Minkwitz's Theorem: Why the Sides of a Progressive Lens Never Come Into Focus
"Straight ahead is perfect. Sideways is a smear."
That sentence is the most common thing people say about their first pair of progressive lenses, and it is usually offered as proof that something went wrong: a bad measurement, a cheap lens, a prescription error. Often it is none of those. The peripheral blur of a progressive lens is not a defect that a better factory forgot to polish out. It is a geometric consequence with a name, a date and a formula: Minkwitz's theorem, published by Gunter Minkwitz in 1963.
Knowing the rule changes what you can reasonably expect from your glasses, and more usefully, what your optician can actually do about the blur.
What a progressive surface is being asked to do
A single-vision lens has one power across its whole surface. A progressive lens has to deliver distance power at the top, a steadily increasing plus addition as you move down, and a stable reading power at the bottom, all on one continuous piece of material. In the region where power changes vertically there is a line down the centre, called the umbilical line, where the surface behaves like a simple sphere at each height and carries no astigmatism. Step sideways off that line and the surface can no longer be spherical: it must curve differently in two meridians. That difference is astigmatism by definition.
So the design question is never whether unwanted astigmatism can be avoided. It is where it should sit and how quickly it should grow.
The theorem, stated properly
Minkwitz's result is a statement about rates, not amounts. Along a line perpendicular to the umbilical line, surface astigmatism grows at roughly twice the rate at which mean power changes along the umbilical line. In the usual shorthand: if the addition builds at one dioptre per 10 mm down the corridor, unwanted cylinder grows at about two dioptres per 10 mm sideways.
Near the corridor this is written as A(x) is approximately 2 times the vertical power gradient multiplied by lateral distance x, where the gradient is the change in power per millimetre down the corridor.
A worked example using the relation alone, not any manufacturer's data. A +2.00 D addition spread over a 12 mm corridor changes power at about 0.167 D per mm. Ten millimetres to the side of the corridor, the theorem predicts roughly 2 x 0.167 x 10, which is about 3.3 D of unwanted cylinder. Three dioptres of astigmatism is not a faint smear; it is a region where text thirty degrees off axis cannot be resolved. Every progressive lens sold has some version of that number somewhere on its surface.
Why designers cannot simply remove it
The relation follows from the requirement that the surface stay smooth and continuous: curvature must be consistent in every direction as you move across the lens, and breaking that consistency produces a visible jump or a machining artefact. What designers control is not whether the astigmatism exists but how it is distributed.
- Corridor rate. Building the addition over a longer vertical distance lowers the gradient and cuts the lateral penalty proportionally. This is the single biggest lever in the whole design.
- Placement. Astigmatism can be pushed outward, downward, or concentrated in a narrow band. A soft design spreads it thinly over a wide area; a harder design keeps a wider clear zone near the corridor and pays for it further out.
- Asymmetry. Most wearers converge more when reading than they diverge, and they read slightly below and inward, so the nasal side of a modern design is treated differently from the temporal side.
Why short corridors feel so much worse
Frame fashion created a problem the mathematics had already predicted. A shallow frame with 28 to 30 mm of depth needs a short corridor so the reading zone still fits inside it. Compressing a +2.25 addition into 10 mm instead of 16 mm raises the vertical gradient by well over half, and the lateral astigmatism rises by the same factor. The lens is not badly made; it is being asked to do more work per millimetre.
Four things people are told that are not true
- "Free-form lenses have eliminated peripheral astigmatism." Digital surfacing improved control, centration, and the ability to vary design by prescription. It does not repeal a result of surface geometry. Free-form lenses distribute unwanted astigmatism more intelligently; they do not delete it.
- "An expensive lens has no blur at the sides." Price buys design sophistication, coating durability and tighter tolerances. It does not buy clear vision at every angle, because such a lens cannot be manufactured.
- "Peripheral blur proves the prescription is wrong." It might: a wrong axis or a misplaced optical centre adds its own error. But two or three dioptres of cylinder twelve millimetres from the corridor is design, not error, and it appears in the same place on every pair of the same design.
- "The reading zone is now forty per cent wider." Advertising usually measures a contour at an astigmatism level no wearer would call clear. Ask which threshold defines the zone before comparing two lenses.
What actually reduces the blur you notice
- Fit before design. Fitting height, pantoscopic tilt and monocular pupillary distance put the clear corridor on your actual line of sight. A well-fitted mid-range lens beats a badly fitted premium one, every time.
- Turn your head, not just your eyes. The clear zone is a corridor, not a window. Head movement is the intended operating mode and becomes automatic for most wearers within a few weeks.
- Match the corridor to the frame depth. In a 26 mm deep frame, a long-corridor design simply loses its reading zone below the rim.
- Give adaptation time, but not indefinitely. Most adaptation happens in the first two weeks and is largely complete by four.
When the blur is a reason to go back
See your optometrist if the blur is one-sided, if it appeared suddenly on a pair you had worn comfortably for months, if distance vision above the corridor is also soft, if you see double images rather than a smear, or if the reading zone sits noticeably too low or too high. Those point to measurement, prescription or lens position problems, which are the correctable category. Persistent, symmetric, both-sides peripheral blur in a well-fitted pair is the theorem doing its job, and the answer is a different design trade-off rather than a remake of the same one.