Free Figure Proportion Drawing Grid Generator

The landmarks are where the system earns its place. This guide marks five: the head, the shoulders at one and a half heads, the pelvis at four, the knees at six, and the feet at the bottom.

Notice which of those move when you change the count. The head, shoulders, pelvis and knees are at fixed head-numbers. Only the feet move, because the feet are wherever the figure ends.

That single decision is the whole mechanism of the canon, and it is worth working through. At eight heads, the pelvis at four is exactly halfway down the figure — legs and torso equal. At nine heads, four is 44.4% of the way down, so the legs are 55.6% of the total.

So a nine-head figure is not an eight-head figure scaled up. It is an eight-head figure with an extra head of leg. That is precisely the fashion-illustration convention, and the reason fashion figures read as elongated in the legs specifically rather than uniformly tall.

Run it the other way and the same mechanism explains child proportions. At six and a half heads the pelvis at four is 61.5% of the way down, so the legs are only 38.5% of the figure — which is exactly why children read as short-legged and large-headed, and why drawing a child as a small adult never works.

The eight-head figure is the academic canon and it is a construction rather than a measurement. Real adult populations average closer to seven and a half heads, and the eight-head figure was adopted because it produces a more imposing result, not because it describes anyone.

The history goes back a long way and is unusually well documented, because the canons were written down. The Egyptian system used a grid rather than a head unit: the classic canon put the standing figure at eighteen squares from the sole of the foot to the hairline, revised to twenty-one in the Late Period, and the change is visible in the art.

The figure proportion overlay page follows the canons, beside the rest of the artist guide category. Vitruvius, writing in the first century BC, gives a figure of eight heads and sets out the proportions that Leonardo drew as the Vitruvian Man around 1490 — the figure inscribed in both a circle and a square, which is a claim about the body's relationship to those two forms rather than a drawing aid.

Renaissance treatises multiplied the systems. Alberti's De statua proposes a measuring apparatus rather than a canon. Dürer's Four Books on Human Proportion, published posthumously in 1528, works through five different figure types systematically and is much closer in spirit to a modern reference than to a rule.

Nineteenth-century academies settled on eight heads as the teaching standard, and it has stayed there, with eight and a half used for heroic figures and nine reserved for fashion. Comic and superhero drawing generally uses eight and a half, which is where its characteristic look comes from.

The guide here covers 6.5 to 9. There is a preset labelled for a child at five heads, but the drawing clamps to a minimum of six — selecting it produces a six-head figure rather than a five-head one. That is a defect in the overlay rather than a design decision, and it is worth knowing before you rely on the child preset for anything.

For genuinely young children — a toddler is closer to four heads and a newborn to three and a half — this guide cannot reach the proportions at all, and a different reference is needed.

The shoulder mark at one and a half heads is the least discussed and one of the most useful. Shoulder placement is where beginner figures most often go wrong, because the head is drawn first at a comfortable size and the shoulders are then placed by feel, usually too close to it.

The knees at six heads on an eight-head figure put them three quarters of the way down, which is another position frequently misjudged. The instinct is to place the knee halfway between pelvis and foot; the canon puts it below that.

What the guide deliberately does not mark is width. Head counts govern the vertical only, and the width of the shoulders, the hips and the ribcage are separate conventions — shoulders at about two heads wide for a male figure, less for a female one, is the usual starting point and it is not drawn here.

That omission is honest rather than lazy. Width conventions vary far more between traditions than height ones do, and a guide that asserted a width would be making a claim the vertical canon does not support.

The most productive use over a photograph is measurement rather than construction. Scale the ladder to a photographed figure and read where that person's landmarks actually fall. Almost nobody is eight heads, and knowing what your model is measures your drawing against them rather than against a convention.

That is also the fastest way to see why a drawing feels wrong. A figure drawn at eight heads from a seven-and-a-half-head model will read as subtly elongated and nobody will be able to say which part is too long, because the answer is all of it, slightly.

For the head itself the Loomis head generator builds it from a sphere, and the face proportion guide divides the face. For construction under a drawing, the sequence is head size first, then the ladder, then the landmarks, then the figure. Drawing the figure first and fitting a ladder to it afterwards tells you what you did rather than what you meant.

The central axis is a plumb line rather than a spine. A standing figure's weight has to be over its supporting foot, and the axis is where you check that — a figure whose mass sits off the axis will look as though it is falling, which is sometimes the intention and usually not.

The guide draws over your own image rather than exporting a fixed template because the interesting question is always about a specific figure. A ladder on a blank sheet is a diagram; the same ladder on a photograph is a measurement.

The export sizes itself from the paper format and the DPI you choose rather than from your screen. SVG keeps every division editable, which is what you want when the ladder is construction that will be erased.

Everything runs in your browser. Images you load are never uploaded, because there is no server for them to reach. The drawing, the transform and the export all happen on your own machine.