Key findings
- 01LooksMeter reports angles, ratios and percentages. It reports no distance in millimetres, because a single photo has no scale.
- 02Every measurement is computed from points located on one front-facing photo. The same photo on the same device gives the same numbers.
- 03This document defines 14 measurements in plain terms, each with its unit and the meaning of a higher reading.
- 04Three things move a reading more than any others: a turned head, a change of expression and camera distance.
- 05Classical targets are shown in results for orientation. The comparison that carries weight is the measured spread of real faces.
- 06The measurements are not clinically validated. They are not a medical assessment.
Section 1
Scope
This document covers the facial measurements that appear in a LooksMeter result: the angles and ratios taken from a photo of a face. It says what each one is, how it should be read and what can make it wrong.
It does not cover scores, tiers or written reports. Those are built on top of the measurements and are outside this standard. It also does not describe how the software locates a point on a face. It states what is measured, so that a reading can be understood and checked by anyone with a ruler and a photo.
Section 2
Principles
- Relative, never absolute. A photo carries no scale. Two faces of different sizes can fill the same frame. LooksMeter therefore reports angles, ratios and percentages only. It does not report a width, a height or a distance in millimetres.
- One front-facing photo. Every measurement comes from a single image of a face looking at the camera. A scan records whether the head was clearly turned, because a turned head distorts every ratio.
- Repeatable. Nothing in the measurement step is random. The same photo on the same device gives the same measurements. We have not yet tested agreement between different devices.
- Defined in the open. Each measurement has one written definition, given below. A reading means what its definition says and nothing more.
- Read against real faces. A measurement is interpreted by its position in a measured distribution. Our norms study provides that distribution.
Section 3
Definitions
The terms used below have their ordinary anatomical meaning. The inner and outer eye corners are the points where the upper and lower lids meet. The pupil line joins the centres of the two pupils. Mid-brow is the point between the eyebrows. The nose base is where the nose meets the upper lip. The philtrum is the groove between the nose base and the upper lip. Cheekbone width is the widest distance across the face at the level of the cheekbones.
Eyes
| Measurement | Definition | Unit | A higher reading means |
|---|---|---|---|
| Canthal tilt | The angle of the line from the inner to the outer eye corner, averaged across both eyes and corrected for head roll. Positive means the outer corner sits higher. | Degrees | The outer eye corner sits higher than the inner corner |
| Eye aspect ratio | Eye width divided by eye height, averaged across both eyes. | Ratio | The eye is longer for its height |
| Eye spacing to eye width | The distance between the inner eye corners divided by the average width of one eye. | Ratio | The eyes are set wider apart |
| Eye width to face width | The average width of one eye divided by face width. | Ratio | The eyes are wider for the width of the face |
Canthal tilt is the only measurement with a sign. A positive value means the outer corner is higher than the inner corner. A negative value means it is lower. The reading is corrected for a head that is tilted to one side, so a small tilt of the head does not change it.
Face proportions
| Measurement | Definition | Unit | A higher reading means |
|---|---|---|---|
| Facial width-to-height ratio (fWHR) | Cheekbone width divided by the height from mid-brow to the upper lip. | Ratio | The face is wider for the height of the midface |
| Midface ratio | The distance between the pupils divided by the height from the pupil line to the upper lip. | Ratio | The midface is shorter for the distance between the pupils |
| Face length to width | Face length from the top of the tracked face area to the chin divided by face width. | Ratio | The face is longer for its width |
The facial width-to-height ratio follows the definition used in the research literature: the width of the face over the height of the upper face. Our review of that measurement covers what it has and has not been shown to predict.
Nose, mouth and chin
| Measurement | Definition | Unit | A higher reading means |
|---|---|---|---|
| Nose width to eye spacing | Nose width at the nostrils divided by the distance between the inner eye corners. | Ratio | The nose is wider for the gap between the eyes |
| Mouth width to nose width | Mouth width divided by nose width at the nostrils. | Ratio | The mouth is wider for the width of the nose |
| Lower lip to upper lip | Lower lip height divided by upper lip height. | Ratio | The lower lip is fuller for the upper lip |
| Chin to philtrum | The distance from the lower lip to the chin divided by the length of the philtrum. | Ratio | The chin is taller for the length of the philtrum |
Vertical sections
| Measurement | Definition | Unit | A higher reading means |
|---|---|---|---|
| Upper section | Top of the tracked face area to mid-brow as a share of face height. | Percent of face height | The section from the top of the tracked face area to the brow is taller |
| Middle section | Mid-brow to nose base as a share of face height. | Percent of face height | The section from the brow to the nose base is taller |
| Lower section | Nose base to chin as a share of face height. | Percent of face height | The section from the nose base to the chin is taller |
The three sections add up to 100% of the tracked face height.
Section 4
Photo protocol
A measurement is only as good as the photo it is taken from. The protocol below removes the largest sources of error.
- Face the camera squarely with the head upright. Do not turn or tilt it.
- Hold the lens at eye level. A camera above or below the face changes every vertical measurement.
- Put distance between the camera and the face. About 1.5 m is the figure supported by the published work on camera distance.
- Keep a neutral expression with the mouth closed and the eyes open as they normally are.
- Use even light from the front. Avoid a strong light from one side.
- Keep hair, glasses and hands clear of the eyes, the brows and the jaw.
- Use the same setup for every photo you mean to compare.
The reasons for the distance rule are set out in our review of camera distance.
Section 5
What moves a reading
The table lists the known sources of error, the measurements each one affects most and what to do about it.
| Source | Measurements most affected | What to do |
|---|---|---|
| Head turned to one side | Every width and every ratio that uses a width. Left and right comparisons. | Retake the photo facing the camera. A scan records when the head was clearly turned. |
| Head tilted to one side | Canthal tilt is corrected for it. Other measurements are little affected by a small tilt. | Keep the head upright. |
| Camera above or below the face | Vertical sections, midface ratio, facial width-to-height ratio, chin to philtrum. | Hold the lens at eye level. |
| Expression | Mouth width, lip ratio, eye aspect ratio. | Use a neutral expression with the mouth closed. |
| Camera distance2, 3, 4 | Nose measurements and the facial width-to-height ratio. | Take the photo from about 1.5 m and crop it. |
| Hair over the forehead | It can change where the tracked face area starts, which affects the upper section and face length to width. | Keep hair off the forehead. Read these two measurements with caution in any case. |
| Uneven light or a blurred photo | Any measurement, through less exact placement of points. | Use even light from the front and a sharp photo. |
Camera distance in more detail
Camera distance has published numbers behind it. In a geometric model a photo from 12 inches made the nose appear 30% wider in men and 29% wider in women than it would with no perspective distortion. A photo from 5 feet showed essentially no difference.2 In measured photos of 30 volunteers the nose was 6.4% longer at 12 inches than at 5 feet.3 In portraits of 45 people a shorter focal length gave a smaller facial width-to-height ratio.4
LooksMeter reads a photo as it is. It does not estimate camera distance and it does not correct for it.
Section 6
Reading rules
- Read a measurement as a position. Place it among the percentiles in the norms study. A reading near the median is typical. A reading beyond the 5th or 95th percentile is unusual. Neither is good or bad by itself.
- Treat classical targets as reference lines. Results show figures such as 1.618 and 1.0 next to some measurements. They are there for orientation. Our review of the golden ratio sets out why they are not a test of attractiveness.
- Compare like with like. Two readings can be compared when the photos were taken the same way. A change in distance, pose or expression between them can look like a change in the face.
- Do not read small differences. We have not yet published how much a reading changes between two photos of the same person. Until we do, treat a small difference between two scans as noise.
- Remember the group. Published anthropometry shows that facial measurements differ between populations.5 Our norms pool everyone who scanned. They are not norms for any one group.
Section 7
Agreement with other methods
A fair question is whether a ratio taken from a photo matches the same ratio taken directly from a face. One study tested this for the facial width-to-height ratio. It measured the same people from 2D photographs, from 3D scans and by direct anthropometry and found high agreement across all three.1
That result is about standard photographs in a research setting. LooksMeter measurements have not been compared with 3D scans or with direct measurement. They should be read as measurements of a photo.
Section 8
What is not measured and not claimed
- No absolute size. No width, height or distance in millimetres or inches.
- No depth. A front-facing photo does not show how far the nose, brow or chin project. Profile measurements need a profile photo and are outside this version of the standard.
- No diagnosis. A measurement does not identify or rule out any medical condition.
- No clinical validation. These measurements have not been validated against clinical instruments and are not a medical device.
- No prediction of character or behaviour from any measurement.
- No inference of age, sex or ethnicity from the measurements described here.
Section 9
Changes
A change to a definition is a new version of this standard. Earlier versions stay listed so that an older result can still be read against the definition it was made under.
| Version | Date | Change |
|---|---|---|
| 1.1 | 4 August 2026 | First publication |
References
- 1Kramer RSS, Jones AL, Ward R. A lack of sexual dimorphism in width-to-height ratio in white European faces using 2D photographs, 3D scans, and anthropometry. PLoS ONE. 2012;7(8):e42705. doi:10.1371/journal.pone.0042705 · PubMed
- 2Ward B, Ward M, Fried O, Paskhover B. Nasal distortion in short-distance photographs: the selfie effect. JAMA Facial Plastic Surgery. 2018;20(4):333-335. doi:10.1001/jamafacial.2018.0009 · PubMed
- 3Pressler MP, Kislevitz ML, Davis JJ, Amirlak B. Size and perception of facial features with selfie photographs, and their implication in rhinoplasty and facial plastic surgery. Plastic and Reconstructive Surgery. 2022;149(4):859-867. doi:10.1097/prs.0000000000008961 · PubMed
- 4Třebický V, Fialová J, Kleisner K, Havlíček J. Focal length affects depicted shape and perception of facial images. PLOS ONE. 2016;11(2):e0149313. doi:10.1371/journal.pone.0149313 · PubMed
- 5Farkas LG, Katic MJ, Forrest CR. International anthropometric study of facial morphology in various ethnic groups/races. Journal of Craniofacial Surgery. 2005;16(4):615-646. doi:10.1097/01.scs.0000171847.58031.9e · PubMed
Every reference was checked against Crossref before publication. Those indexed in PubMed were checked there as well. Findings are summarised in our own words. Read the original before relying on a number.
How to cite
LooksMeter Research. Measurement standards: how LooksMeter defines and reads each facial measurement. Version 1.1. 29 October 2026. https://www.looksmeter.com/research/measurement-standards
About this document
LooksMeter is a consumer appearance analysis product. This document is written by the LooksMeter team for general information. It has not been peer reviewed or reviewed by a medical professional. The measurements it describes are not clinically validated and are not a diagnosis or a medical device. See a qualified clinician about any health or surgical decision.
Corrections are welcome at [email protected]. Material changes are recorded by raising the version number above.
Continue