What does "28mm scale" actually divide by? The hidden ratio behind heroic-scale terrain

You did the conversion properly. You looked up the real-world height, divided by the scale ratio, cut the foam, and the finished piece still sits wrong next to your army — a shade too tall, a doorway that reads a little grand, a wall your models look faintly childlike beside. Nothing in the arithmetic was wrong. The problem is upstream of the arithmetic, in the number you divided by, because "28mm scale" is not a ratio at all. It is a figure height, and turning it into a ratio requires an assumption that nobody writes on the box.

Our companion piece on terrain scale math covers the conversion itself — feet to millimetres, ratio division, doorways, foam layers. This one goes after the constant: where 1:60 comes from, why your own miniatures may not agree with it, and what to do about it.

Do the division yourself and you get a different answer

Here is the arithmetic anyone would do unprompted. A six-foot human is 1,828.8 mm tall. If that human is 28 mm tall as a miniature, the scale ratio is 1,828.8 ÷ 28, which is about 1:65. Do the same for a 32 mm figure and you get about 1:57.

Now look at what the converters actually use. The 28mm tabletop preset divides by 60. The 32mm preset divides by 56. Neither matches the naive calculation, and the 28mm case is off by nearly nine percent. That is not a rounding artefact — it is a different assumption, and it is worth knowing which one you have inherited.

Work backwards and each preset confesses its assumed human

The cleanest way to interrogate a scale constant is to multiply instead of divide. If 28 mm of miniature corresponds to 1:60, then the real person being represented is 28 × 60 = 1,680 mm — 5 ft 6 in. Run the 32mm preset the same way: 32 × 56 = 1,792 mm, or 5 ft 10.5 in.

So the two heroic presets do not merely disagree about miniature height; they disagree about how tall a person is. That has a consequence people rarely anticipate. Going from 28mm to 32mm sounds like a 14 percent jump, because 32 ÷ 28 is 1.14. But the terrain only grows by 60 ÷ 56, about 7 percent — half of what the figure heights imply — precisely because the taller miniature is standing in for a taller person. If you have ever converted a piece "up to 32mm" and found the change underwhelming, that is why.

The eyes have it: why 5 ft 6 in is not a mistake

A 5 ft 6 in soldier sounds short until you remember what classic 28mm actually measured. The long-standing wargaming convention takes the height from the sole of the foot to the eye line, not the top of the head — the head, helmet, crest, and hair all sit above the quoted number. Measure a nominally 28 mm figure over the top of its skull and 32-33 mm is entirely normal. A ratio somewhere around 1:60 falls out naturally once you account for that missing head.

Modern "heroic" sculpting muddies this further. Heroic proportions exaggerate hands, weapons, and heads relative to the body, so a 32mm heroic figure is not simply a 28mm figure enlarged by 14 percent — the body under the armour may represent barely a taller person at all. Different manufacturers land in different places, which is exactly why no single constant can be correct for every range on your shelf. The presets encode a defensible convention. They cannot encode your specific miniatures.

Measure your own army and get your own ratio

Fortunately this is a two-minute job with a ruler. Stand a representative rank-and-file infantry model on the table — not a hero, not a model on a scenic base — and measure from the bottom of its feet, ignoring the base, to the top of its head. Then divide the real-world height you want that model to represent by what you measured.

Say your figure comes in at 34 mm to the crown and you want it to read as a six-foot soldier. That is 1,828.8 ÷ 34, or about 1:54 — closer to the 32mm preset (1:56) than to the 28mm one, even if the range is sold as 28mm. Build that army's terrain at 1:60 and every structure will be about ten percent too big, which is precisely the amount that reads as "slightly off" without ever announcing itself.

Hitting a ratio the dropdown does not offer

The converter has no custom-ratio field, but it does not need one, because the whole operation is a single division. To get an effective ratio the menu does not list, scale your input instead: multiply the real-world dimension by the preset's ratio divided by your true ratio, then convert as normal.

Concretely, suppose you want 1:65 and you are using the 28mm preset (1:60). The factor is 60 ÷ 65.3 = 0.919. A 30 ft tower becomes 27.6 ft on the input line. The tool reports 140 mm — exactly what 30 ft at 1:65 gives you, and 12 mm shorter than the 152 mm the unadjusted 30 ft would have produced. Half an inch of tower, invisible in the arithmetic, obvious on the table. For the 1:54 army above, the factor with the 32mm preset is 56 ÷ 54, or 1.04 — everything gets entered four percent larger.

Note that this ambiguity applies only to the figure-named scales. HO (1:87) and the military modelling scales — 1:35, 1:48, 1:72 — are ratios by definition, with no implied human anywhere in them. If you want the conversion to be unarguable, those presets always are.

What the foam-layer count quietly leaves out

The layer figure is an honest division rounded up, and there are three things it cannot know about.

The first is glue. A stack of seven sheets has six joints. Thin PVA squeezed flat adds perhaps a tenth of a millimetre each and is irrelevant; a hot-glue bead adds closer to a millimetre, so six joints put roughly 5 mm on the finished height. On a 15 ft palisade wall at 1:60 — a 76 mm target — that is another seven percent on top of the height the rounding already added.

The second is that nominal sheet thickness is a marketing figure. XPS sold as "half inch" is 12.7 mm, not 12. Enter 12 for that 76 mm wall and you are told seven layers; the sheets you actually own stack to 89 mm. Measure a sheet with calipers once and use the real number.

Third, and most quietly: the rounding is a plain ceiling on a floating-point division, so a stack that lands exactly on a whole number can occasionally report one layer more. The 15 ft wall at 12.7 mm sheets is a live example — six sheets hit 76.2 mm perfectly, and the tool says seven. At the same settings, 10 ft and 20 ft come out clean. Treat the layer count as a shopping list rather than a build height: when the scaled height divides evenly by your sheet thickness, the last layer is optional.

One more habit worth forming: the layer count is computed from whichever dimension you typed, with no notion of what that dimension was. Enter a building's footprint width and you will be told how many sheets to stack for a width — a number that means nothing. Convert heights for layers, and everything else for reference.

Set the constant before you cut

Scale conversion has a reputation for being trivial, and the division genuinely is. The judgement is entirely in the constant, and the constant is a convention that your particular miniatures may or may not share. Measure one model, derive your own ratio, and either pick the preset that matches it or nudge your inputs by the correction factor.

Our miniature scale and terrain converter handles the conversion in feet, metres, inches, or millimetres across 28mm, 32mm, HO, 1:35, 1:48, and 1:72, and gives you the scaled size in millimetres and inches plus a foam-layer count. Feed it a ratio you have checked against your own army rather than one you inherited from a dropdown, and the pieces will finally sit right next to the models they were cut for.