Question 1

Can gonial angle be measured from a photograph?

Accepted Answer

Gonion is a landmark on the mandibular BONE. In a photograph it is hidden under the masseter muscle and subcutaneous fat. What a photo-based tool measures instead is the angle of the soft-tissue silhouette: the muscle-plus-fat envelope, not the jaw underneath it. That outline also swings by tens of degrees with a few degrees of head yaw. Gonial angle is a lateral-cephalogram or CBCT measurement (adult norms ≈118–128°), and it cannot be recovered from a selfie. Canthion reports instead: Two things a photo can see: jaw taper, the soft-tissue angle from the jaw angle to the chin, levelled to your pupils, and jaw definition, the shadow step across the jaw line. Both come from your ¾ views. Jaw:cheek width comes from the front. Each carries an error bar and a grade.

Question 2

Can ramus : mandibular body ratio be measured from a photograph?

Accepted Answer

The ratio requires condylion, the top of the mandibular condyle, which sits inside the temporomandibular joint and is not externally visible at all. It also requires gonion (see above). Neither exists in a photograph. Cephalometric ideals (~0.68–0.75, Jarabak-type ratios) apply to radiographs. Canthion reports instead: Jaw taper and jaw:cheek width. The vertical ramus is hidden in any photo, so the ratio itself is not attempted. The two silhouette reads carry what it would have told you.

Question 3

Can nasofacial angle, and any profile angle taken against a plane be measured from a photograph?

Accepted Answer

There are two kinds of profile angle and only one of them is refusable. A THREE-POINT angle whose vertex and both arms are points on the visible outline — the nasal root, the corner between the nose and the upper lip, the crease between lip and chin, the corner under the chin — is measurable from a photograph, because an angle between two chords does not depend on scale and does not depend on how the camera was rolled. Canthion measures those, traces them off your own silhouette against the wall Canthion reports instead: The angles that a silhouette genuinely contains, each traced off your own outline and drawn on your photograph: the nasal root, the nose-to-lip corner, the lip-to-chin crease, the angle under your chin, and the brow → nose-base → chin line with which way it be

Question 4

Can ricketts E-line, Steiner S-line, Holdaway H-line be measured from a photograph?

Accepted Answer

These are lip-to-line distances with norms of a few millimetres (E-line: upper lip −4 ± 2 mm). They require a calibrated lateral in Frankfort horizontal, and the H-line additionally needs bony nasion and point B. Ethnic variation is large. African, East Asian and South Asian populations show more anterior lips than Ricketts' original White sample, so applying the norm across ethnicities systematically labels normal lips as protrusive. Canthion reports instead: Where your lips sit between your nose tip and your chin, read off your own profile: a direction, and a share of the nose-tip→chin distance, with its ±.

Question 5

Can orbital vector / infraorbital rim position be measured from a photograph?

Accepted Answer

Defined in lateral view as the sagittal relationship of the cornea to the inferior orbital RIM. The rim is bone. It cannot be computed from a frontal photo at all, and from a lateral photo only as a soft-tissue proxy. The clearest example of a metric consumer tools claim and provably cannot measure. Canthion reports instead: The parts of the eye socket a photo can actually read: under-eye darkness (ΔL*), eye openness and canthal tilt. Each is scored and placed on the masculinity axis.

Question 6

Can ogee curve score be measured from a photograph?

Accepted Answer

The ogee is a qualitative description of the S-shaped malar–submalar contour seen on a three-quarter view, used in custom-implant planning. There are NO published quantitative norms for it. Any numeric "ogee score" is fabricated. Canthion reports instead: The visible part of the curve: cheek shading in the ¾ view, meaning light on the cheekbone against a flat mid-cheek, plus midface width.

Question 7

Can golden ratio / Marquardt Phi mask fit be measured from a photograph?

Accepted Answer

Refuted with morphometric methods. Holland (2008) showed via Procrustes and thin-plate-spline analysis that the mask is ill-suited to non-European populations and best describes MASCULINIZED white women — the opposite of what raters prefer. Khoshab et al. (2022), a PRISMA systematic review of 27 studies across six ethnic categories, found no ethnic/gender group consistently approximates either the neoclassical canons or Φ. Pallett et al. (2010): "there was never any proof that the golden ratio w Canthion reports instead: The ratios people actually preferred, instead of 1.618: Pallett's 0.36 vertical and 0.46 horizontal, scored like every other proportion here.

Question 8

Can facial width-to-height ratio (fWHR) be measured from a photograph?

Accepted Answer

As an individual-level trait indicator, discredited. Kramer (2015) found no sexual dimorphism in fWHR across photographs, 3D scans and direct anthropometry. Kordsmeyer et al. (2022) showed that bizygomatic width alone is dimorphic once size is adjusted for: the ratio's denominator destroys the signal. Deaner et al. (2012): body weight, not fWHR, predicts aggression in hockey players. And a brow raise changes the number on its own. Canthion reports instead: Width against height, with your hairline placed for you or taken from your hair-back photo, and scored. Alongside it, a masculinity axis built from the traits that genuinely separate the sexes.

Question 9

Can chin projection vs a bony reference plane (zero meridian, etc.) be measured from a photograph?

Accepted Answer

Gonzalez-Ulloa's zero meridian drops a vertical from nasion with the head in Frankfort horizontal. It needs a lateral view, and it needs a reference plane that can be reproduced (FH varies ±4.6° between individuals). Around eight mutually inconsistent published methods for chin position exist: Goode, Merrifield, Silver, Legan, Burstone, Riedel, Steiner, Byrd. That there are eight of them is itself evidence that none is a ground truth. Canthion reports instead: From the profile, which way the chin sits relative to the lips. From the front, how much of the lower face the chin takes up. The masculinity read on the chin and the beard call follow from those two.

Question 10

Can fitzpatrick type / Individual Typology Angle from RGB be measured from a photograph?

Accepted Answer

ITA requires calibrated CIE L*a*b* colorimetry under a known illuminant. sRGB from a phone under unknown light cannot yield it. Modern computational photography makes it worse: the tone mapping is skin-tone-specific, and it actively destroys colorimetric fidelity. Fitzpatrick is a different thing again. It was designed to predict UV erythema response, not to classify skin colour, and types IV–VI collapse most of the world into three bins. Canthion reports instead: Skin depth (L*), undertone, chroma and colour season, taken from a white-balanced sample. They are described and used for your palette, not forced into a phototype.

Question 11

Can any absolute measurement in millimetres be measured from a photograph?

Accepted Answer

Scale is unrecoverable from a single uncalibrated 2D image. Without a reference object in frame or a metric 3D capture (TrueDepth, stereophotogrammetry), a photo cannot say whether a face is 130 mm or 145 mm wide. It can only say what the proportions are. That is why every Canthion output is a ratio or an angle. Canthion reports instead: Everything as ratios of your own features: iris width, inter-pupil distance, face height. That is the only scale a single photo supports. Each ratio also gets a score out of 100.

Measurement	What Canthion reports, and how	Why the clinical number needs more than a photo
Gonial angle bony	Two things a photo can see: jaw taper, the soft-tissue angle from the jaw angle to the chin, levelled to your pupils, and jaw definition, the shadow step across the jaw line. Both come from your ¾ views. Jaw:cheek width comes from the front. Each carries an error bar and a grade. more →	Gonion is a landmark on the mandibular BONE. Lim et al. 2022: bigonial breadth from 2D photogrammetry off by 9.3 mm vs calipers despite ICC 0.91
Ramus : mandibular body ratio bony	Jaw taper and jaw:cheek width. The vertical ramus is hidden in any photo, so the ratio itself is not attempted. The two silhouette reads carry what it would have told you. more →	The ratio requires condylion, the top of the mandibular condyle, which sits inside the temporomandibular joint and is not externally visible at all. It also requires gonion (see above).
Nasofacial angle, and any profile angle taken against a plane lateral	The angles a silhouette genuinely contains, each traced off your own outline and drawn on your photograph: the nasal root, the nose-to-lip corner, the lip-to-chin crease, the angle under your chin, and the brow → nose-base → chin line with which way it bends. Then chin–lip direction, forehead slope, the dorsum line, and nasal projection as a ratio of your own face. All carry error bands. None are in millimetres. more →	There are two kinds of profile angle and only one of them is refusable. Derakhshan et al. 2024, Laryngoscope · Madsen, Sampson & Townsend 2008, Eur J Orthod · Naini et al. 2017 (mentolabial angle norms: SD ±20.7° in males)
Ricketts E-line, Steiner S-line, Holdaway H-line lateral	Where your lips sit between your nose tip and your chin, read off your own profile: a direction, and a share of the nose-tip→chin distance, with its ±. more →	Lip-to-line distances with norms of a few millimetres (E-line: upper lip −4 ± 2 mm). Ricketts 1960, Am J Orthod 46:330
Orbital vector / infraorbital rim position bony	The parts of the eye socket a photo can actually read: under-eye darkness (ΔL*), eye openness and canthal tilt. Each is scored and placed on the masculinity axis. more →	Defined in lateral view as the sagittal relationship of the cornea to the inferior orbital RIM. The rim is bone.
Ogee curve score invented	The visible part of the curve: cheek shading in the ¾ view, meaning light on the cheekbone against a flat mid-cheek, plus midface width. more →	The ogee is a qualitative description of the S-shaped malar–submalar contour seen on a three-quarter view, used in custom-implant planning.
Golden ratio / Marquardt Phi mask fit refuted	The ratios people actually preferred, instead of 1.618: Pallett's 0.36 vertical and 0.46 horizontal, scored like every other proportion here. more →	Refuted with morphometric methods. Holland 2008, Aesthetic Plast Surg · Khoshab et al. 2022, Aesthet Surg J · Pallett, Link & Lee 2010, Vision Res
Facial width-to-height ratio (fWHR) refuted	Width against height, with your hairline placed for you or taken from your hair-back photo, and scored. Alongside it, a masculinity axis built from the traits that genuinely separate the sexes. more →	As an individual-level trait indicator, discredited. Kramer 2015, Evol Psychol · Kordsmeyer et al. 2022, Biol Lett
Chin projection vs a bony reference plane (zero meridian, etc.) lateral	From the profile, which way the chin sits relative to the lips. From the front, how much of the lower face the chin takes up. The masculinity read on the chin and the beard call follow from those two. more →	Gonzalez-Ulloa's zero meridian drops a vertical from nasion with the head in Frankfort horizontal. Ahmad & Rohrich 2014, Aesthet Surg J: origin of the zero meridian
Fitzpatrick type / Individual Typology Angle from RGB uncalibrated	Skin depth (L*), undertone, chroma and colour season, taken from a white-balanced sample. They are described and used for your palette, not forced into a phototype. more →	ITA requires calibrated CIE Lab* colorimetry under a known illuminant. Osto et al. 2022: ITA and Fitzpatrick are not equivalent · Almeida et al. 2023: overview of skin typing scales
Any absolute measurement in millimetres scale	Everything as ratios of your own features: iris width, inter-pupil distance, face height. That is the only scale a single photo supports. Each ratio also gets a score out of 100. more →	Scale is unrecoverable from a single uncalibrated 2D image. Zielonka, Bolkart & Thies 2022 (MICA): the metric-reconstruction problem

Science

Product

Canthion

Every bone-level measurement
and how a photograph gives it.

Every bone-level measurementand how a photograph gives it.Every bone-level measurementand how a photograph gives it.X-rayPhoto

Every bone-level measurement
and how a photograph gives it.