Canthal tilt test

Canthal tilt is the slope of the line between the inner corner of the eye, the medial canthus, and the outer corner, the lateral canthus. The tool runs a face mesh on your photo, finds those two points on each eye, and compares the height of the outer corner with the inner one. Outer corner higher: positive, and the eye reads upturned. Level: neutral. Outer corner lower: negative, and the eye reads downturned. Both eyes are reported separately, because they are rarely identical, and the whole calculation happens on your device with nothing sent anywhere. The output is a band, and that is deliberate. The full definition, and why forums care about it, is under canthal tilt in the glossary; the short version is that the outer corner is anchored by a ligament to the rim of the orbit, and the skull decides where that rim sits.

The tilt is a small slope, and the camera adds its own. The rules below matter more for this tool than for almost any other on the site.

• Head straight, no turn. The tool levels a head tipped toward one shoulder before it measures, but a head turned to the side shifts the corners, and the result is marked low confidence.
• Camera exactly at eye level. A camera above the eyes with the chin dropped makes both eyes read more upturned; a camera below does the opposite. Neither is your tilt.
• Arm's length or further. A close lens bends the outer face and shifts where the outer corners appear.
• Eyes open normally, looking at the lens. Squinting pulls the outer corners; looking up or down moves the lids over the corners.
• Brows relaxed. A frown drops the brow tail over the outer corner and hides where the corner actually is.

You will see a band for each eye: positive, slightly positive, neutral or negative. Slightly positive is the ordinary case: most adults' outer corners sit a little higher than the inner ones. Neutral reads level and balanced; clearly positive reads alert. Clearly negative is less common, and if the tool reports it on a square-on, eye-level photo, retake once with the camera at eye height and the chin level before you believe it. If the two eyes land in different bands, that is usually a small turn of the head or a genuine slight asymmetry, and the second photo will tell you which. Do not read the degree figure as a measurement of your anatomy. It is how far from level the line appeared in this photo, which is partly your anatomy and partly the camera, and the tool cannot separate the two.

A negative canthal tilt means the outer corner of the eye sits lower than the inner corner, so the line between them slopes down. It is a normal variation, not a defect, and on its own it is not a health issue. It is also less common than the internet suggests: most adults have a slightly positive tilt, and most photos that look negative were taken from a little below eye level or with the chin raised, which turns the outer corners down in the image. Check that first, with the camera exactly at eye height.

The forums treat a negative tilt as a flaw, and there is a grain of evidence behind the worry: eyes that turn down at the outer corner can read as tired. In real ratings, though, tilt is a small factor next to skin, sleep and the under-eye area, and faces with a downturned eye are rated highly all the time. The tilt is bone and ligament and does not change with exercises or mewing. The downturned eyes test answers the question directly, and the upturned eyes test covers the opposite case.

The red-line filter that went viral on TikTok draws the same line from inner to outer corner. The difference here is that the tool levels the head before it measures, so a tilted phone or a tilted head does not count as a tilted eye.

It cannot give you your anatomical angle. The tool shows the measured slope in degrees for comparing your own photos, but read the band, not the decimal. A photograph flattens the orbit onto a plane, and the apparent slope changes with a head turn, camera height, lens distance and even which way you are looking. The differences those introduce are larger than the differences between most people's real tilts. The band is the honest resolution of a photo, and the full rating, under the methodology, reports tilt as a band rather than in degrees. It also cannot tell you whether your tilt is a problem. Forums frame anything short of clearly positive as a loss. The perception research does not. The 2011 review of facial attractiveness supports averageness, symmetry and sex-typical features; there is no upturned-eye rule in it, and a neutral tilt on a rested, well-framed eye reads well.

Structural. The outer corner is held by the lateral canthal ligament, which attaches to bone, and nothing non-surgical relocates it. Eye exercises, pulling at the corner, tape and squinting drills stretch thin eyelid skin and nothing else. BecomeTen tags the tilt structural, the plan never promises to change it, and the potential image is forbidden from redrawing it. Canthoplasty exists; BecomeTen does not recommend surgery at any tier. What is workable is what surrounds the corner. A brow tail that drops low over the outer corner makes a neutral eye read downturned, and tidying it is the cheapest lever on this page. Puffiness at the lower outer corner fills in the line; a study photographing the same people rested and sleep-deprived found the tired faces were judged to have more hanging eyelids and swollen eyes. And the photo itself: chin down and camera high fakes a positive tilt, which is why the full rating grades the capture rather than rewarding it.