Midface ratio: what "compact midface" actually measures

"Compact midface" shows up constantly in face-rating communities as shorthand for a specific look: eyes and mouth sitting closer together, cheekbones reading high and forward, the space between brow and upper lip looking short relative to face width. The usual calculation is interpupillary distance divided by the distance from the pupil line down to the top of the upper lip, with a narrow target range, often cited somewhere between 0.85 and 1.0, presented as the cutoff between a compact midface and a long one.

The measurement itself is real and reproducible. What is worth separating out is which part of that discussion is anthropometric fact, which part is aging biology, and which part is a specific numeric threshold that nobody has actually validated against attractiveness ratings.

What is actually being measured

Midface, in the anthropometric literature, refers to the middle third of the face: roughly the brow line down to the base of the nose, sometimes extended down to the upper lip depending on the source. Farkas, Hreczko, Kolar, and Munro's 1985 study in Plastic and Reconstructive Surgery, the largest direct-measurement study of facial proportion in young adults, measured this region as one of three vertical thirds and found it does not sit at a fixed one-third share of total face height, it varies by several millimeters against the other two thirds even within a single population sample.

The forum version of midface ratio is a narrower, different calculation, built for a quick before-and-after visual rather than a clinical workup: two pupil-anchored points instead of four thirds-anchored landmarks. It is measuring a real distance. It is not the same distance the surgical literature has actually validated a range for.

What the aging research actually shows

The part of the compact-midface claim that does hold up is the aging mechanism, just not in the form forums usually state it. Cevik Cenkeri, Sarigul Guduk, and Derin Cicek published an MRI study in Dermatologic Surgery in 2020 tracking the superficial fat compartments of the midface over time. It found measurable volume loss and downward redistribution of midface fat with age, part of a broader body of work on facial fat compartments that traces back to Rohrich and Pessa's anatomical mapping of the face into discrete, independently aging compartments.

That volume loss and descent is a real, imaging-confirmed mechanism, and it does functionally lengthen the visual midface over time as the cheek deflates and the lower lid to cheek junction drops. It is a legitimate reason a compact-reading midface correlates with a younger-reading face in casual observation. What it does not do is hand you a validated ideal ratio. The MRI data describes change over time within an individual, not a cutoff that separates attractive from unattractive across a population.

Why the forum number spread anyway

• It is easy to self-measure from a single front-facing photo with two rulers' worth of landmarks, no clinical equipment required.
• It borrows real credibility from adjacent, legitimate research, facial fat compartment aging, without carrying over that research's actual conclusions.
• A single ratio with a narrow ideal band is more shareable as a rule of thumb than "proportion changes with age via multiple independently aging fat compartments, and no study has scored it against attractiveness."
• It gives people something specific and numeric to fixate on, which is part of what makes it spread in self-rating communities even without population-level validation behind the exact cutoff.

What Facet measures instead

Facet's overall harmony module scores midface proportion as the middle third's share of total face height, measured from MediaPipe's landmark output rather than a two-point pupil-to-lip approximation, and compared against gender and ethnicity-keyed population bands drawn from the anthropometric literature rather than a single forum-sourced ratio. A midface that reads visually compact is not automatically scored higher; it is scored against where it actually falls relative to a measured population, the same approach used for facial thirds.

A full walkthrough of how the harmony module reports midface proportion, alongside the other nine modules, is posted at [/sample-report](/sample-report). The citation list behind every threshold, including both studies above, is at [/methodology](/methodology).