Pupillary Distance Measure

What is pupillary distance and why do glasses need it?

Pupillary distance is the distance between the centres of your pupils, in millimetres. A lens has an optical centre, and that centre needs to sit in front of your pupil. If it does not, you look through the lens off-axis, which introduces prism — the practical result is eye strain, headaches, or a subtle sense that the glasses are wrong that people often cannot name.

The stronger your prescription, the less error it takes to matter. At a low prescription a couple of millimetres is usually unnoticeable. At a high one it is not.

How to measure pupillary distance from a photo, without a ruler

The obstacle with any photo measurement is scale: a picture tells you how many pixels apart your pupils are, not how many millimetres. Every retailer's tool solves this by asking you to hold a reference object of known size — usually a credit card, since they are a standard 85.6 mm wide — against your forehead.

There is a simpler reference, and it is already on your face. The horizontal visible diameter of the human iris is remarkably consistent: about 11.7 mm across adults, varying between roughly 11.0 and 12.5. A face mesh that detects the iris ring therefore has a ruler in every photo. Pupil-to-pupil distance in pixels, divided by iris width in pixels, times 11.7, gives millimetres.

That is what this tool does, which is why it does not ask you to find a card first.

Binocular and monocular PD: which number your optician wants

Binocular PD is the single number between your two pupils — the one most online retailers ask for, and the only one single-vision lenses usually need. It is written as one figure, for example 63.

Monocular PD is the distance from each pupil to the centre line of your face, written as two figures, for example 32/31. Most faces are slightly asymmetric, so those two numbers are frequently unequal. Progressive lenses, high prescriptions and lenses with significant prism are made from the monocular pair, because centring each lens separately is the whole point.

This tool reports both, and flags a left–right difference above about a millimetre, since that is the point where it stops being noise.

Average pupillary distance, and where yours is likely to fall

Adult PD spans roughly 54 to 74 mm. The commonly quoted averages are around 63 mm for men and 60 mm for women, and children's are smaller and change as they grow, which is why a child's PD should be remeasured rather than reused.

If your measurement lands far outside that range, the likeliest explanation is the photo rather than your face. A head turned even slightly shortens the apparent distance, and a photo taken very close introduces perspective that pulls the eyes together.

How accurate is a photo-based PD measurement?

Honestly: around ±2 mm, and there are three separate reasons rather than one.

The scale reference itself varies — iris diameter is consistent across people but not identical, and that alone contributes most of a millimetre. Head rotation shortens the measured distance by the cosine of the angle, so ten degrees of turn costs about one and a half per cent. And a photo taken close to the face has perspective distortion that pulls the eyes inward relative to the nose.

For single-vision lenses at a modest prescription, ±2 mm is generally tolerated and this measurement is fine. For progressive lenses, high prescriptions, or anything with prism, it is not — have those measured in person by an optician with a pupillometer. That is a real limit, and a tool that does not tell you about it is not being straight with you.

Getting the most accurate reading from this PD tool

Take the photo at arm's length or further, straight on, with your head level and your eyes open and looking directly at the lens. Distance matters more than most people expect — the further back the camera, the less perspective distortion, and using a rear camera at arm's length beats a front camera held close.

Take three photos and compare. If the readings agree within a millimetre, the measurement is stable. If they scatter, something in your setup is moving, and the answer is more distance and a more level head rather than more attempts.