Lens Distortions

During a recent lecture, I asked a group of art students a simple question: do lenses distort the world? They agreed unanimously and without hesitation. It felt obvious, inevitable.

And in one sense, that answer is not wrong. But it carries something else with it, something less secure. It sits alongside a quieter assumption: that some lenses distort more than others, and that somewhere between these extremes lies a form of seeing that is, if not entirely neutral, then close enough. A standard lens. A “normal” view.

Wide-angle lenses exaggerate. Telephoto lenses compress. Fish-eye lenses bend the world into snowglobe bubbles. Against these, the standard lens appears calm, undramatic, and, crucially, faithful. It looks like how things look.

That is the familiar picture. It holds together so long as one does not ask what the term “how things look” actually means.

A standard lens corresponds, more or less, with the way things appear to us under ordinary conditions. That correspondence is what makes it feel unremarkable. The image it produces does not draw attention to itself as constructed or unusual. It passes without comment. It feels like seeing. But that sense of naturalness is not neutral. It is specific. It belongs, as Thomas Reid might have put it, to "beings who see as we do".

Consider a simple case. A circular bottle top seen from an angle presents an elliptical profile. That can be drawn, measured, even geometrically described. But we do not see an ellipse and infer a circle behind it. We see the circular bottle top, as it appears from here. The geometry of its appearance is entirely real, but it does not replace the thing with something else. It does not interpose a representation. The same is true more generally. We do not first encounter distortions and then correct for them. We encounter things, and the conditions under which they appear usually go unnoticed.

This begins to matter when we return to lenses.

What a lens does is fix certain relationships of size, distance and perspective. It establishes how objects will be related to one another in the image. These relations can vary. Some stretch space. Some compress it. Some curve it in ways that are immediately striking. Others correspond more closely with what we are used to. The key point is not that one set is correct and the others deviate. It is that certain arrangements become familiar enough that they no longer register as arrangements at all. They pass as straightforward depiction.

One way to see this more clearly is to step outside the usual continuum.

Telecentric lenses, widely used in manufacturing industries, do not behave like wide-angle, standard, or telephoto lenses. Within the working range of these lenses, objects do not diminish in size as their distance increases. A distant cube is rendered the same size as one held close to the lens. Depth does not register as a scaling difference.

This is not simply more or less distortion. It is a departure from the expectations and conditions that organise ordinary pictorial space. The result does not look like an exaggeration of familiar seeing. It looks strange in a quieter, harder-to-place way, because the usual cues are absent. Set against this, the standard lens looks less like a neutral midpoint and more like a particular calibration—one that happens to match the expectations (and moreover the optical morphology) we bring with us, as beings who see as we do. Those expectations are not incidental. They are reinforced by long habit, by the history of picturing, and by the technology that has standardised certain formats of viewing. Photographs, screens, prints—these continually return us to the same spatial conventions until they settle into something like visual common sense.

At that point, deviation becomes visible, and alignment invisible.

The language surrounding lenses does little to disturb this. In optics, the term “objective” persists as an identifier of lens elements. The French objectif and German Objektiv carry the same root. The word suggests a relation to the object itself, to what is there, as though the lens simply delivered it without remainder. Yet the lens does not stand outside the formation of the image. It is part of the arrangement that determines how things appear within it. It fixes relations that we ordinarily do not notice. The suggestion of objectivity does not remove this; it diverts attention from it.

The earlier confidence—“all lenses distort”—turns out to be both true and misleading. It is true in the sense that every lens produces an image according to specific optical conditions. But it is misleading because it treats distortion as something that varies from minimal to extreme, as though there were a baseline against which all others could be measured. What falls out of that picture is the more ordinary fact: that what looks undistorted is simply what has become familiar because it conforms to optical inevitabilities to which our perceptual system is already attuned. We do not perceive distortion and then recover the world behind it. We perceive the world, and certain conditions of its appearance recede from notice. Lenses, and perspectival imagery in general, extend and stabilise—indeed normalise--those conditions. They make a particular organisation of space feel inevitable.

From that viewpoint, the question changes. It is no longer a matter of asking which lens is most faithful. It is a matter of asking which relations are being fixed, and how easily those relations disappear into what feels like simple seeing.

To attend to distortion, then, is not to uncover a flaw at the margins of optics. It is to notice something that is ordinarily left unremarked: that how things appear is always bound up with the conditions under which they are seen, and that those conditions, once familiar enough, no longer present themselves as conditions at all.