M were the advanced setting. It is not. They all ask one question: of the three numbers, which do you want to decide?

Mode	What you hold	Suits
A (aperture priority)	Aperture	You decided first what should be sharp — portraits, landscape, most things
S (shutter priority)	Shutter	You decided first whether movement freezes or streaks
M with auto ISO	Aperture and shutter	You care about both, and let ISO pay
M (fully manual)	Everything	Light that does not change: studio, copy work, long night exposures

What most people actually want is the third, and it is routinely overlooked: you set aperture and shutter, ISO floats. Which is this chapter’s whole argument in one setting — you keep the two numbers that shape the picture, and hand the camera the one that is purely a cost.

Fujifilm’s physical dials make this natural: an aperture ring on A is automatic, a shutter dial on A is automatic, and leaving both off A with ISO on auto puts you in the third mode. You do not need the name of the mode, only the decision about which number is yours.

What this chapter deliberately leaves out

This is a recap, not a course in exposure. Several related things belong to later chapters and are left blank here so the entrance does not become a wall:

• Metering patterns (spot, centre-weighted, multi) — they change which part of the frame gets averaged. Multi plus compensation covers most situations.
• Reading a histogram — all of chapter nine.
• Dynamic range and highlight protection — chapters eight and nine, and chapter nineteen for the Fujifilm implementation.
• Focus modes and tracking — unrelated to exposure; it belongs to the technique of the moment (chapter fifteen).
• Flash — outside this site’s scope.

And one more important thing is missing, because it was never part of exposure at all: what this photograph is about. The triangle can make a picture correct, and a correct picture can be entirely uninteresting. Deciding what to photograph is chapter fourteen’s subject, and it ought to happen before the camera comes up.

Six common symptoms and their causes

1. Everything is blurred, including walls and pavement → camera shake, shutter too slow → get above one over the equivalent focal length, or raise ISO to buy shutter speed. Stabilisation helps here.
2. The person is blurred and the background is sharp → subject movement, shutter too slow for it → a different problem from shake, and stabilisation cannot help; add one or two stops of shutter speed.
3. Snow comes out grey → the meter averaged a bright field to middle grey → compensate +1 to +2.
4. Night scenes look washed out, the blacks are not black → the meter averaged a dark field to middle grey → compensate −1 to −2.
5. You stopped down to f/22 for sharpness and everything went soft → diffraction → return to about f/8; if you need more depth, use zone focusing (chapter fifteen) or focus stacking.
6. Interiors are dark and blurred and you will not raise ISO → treating ISO as a last resort, so shutter speed falls to the shake threshold → set the auto-ISO ceiling to 6400 and protect the shutter. Noise can be worked on; blur cannot.

Where the word “stop” comes from

The word survives from a time when it named a physical object: something you pushed in to stop the light.

Mid-nineteenth-century lenses had no adjustable iris. To reduce the aperture, a photographer slid a metal plate into a slot in the barrel, each plate pierced with a hole of a different size. Changing aperture meant swapping plates. The set was named after its populariser: Waterhouse stops.

The adjustable iris eventually replaced the plates, but the word stayed and expanded to cover the whole exposure system. Saying “a stop faster” today still uses the name of a strip of brass.

ISO carries a comparable history. Film speed had two parallel standards: the American ASA, which was linear, so doubling the number doubled sensitivity, and the German DIN, which was logarithmic. In 1974 the International Organization for Standardization merged them, and ISO replaced both. The ISO 100 and ISO 400 on your camera inherit ASA’s numbering — which is why they double rather than increment.

None of this improves your photographs, but it explains something: every oddity in the exposure system is a historical deposit. The numbers were not designed to be memorable; they were accreted by practice. Finding them unintuitive is the correct reaction.

Where this goes

With all of the above set, a photograph will be correct — sensibly bright, sharp where it should be, soft where it should be. Correct and good are different things, and the next twenty chapters are about that gap.

One thing to carry forward: every time you move any of these three numbers, you are simultaneously making a decision about what the photograph is saying. From here on, make that decision on purpose.

A table for the field

What you decided first	Lock this	Then
Freeze it, or streak it	Shutter	Aperture pays the light; ISO covers the shortfall
Blurred background, or all sharp	Aperture	Shutter pays the light; ISO covers the shortfall
Very dark, want to keep everything	Protect the shutter first	Aperture wide open, ISO on auto
A large white field (snow, fog, walls)	Compensate +1 to +2	Check in the EVF that white has not gone grey
A large dark field (night, dark ground)	Compensate −1 to −2	Check in the EVF that black has not lifted
Nothing in particular	f/5.6, 1/250, auto ISO	That covers nine-tenths of everything