Free Manuscript Grid Generator

The historical canons are attempts to derive those numbers rather than choose them. The best known is the one Jan Tschichold reconstructed and published in The Form of the Book, crediting it to the Dutch scholar Johannes van de Graaf, who had identified it in medieval manuscripts.

The Van de Graaf canon divides the page into ninths. On a spread, the inner margin is one ninth of the page width, the outer two ninths, the top one ninth of the height and the bottom two ninths. The construction is done entirely with diagonals — no measurement — and it produces a text block with the same proportions as the page.

That last property is the point. The block being similar to the page means the two rectangles are related, and the relationship holds at any page size. Scale the book up or down and the design is unchanged, which is exactly what you want from a rule meant to serve a whole tradition rather than one edition.

This overlay's defaults produce that similarity directly. Side margins of 14% leave a block 72% of the page width; a 12% top and 16% bottom leave 72% of the height. Both 72%, so the block has the page's proportions exactly.

Where it differs from Van de Graaf is that the side margins here are symmetric. The classical canon is a spread canon: the inner margin is half the outer, because two facing pages share a gutter and the visual unit is the opening rather than the page.

The top-to-bottom relationship is the other classical instruction and it survives in this overlay's defaults. Twelve percent top against sixteen percent bottom is a 3:4 ratio, which places the block above the geometric centre of the page.

The reason is optical rather than geometric. A block placed at the measured centre reads as sitting low, for the same reason that a capital letter centred by measurement in a box looks as if it has sunk. The eye's estimate of the middle of a vertical space is above the true middle, and every typographic tradition compensates.

How much to compensate is where the arguments start. Van de Graaf's ninths give a 1:2 top-to-bottom ratio, which is a strong lift. The 3:4 here is gentler. Nineteenth-century commercial book work often used something close to equal, which is why so many books of that period look bottom-heavy to a modern eye.

The other classical system is the Golden Canon, sometimes attributed to Villard de Honnecourt, which produces margins in the ratio 2:3:4:6 — inner, top, outer, bottom. On a page with the proportions 2:3, that canon gives a text block of exactly the same proportions, and the block occupies four ninths of the page area.

Tschichold's own preference was for a page proportioned 2:3 with the Golden Canon applied, and he argued at length that the octavo formats derived from folding a sheet in halves — which produce the √2 proportion — were less satisfactory. That argument is unresolved and probably unresolvable, but it is a real argument about a real difference.

The diagonals drawn across the block are the construction tool all of these canons share. Both historical methods are performed by drawing diagonals on the page and the spread and finding where they cross; the margins fall out of the geometry rather than being measured.

Which is why they are worth having on the drawing even when you are not following a canon. If the block's diagonals are parallel to the page's diagonals, the two rectangles are similar. That is a check you can make by eye in a second, and it is the single most useful property to preserve.

For a page that needs columns rather than one block, the column grid tool takes over, and the baseline grid tool handles the vertical rhythm inside the block. For contemporary work the manuscript grid is more common than its simplicity suggests. Every novel, every essay collection, every long-form article and most academic papers use it. So does almost every well-set screen article, where the single centred column of text is a manuscript grid whose margins happen to be responsive.

The measure — the line length — is the practical constraint that ties back to the margins. Between about forty-five and seventy-five characters per line is the usual guidance, with sixty-six often cited as the ideal for continuous reading. On a fixed page size, that constraint sets the side margins once the type size is chosen.

Which direction you reason in depends on the job. For a book with a specified trim size, choose the type size and the measure follows, then the margins. For a page where the margins are the design decision, the type size follows. Doing both by feel simultaneously is how pages end up with a measure of ninety characters.

The most productive use of an overlay like this is reading an existing page rather than designing a new one. Scan or photograph a book page you like, lay the grid over it and adjust until it fits. Now you have the margins as percentages, which transfer to any page size.

That process is more informative than any rule, because it tells you what a specific designer decided for a specific kind of book. A poetry collection and a technical manual want quite different pages, and no canon distinguishes between them.

A caution about photographs: a page shot at an angle will not give you honest margins, because the perspective compresses one side. Scan flat, or photograph from directly overhead with the page pressed flat, or the numbers you read off will be wrong in a way that is hard to spot.

The export sizes itself from the paper format and the DPI you choose rather than from your screen. SVG keeps the block and its diagonals as editable lines, which is what you want when the construction goes into InDesign or Affinity as a guide layer. The manuscript grid overlay page works through them, beside the rest of the typography grid category.

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.