A fringe changes one measurable thing on a long face: it moves the line where the visible face starts. The hairline stays where it is, but a straight, opaque fringe hanging over the forehead hides the first few millimetres of it, and the length a tape would find from the new edge to the chin is shorter by exactly that amount. Divide the shorter length by the same cheekbone width and the face length ratio falls.
That is the whole mechanism, and it is arithmetic rather than opinion. What the arithmetic cannot tell you is which fringe looks good, how dense it has to be before the hairline stops showing through, or whether a stranger reads the face as shorter. This page gives the millimetres and the ratio they move, and says plainly where the chain stops.
What a fringe can change, and what it cannot
The reading this site classifies a face on is face length divided by cheekbone width, called R1 in the rule set. Face length runs from the centre of the hairline to the tip of the chin; cheekbone width runs across the tops of the cheekbones. A fringe does not move either endpoint. It covers the upper part of the first one, so the line a viewer — or a tape — can see starts lower. Call the covered amount c. The visible reading is (length − c) divided by the same width.
Two consequences follow, and they are worth separating. The first is that the effect is linear: every extra millimetre of coverage lowers the visible ratio by the same amount, and that amount is 1 divided by the cheekbone width. The second is that the effect is bounded by the forehead: no fringe can cover more than the distance from the hairline to the brow without covering the brow too, and most foreheads are 50 to 60 mm tall. A face that starts far into the long band may need more coverage than the forehead has.
Do bangs make your face look shorter is, in this framework, two questions. If it means the visible length a ruler would find, the answer is yes, by c divided by the cheekbone width, and the tables below give c. If it means how a stranger reads the face, there is no published scale that converts a ratio into a perception, and this page does not invent one. The arithmetic is offered as the measurable half, not as a promise about the other half.
One assumption has to be stated because it is what makes the numbers usable. The fringe has to be opaque enough that the hairline does not show through it, and it has to cover the forehead continuously across the line where the length is measured. A see-through fringe, a parted fringe, or a fringe with gaps leaves the hairline visible and the subtraction overstates the change. The arithmetic describes the geometry of an opaque curtain, not the average behaviour of all fringes.
No account · No ad pixels · No photo upload
Detection runs on-device with MediaPipe. Nothing is uploaded — no account, no email, no storage.
The millimetres, at three cheekbone widths
The table uses 1.63 as the line to reach. That is the top of this site's oval band on the tape scale — oval is centred at 1.45 with a tolerance of 0.18 — and it is a convention rather than a natural boundary. A face above 1.63 is outside the oval band in the direction of oblong; a face at 1.83 is at the top of the oblong band. The point of the table is not that 1.63 is the right target. It is that once you pick a target, the coverage needed is a subtraction you can do with your own two numbers.
Read the row nearest your own reading, then the column nearest your cheekbone width. A face at 1.75 with a 136 mm cheekbone width needs 16.3 mm of coverage to reach 1.63. A face at 1.80 at the same width needs 23.1 mm. The same readings at a 126 mm width need less — 15.1 mm and 21.4 mm — because the same absolute coverage is a larger fraction of a narrower face. Width is the denominator, so a narrower face gets more ratio movement from the same millimetres.
The six rows stop at 2.00 because the coverage column is approaching the height of an average forehead. At 2.00 and a 146 mm width, reaching 1.63 would take 54.0 mm of coverage, which is most of a 60 mm male forehead and more than a 50 mm female one. That is the edge of what a fringe can do. Below the table, the coverage is plausible; above it, the arithmetic is still correct and the hair is no longer available.
Millimetres of forehead a fringe has to cover to bring a measured face length ratio down to 1.63, at three cheekbone widths. The inches are the millimetres divided by 25.4.
| Your measured R1 | Cover at 126 mm width | Cover at 136 mm width | Cover at 146 mm width |
|---|---|---|---|
| 1.65 | 2.5 mm / 0.10 in | 2.7 mm / 0.11 in | 2.9 mm / 0.11 in |
| 1.70 | 8.8 mm / 0.35 in | 9.5 mm / 0.37 in | 10.2 mm / 0.40 in |
| 1.75 | 15.1 mm / 0.59 in | 16.3 mm / 0.64 in | 17.5 mm / 0.69 in |
| 1.80 | 21.4 mm / 0.84 in | 23.1 mm / 0.91 in | 24.8 mm / 0.98 in |
| 1.83 | 25.2 mm / 0.99 in | 27.2 mm / 1.07 in | 29.2 mm / 1.15 in |
| 2.00 | 46.6 mm / 1.83 in | 50.3 mm / 1.98 in | 54.0 mm / 2.13 in |
Every figure is (your R1 − 1.63) × cheekbone width, with the width in millimetres. The 1.63 line is this site's oval band top on the tape scale (oval 1.45 ± 0.18), not a published styling standard; 1.83, the top of the oblong band, is where the last three rows sit. The inch figures are the millimetre figures divided by 25.4 and rounded separately, which is why a printed inch can differ from its millimetre by a hundredth. Coverage means vertical millimetres from the hairline to the lowest point of the fringe edge, measured on the same vertical line as the face length. It assumes an opaque fringe that covers the forehead continuously across that line. Nothing in the table says the result looks shorter to a viewer; it describes the visible length a ruler would find.
Four coverage bands, and what each one asks of the fringe
A search for fringe for oblong face returns styling advice, not a millimetre table. The four bands below convert the table into four practical requests, and the cut between them is this page's own reading aid rather than a published standard. What is published is the arithmetic; the bands are a way to hold it in your hand.
The first band is no coverage at all: a reading at or below 1.63 needs nothing from a fringe to move the ratio, because it is already inside the oval band. A short fringe that sits above the brow, a see-through fringe, or a centre part all leave the hairline visible and belong here. If your reading is in this band, the fringe decision is a styling decision and the arithmetic has no stake in it.
The second band is 1.63 to 1.70, which asks for about 0 to 10 mm of coverage at the three widths. That is a short fringe whose lowest point sits just below the hairline — enough to hide the first centimetre of forehead, not enough to reach the brow. The third band is 1.70 to 1.75, about 9 to 17 mm, which is a brow-grazing fringe. The fourth is 1.75 to 1.83, about 15 to 29 mm, which is a full fringe reaching the brow or just past it. Above 1.83, the request runs past 29 mm and toward the full height of the forehead; at 2.00 it is 47 to 54 mm, which for most people means the fringe has to reach the brow and stay dense.
The bands are cumulative in coverage but not in style. A 1.75 reading does not need a heavier fringe than a 1.70 one; it needs about six more millimetres of vertical cover, which can come from a lower hairline of the fringe, a denser fringe, or a slightly different part. The number tells you how much more forehead has to disappear, not how much more hair to cut.
Why the same fringe reads differently at 1.70 and 1.80
Long face with bangs is a phrase that treats the fringe as a single object, but the same fringe covers the same millimetres on every face. What changes is the denominator. A 16 mm fringe on a 126 mm face removes 0.127 of ratio; on a 136 mm face, 0.118; on a 146 mm face, 0.110. The fringe is identical, the forehead coverage is identical, and the effect on the reading differs by about 15 per cent across the width range.
That is why a fringe that visibly changes one long face can look like it did nothing on another. It is not that the second fringe was cut worse or styled worse. The second face is wider, so the same absolute coverage is a smaller share of the width, and the reading moves less. The table's three columns are the correction for that, and they are the reason the page asks for the cheekbone width before it asks for the fringe.
The other variable is where the reading starts. A face at 1.70 is only 0.07 above the 1.63 line, which is less than half of one published tolerance on this scale. A face at 1.80 is 0.17 above it. The first can reach the line with a short fringe; the second needs a full one. The difference between those two fringes is not a matter of taste — it is the difference between covering 9.5 mm and 23.1 mm at the same width.
The outline check that comes before the fringe
How bangs change face proportions depends on which outline the face is, and the ratio alone does not settle that inside the overlap. This site's oval band runs 1.27 to 1.63 and its oblong band 1.47 to 1.83, so a reading between 1.47 and 1.63 is inside both. In that shared band the ratio cannot tell you whether the face is oval or oblong, and the fringe advice that follows from each is different.
The check is the side line. Pull the hair back, put the mirror at eye level, and look at the outline from the temple to the jaw corner. If it drops almost straight, the face is oblong in the sense the guides use, and the length recommendations apply whatever the ratio says. If it curves inward, the face is oval, and the fringe is optional rather than corrective. That criterion has been published since 2013 and has not been revised by any of the guides this site has recorded, while the ratio borders have moved four times in ten months.
The practical order is therefore: read the side line first, then the ratio, then the width. The side line tells you whether the face is in the long family at all. The ratio tells you how far into it the face sits, but only against a named border. The width tells you how many millimetres of coverage buy a given change. A fringe chosen from the ratio alone, without the side line, can be the right number for the wrong outline.
There is a second check for the outline that costs one subtraction. Compare the jaw width with the cheekbone width. On the tape scale this site publishes about 0.80 for oval and about 0.92 for oblong, jaw over cheekbones. The two bands touch, but the direction holds: the closer the jaw is to the cheekbones, the straighter the side reads. If the side line and the jaw ratio disagree, trust the side line first, for the same reason the guides have never revised it.
What the arithmetic leaves out
Hair density is the first omission. The table assumes the fringe hides the hairline completely. Most fringes do not: a fine fringe, a grown-out fringe, or a fringe with a gap lets the hairline show through, and the visible starting line stays higher than the arithmetic assumes. The correction is not a different formula; it is to measure the coverage on the finished fringe, not on the cut you asked for.
The second omission is the shape of the fringe edge. The table measures to the lowest point of the edge, because that is the point that sets the visible start of the face. A curved fringe, a curtain fringe, and a blunt fringe with the same lowest point have the same effect on this reading, even though they look different. A V-shaped fringe whose centre hangs lower than its ends has its lowest point at the centre, and that is the point to measure.
The third omission is the rest of the silhouette. A fringe changes the hairline, but the hair around the face also changes where the outline is read. Width at the temples, volume at the sides, and the length of the front pieces all move the perceived outline without changing the face length ratio. The arithmetic here is deliberately narrow: it describes the one thing a fringe covers, and it does not claim to describe the whole hairstyle.
The fourth omission is the camera. A photograph taken at a different distance or focal length changes the proportions in the image, and a fringe that reads one way in a mirror can read another in a phone portrait. The tape reading is the stable one; the photograph is a projection. If the two disagree, the tape is the number this page is written about.
The record, in the order it arrived
The dated record below is the one this site already published on the oval-versus-oblong page, with the fringe-length page's forehead figures added as the working height. The last row is this page, and it is the only row that turns those numbers into millimetres of coverage.
What has been published about the long-face border and the forehead a fringe can cover, oldest first. The external rows are identities by kind, not links.
| Date and kind of source | What it says | What it leaves you holding |
|---|---|---|
| 22 November 2025 — an editorial guide on a face-shape tool's site | Oval near 1.5, oblong starts at 1.6, length divided by cheekbone width | A border and a denominator, but no fringe measurement |
| 16 July 2026 — a guide site devoted to a single shape | Oval 1.25 to 1.45, a borderline band to 1.5, oblong at 1.5 and above; divided by the average of three widths | A different denominator, which moves the same face across the border |
| 14 August 2026 — an AI detector's comparison article | Oval near 1.5, oblong near 1.75 or more, divided by cheekbone width; no hard cutoff between them | The widest border in the record, and an explicit statement that the border is soft |
| 12 September 2026 — this site's published rule set | Oval 1.45 ± 0.18, oblong 1.65 ± 0.18 on the tape scale; the two bands share 1.47 to 1.63 | The two bands and the overlap this page uses as its target line |
| 16 October 2026 — this site's fringe-length page | Forehead height of approximately 50 mm for women and 60 mm for men, with the distance from hairline to brow built as a millimetre ladder | The height a fringe has to work within, and the reason 29 mm is not the same request as 54 mm |
| 3 November 2026 — this page | Coverage needed to bring a measured ratio to 1.63, at 126, 136 and 146 mm cheekbone widths | A subtraction you can run on your own two numbers, and no claim about how the result is perceived |
The first three rows are the external borders recorded on this site's oval-versus-oblong page; none of them discusses a fringe, and none is reprinted here as a styling standard. The 12 September row is this site's own rule set, published on the about page and implemented in the classifier, not a fourth external guide. The 16 October row is this site's fringe-length page, which supplies the forehead heights but measures the fringe's own length rather than the forehead it covers. The last row is the only one whose numbers are derived for this page, and its derivation is in the table note above.
A fringe changes a long face by covering forehead, and forehead is a length. At a 136 mm cheekbone width, 10 mm of coverage lowers the visible face length ratio by about 0.07; a face at 1.75 needs about 16 mm to reach 1.63, the top of this site's oval band. The same fringe does less on a wider face and more on a narrower one, because the width is the denominator.
What the arithmetic does not do is choose the fringe. It does not know whether your hair is dense enough to hide the hairline, whether the fringe edge is curved, or whether the rest of the silhouette is working with it or against it. It gives you the one number a fringe actually changes, and it stops there.
Measure the side line first, then the length and the cheekbone width, then read the row. If the row asks for more coverage than your forehead has, the answer is not a heavier fringe; it is that this particular lever has reached its limit.
Questions this page answers
How many millimetres of forehead should a fringe cover on a long face?
It depends on how far into the long band your reading sits and how wide your cheekbones are. At a 136 mm width, a face at 1.70 needs about 9.5 mm to reach 1.63, a face at 1.75 needs about 16.3 mm, and a face at 1.80 needs about 23.1 mm. At 126 mm the same readings need 8.8, 15.1 and 21.4 mm; at 146 mm they need 10.2, 17.5 and 24.8 mm. The formula is (your reading minus 1.63) times your cheekbone width.
Does a see-through fringe change the ratio as much as a dense one?
No. The arithmetic assumes the fringe hides the hairline completely across the line where the face length is measured. A see-through fringe, a parted fringe, or one with gaps leaves the hairline visible, so the visible face still starts higher than the calculation assumes and the reading moves less. If you want the number to describe the finished look, measure the coverage on the finished fringe rather than on the length you asked the stylist for.
Why does the same fringe do less on a wider face?
Because the cheekbone width is the denominator of the ratio. Sixteen millimetres of coverage removes 0.127 from a 126 mm face, 0.118 from a 136 mm face, and 0.110 from a 146 mm face. The fringe and the covered forehead are identical; only the divisor changes. That is why the table has three columns and why the page asks for the width before it gives a millimetre figure.
What if the coverage I need is taller than my forehead?
Then a fringe has reached the limit of what it can do for the reading. At a 146 mm width, a face at 2.00 would need 54 mm of coverage to reach 1.63, which is most of a 60 mm forehead and more than a 50 mm one. The arithmetic is still correct, but the hair is not there to cover it. At that point the useful move is to stop treating the fringe as a corrective and look at the side line, the silhouette and the other readings instead.