Seeing: How a Picture Is Built

Standing Point: The Decision You Make Before the Shutter

Whether the background grows or shrinks, crowds together or spreads apart — the lens did not do that. Your feet did. The lens only decided how much of it you kept.

Changing lenses does not change the spatial relationships inside a photograph. Walking toward or away from your subject does. This chapter separates three words that get used interchangeably — perspective, compression, and blur — and hands each of them back to its actual cause.
Changing lenses does not change the spatial relationships inside a photograph. Walking toward or away from your subject does. This chapter separates three words that get used interchangeably — perspective, compression, and blur — and hands each of them back to its actual cause. Florian Schott, 2007 · CC BY-SA 3.0 · Source

The previous chapter was about where the four edges fall. This one steps back one move further, because before you decide how to cut, you have already decided something more fundamental: where you are standing.

There is a sentence almost every beginner says within their first year — “I put the long lens on and the background completely changed.” The observation is accurate. The explanation is not. And getting it wrong is expensive, because it teaches you that framing problems are solved by buying lenses. So you buy a lens, and the pictures come out the same.

This chapter argues one thing:

The lens decides how much you keep. Your feet decide how the things you kept relate to each other.

The argument comes in three layers and the order matters. The first shows that the lens does nothing. The second identifies what actually did it. The third handles the single genuine exception — which works in the opposite direction from what most people assume.

Layer one: stand still, change lenses, and you have cropped

Start with the dullest possible experiment. Put the camera on a tripod, point it at a wall that faces you squarely, and change focal length without moving a centimetre.

50mm: the whole frontage, the red-and-white shutters, cars along the kerb. The camera is fixed for this and the next two frames — nothing but the focal length changes.
50mm: the whole frontage, the red-and-white shutters, cars along the kerb. The camera is fixed for this and the next two frames — nothing but the focal length changes. Mich.kramer, 2011 · CC BY-SA 3.0 · Source
200mm from the same spot. The spacing between windows, the density of the brick courses, the angle of every shutter — all identical to the 50mm frame. This is a piece of that picture, not a different one.
200mm from the same spot. The spacing between windows, the density of the brick courses, the angle of every shutter — all identical to the 50mm frame. This is a piece of that picture, not a different one. Mich.kramer, 2011 · CC BY-SA 3.0 · Source
300mm, same spot again, down to a single window. Enlarge the 50mm frame and crop it to this area and you get the same image. The difference is pixels, not geometry.
300mm, same spot again, down to a single window. Enlarge the 50mm frame and crop it to this area and you get the same image. The difference is pixels, not geometry. Mich.kramer, 2011 · CC BY-SA 3.0 · Source

The subject here is a flat facade parallel to the sensor, and that is not an accident. Put any depth in the scene and you will catch yourself wondering whether something changed just a little. A flat wall closes that door: every point on it is the same distance from the camera, so perspective cannot change. You can measure it with a ruler.

Which makes the conclusion a hard one. From a fixed position, the telephoto frame is a crop of the wide frame. The lens did not pull the background closer, did not flatten anything, did not alter the relative size of any two objects in the scene. It changed the size of the rectangle.

You do not have to take my word for it. Take a zoom outside, find a wall facing you, shoot two frames without moving, then crop the wide one to match the long one and lay them on top of each other.

Layer two: keeping the subject the same size forces you to move

So why does everyone experience lens changes as changes to the picture? Because nobody actually stands still while changing lenses.

What really happens is this. You fit the long lens, find your subject fills too much of the frame, and step back. Or you fit the wide lens, find they have shrunk to nothing, and walk in. You think you changed lenses. You changed lenses and your standing point, every single time — and the standing point is the variable that moved the furniture.

The five frames below are one model, one footbridge, one aperture. To keep her roughly the same size in the frame, the photographer picked a new standing point for every focal length.

24mm. The whole bridge runs away into the gorge, both walls of the ravine are in shot, the far cliff is legible. Filling the frame with a person at this focal length means standing very close to her. One caveat for the whole set of five: they are not framed to a strictly identical subject size. She grows slightly larger step by step, smallest here and largest at 400mm. Read the change in background extent, not the constancy of the subject.
24mm. The whole bridge runs away into the gorge, both walls of the ravine are in shot, the far cliff is legible. Filling the frame with a person at this focal length means standing very close to her. One caveat for the whole set of five: they are not framed to a strictly identical subject size. She grows slightly larger step by step, smallest here and largest at 400mm. Read the change in background extent, not the constancy of the subject. Florian Schott, 2007 · CC BY-SA 3.0 · Source
50mm, several steps back. The bridge survives but the ravine has narrowed. She is about the same size; what changed is the amount of world behind her.
50mm, several steps back. The bridge survives but the ravine has narrowed. She is about the same size; what changed is the amount of world behind her. Florian Schott, 2007 · CC BY-SA 3.0 · Source
100mm, further back again. The bridge is down to a hint of structure and the background is mostly rock and foliage.
100mm, further back again. The bridge is down to a hint of structure and the background is mostly rock and foliage. Florian Schott, 2007 · CC BY-SA 3.0 · Source
200mm. The bridge has effectively gone and the background has collapsed into a soft wall of stone — this is what people mean by telephoto compression. Set it beside the first frame: the same person on the same bridge, with almost the entire world behind her gone.
200mm. The bridge has effectively gone and the background has collapsed into a soft wall of stone — this is what people mean by telephoto compression. Set it beside the first frame: the same person on the same bridge, with almost the entire world behind her gone. Florian Schott, 2007 · CC BY-SA 3.0 · Source
400mm, the far end of the same ladder. Nothing behind her identifies the place any more — only colour and tone remain. This frame and the 24mm frame are the same person, the same bridge, the same afternoon.
400mm, the far end of the same ladder. Nothing behind her identifies the place any more — only colour and tone remain. This frame and the 24mm frame are the same person, the same bridge, the same afternoon. Florian Schott, 2007 · CC BY-SA 3.0 · Source

Set the two layers side by side and the logic closes:

Standing pointSubject sizeBackground extentSpatial relationships
Layer one (brick wall)fixedgrowsshrinksunchanged
Layer two (footbridge)changes every frameroughly constantshrinkstransformed

Layer one established that the lens cannot alter spatial relationships. In layer two they were altered. There is exactly one remaining candidate: the walking.

The cause is geometry and fits in one sentence. How large something appears is inversely proportional to its distance from the camera. Stand one metre from a person with a cliff a hundred metres behind them and the ratio is 1:100. Back off to ten metres and the cliff is at 109 — the ratio is now 1:10.9. Your feet rewrote the proportion between foreground and background; the lens merely framed a relationship that had already been settled.

“Compression” is therefore a slightly misleading word. The long lens compresses nothing. You walked far enough away that the gap between subject and background stopped mattering relative to the total distance, so they started to look equally far off.

Blur and extent are two different things

Now for the other conflation. In that ladder, the 200mm background did not only get smaller — it got softer. It is easy to bundle “soft” and “compressed” into one property, and from there to start shopping for a lens with nicer bokeh to fix a composition problem.

They come apart cleanly. These two frames are the same shoot, same standing point, same 50mm lens. Only the aperture changed.

f/1.8. The rock face and trees dissolve; the handrail survives only as structure. Worth stating: this pair is not perfectly controlled — the framing shifts slightly between frames and the exposures do not match, so something besides the aperture is moving. Read the sharpness of the background, and whether its extent changes with it.
f/1.8. The rock face and trees dissolve; the handrail survives only as structure. Worth stating: this pair is not perfectly controlled — the framing shifts slightly between frames and the exposures do not match, so something besides the aperture is moving. Read the sharpness of the background, and whether its extent changes with it. Florian Schott, 2007 · CC BY-SA 3.0 · Source
f/13, same position, same lens. Branches, rock texture and the mesh of the walkway all return — and they return in exactly the same places, occupying exactly the same portion of the frame as at f/1.8. Aperture changed how clearly you see the background. It did not change what the background is, or how much of it there is.
f/13, same position, same lens. Branches, rock texture and the mesh of the walkway all return — and they return in exactly the same places, occupying exactly the same portion of the frame as at f/1.8. Aperture changed how clearly you see the background. It did not change what the background is, or how much of it there is. Florian Schott, 2007 · CC BY-SA 3.0 · Source

Put the two comparison sets together and you have the point of the chapter:

What you changedBackground sharpnessBackground extent and scale
Aperture only (position and focal length fixed)changes a great dealunchanged
Standing point only (subject size held)changes slightlychanges a great deal

This is also why depth of field does not get a chapter of its own here. Taught separately, that table never gets assembled. When you want a cleaner background, ask which kind of clean you mean — less legible, or less of it. The first is an aperture; the second is a step backwards. They produce genuinely different photographs and only you know which one you wanted.

What the 85mm portrait rule is really about

Every equipment guide will tell you that 85mm to 135mm is the portrait range. The advice is sound; the reason usually given for it is not. Nothing about those numbers is inherently flattering, and they do not “compress features.”

Layer two already explained it. A face is three-dimensional. The nose is nearer than the cheek, the cheek nearer than the ear. The closer you stand, the larger those few centimetres loom as a fraction of the total distance, so the nose enlarges and the ears retreat. At forty centimetres the nose may be at seventy per cent of the ear’s distance and the proportions fall apart. At a metre and a half the same few centimetres are negligible and the face reads normally.

A cat photographed from a few centimetres away. The nose is enormous, the ears have fled to the back of the head. No wide-angle lens did this — the few centimetres did it.
A cat photographed from a few centimetres away. The nose is enormous, the ears have fled to the back of the head. No wide-angle lens did this — the few centimetres did it. Theodore.shouse, 2025 · CC0 · Source
The same thing happening to a dog, here through an ultra-wide converter. Two animals rather than one, so that the effect reads as a rule about distance and not as a quirk of a particular face.
The same thing happening to a dog, here through an ultra-wide converter. Two animals rather than one, so that the effect reads as a rule about distance and not as a quirk of a particular face. Jsw618, 2024 · CC0 · Source

So the “portrait focal length” is a proxy. On full frame, filling the frame with a face at 85–135mm puts you about 1.2 to 1.5 metres away, and that distance is what does the flattering. The focal length just parks you there automatically.

The inference runs backwards too, usefully. If all you own is a 35mm you are not barred from portraits — you are barred from filling the frame with a face from up close. Step back to a metre and a half and shoot head-and-shoulders or full length and the proportions behave. Equally, a 200mm at forty centimetres, if it could focus that near, would enlarge a nose exactly as much.

Six symptoms and what causes them

  • A person in front of a spectacular landscape, and in the photograph the landscape is an unrecognisable smudge → you stood too close, which pushed the background away → step back and go longer to return the person to the same size; the landscape comes with you.
  • The distant mountain refuses to look imposing no matter how long the lens → apparent size depends only on your distance to the mountain, and you cannot walk that far → the fix is not more focal length, it is a standing point with a foreground so the mountain has something to be measured against.
  • People at the ends of a group shot look wider than the ones in the middle → wide lens plus edge of frame: they are turned obliquely to the camera and are physically nearer than the centre → tighten the group inward, or step back and go longer.
  • Your nose in selfies → the length of your arm is the shooting distance, and it is too short → a stick or another person; switching to a different front camera does nothing.
  • Cluttered background, and stopping wide open did not help → a wider aperture blurs the clutter, it does not remove any of it → move: find an angle with a simpler background, or back off and let a long lens push the clutter out of frame.
  • Legs look short (or absurdly long) in a full-length shot indoors → camera height, not focal length: shot from chest height the lower body foreshortens; drop to waist height and the proportions return → the distance rule applies vertically as well.

Layer three: edge stretching, the one real exception

Both previous layers insisted that focal length does not change geometry. Here is the one place where the angle of view genuinely acts on its own — and it acts in the opposite direction from the folklore.

Photograph a group with a very wide lens and the people at the edges are stretched sideways. Put a ball in the corner of the frame and it becomes an ellipse. This is not a lens failing to do its job. It is the unavoidable consequence of a rectilinear lens doing its job well.

A rectilinear lens has one design goal: straight lines in the world stay straight in the picture. Holding that promise across a very wide angle forces objects near the edges to be drawn long, because an object at the edge is seen obliquely and sits closer to the lens than anything in the middle. Those two facts compound, and the projection stretches. A fisheye takes the other bargain: it gives up straight lines and keeps edge shapes intact. Neither is wrong; they are different trades against the same geometric constraint.

Recognising it is easy. The stretching happens only near the edges, and only to things with volume. Flat subjects — a poster on a wall, floor tiles — survive the corners untouched. Faces and balls do not. Which leaves two remedies: move anything solid and important toward the centre, or stop using so wide an angle.

Resist reading this section as “so focal length matters after all.” What it changes is the shape of the projection, not the distances between things. Those three brick-wall frames still hold: changing focal length never moves two objects relative to each other.

The film industry built a trick out of holding these two facts in tension. The dolly zoom — Hitchcock’s stairwell in Vertigo — moves the camera and changes the focal length simultaneously, so the subject stays exactly the same size while the background rushes in or falls away. It looks unsettling precisely because it separates two things that always travel together in ordinary experience. It is layer one and layer two performed at the same time, on purpose.

The standard lens was an accident

50mm gets called the “standard” lens, usually with the explanation that it approximates the human eye. That account has problems, and the real story is more interesting.

Measured honestly, the diagonal of a 35mm frame (24×36mm) is about 43mm — so the mathematically standard lens would be 43mm, not 50. Where did 50 come from? When Barnack designed the Leica in the 1920s he turned cine film on its side and doubled the frame to 24×36. The lenses available to him were built for the 18×24mm cine format and mostly failed to cover the new, larger frame. The ones that did cover it and would still fit that small camera happened to be around two inches — roughly 50mm.

So 50mm became standard not because it matches human vision, but because it was what happened to be lying around. Every other manufacturer followed, and a century later it feels like intuition.

The number is not the lesson. The pattern is: a great deal of photographic common sense originates in the equipment constraints of some particular decade rather than in anything about seeing. The 85mm portrait rule is the same species — it is really a sentence about a metre and a half. Whenever you meet a rule stated as self-evident, the useful question is: which real variable is it standing in for?

Three numbers for Fujifilm users

The X series is APS-C, so the crop factor is 1.5×. This site always states the actual focal length and gives the full-frame equivalent in brackets where the comparison helps:

LensEquivalentWhat this chapter says about it
23mmabout 35mmThe environmental focal length. You will be close to people with it, so mind the distance for anything tighter than head-and-shoulders.
35mmabout 53mmThe least opinionated angle available; the background covers roughly what you saw standing there.
56mmabout 84mmLands inside the portrait range — which is to say, filling the frame with a face puts you at the safe distance automatically.

The difference between the three is not sharpness. It is where each one obliges you to stand. That is the real argument for a beginner staying on one lens for a while: what you need is not more options but a body that has learned what one focal length feels like in metres.

Decisions in the field

What you wantWhat to changeWhat to leave alone
Less stuff in the backgroundStep back, go longer to restore subject sizeAperture — opening up blurs clutter without removing any
A softer background, same contentOpen the aperture, or move the subject away from the backgroundYour standing point — move and the content changes
Show the environment around a personGet close, go wideDo not get so close that facial proportions collapse
A face with normal proportionsStand about 1.5m away and use whatever focal length fills the frameDo not substitute changing lenses for stepping back
An imposing distant mountainFind a position with a foreground to measure it againstDo not reach for a longer lens
No stretched faces at the ends of a groupTighten the group or step back and go longerDo not leave anyone in the corner of an ultra-wide frame

One sentence if you keep only one: let your feet decide the relationships and the lens decide the extent. Do it the other way round and you will keep buying lenses.

That said, one day you will genuinely want another one. When that day comes, work out which sense of distance you are after first, and only then look at what will go on the body: X mount lenses, grouped by what you are shooting. That page ranks nothing; it connects the kind of picture you want to the lenses you can actually buy.

Further watching

  • Lens Compression Doesn't Exist - Here's Why

    Fstoppers · 7 min

    Why this one At 0:56 the conclusion is stated outright, and it is the one this chapter spends three layers building: the look does not come from the lens but from the distance between camera and subject, and camera and background. At 1:23 they do exactly what layer one does here — crop the wide frame to the telephoto field of view and compare. From 3:30 a bicycle stands in for the wall, shot from one position at several focal lengths and then overlaid in software. The hand-drawn schematic from 1:45 to 2:30 marks 1ft and 50ft, which is the same ratio arithmetic this chapter runs.