Free Proportional Transfer Grid Generator

The first failure is the one everybody knows and nobody checks: the two sheets have different proportions. If the source is 210 × 297 and the target is 400 × 500, the same cell count gives source cells of one shape and target cells of another, and every shape in the copy is stretched by the difference.

That failure is invisible while you work. Inside each cell the copy is faithful, so nothing looks locally wrong. Because the distortion is uniform across the whole drawing, it is the hardest kind of error to see and the easiest to prevent.

The prevention is one comparison. If the source and target cells reported are both square, or both the same rectangle, the transfer is proportional. If one is square and the other is not, it is not. That is the entire check, and the reason this overlay reports both numbers rather than a scale factor.

A scale factor is a less useful number than it looks, because it only exists when the proportions match. Two sheets with different aspect ratios have two different scale factors — one horizontal, one vertical — and quoting either as "the" scale factor is how the distortion gets hidden.

The defaults here are the A-series case, which is the one where the arithmetic is friendliest. A4 is 210 × 297 mm and A3 is 420 × 594, so the target is exactly twice the source in each direction. At 8 × 8 cells, the source cell is 26.25 × 37.13 mm and the target cell is 52.5 × 74.25.

That doubling is not a coincidence. The A-series is built on the √2 proportion precisely so that halving the long side of any sheet produces the next size down with the same proportions, and every A-series enlargement is therefore automatically proportional. It is the only paper system with that property built in.

The area consequence is worth stating because it surprises people planning work. Doubling both dimensions quadruples the area, so an A4 drawing enlarged to A3 has four times as much surface to cover. Time and material both scale with area, not with the linear factor.

The second failure is the practical one: choosing a cell count that produces cells the wrong size on the target. This is where the readouts earn their place.

On a 4-metre wall at 12 columns, a cell is 333 mm — a third of a metre, which is a comfortable unit to work in with a chalk line and a straight edge. At 50 columns it is 80 mm, which means drawing and following two thousand five hundred small squares up a ladder.

Conversely, on a small target a low cell count gives cells so large that each one contains most of a feature, and the grid stops constraining anything. The rule of thumb is that the smallest feature you need to place should be about one cell across — smaller and the grid cannot tell you where it goes, much larger and the grid is doing work you did not need.

Between those two limits there is usually a comfortable answer, and the readouts are how you find it in one look rather than by trial.

The technique's history is long and unglamorous. Egyptian tomb painting used proportional grids for figures, with canons specifying the number of squares from the sole of the foot to the hairline — eighteen in the classic canon, twenty-one after the Late Period reform.

The proportional transfer overlay page covers the history, beside the rest of the artist guide grids. Renaissance workshops used it for transferring cartoons to walls and panels, alongside pouncing, which is the same operation performed with a pricked outline and charcoal dust. Alberti describes a version in De pictura in 1435 and Dürer's woodcuts of the 1520s show the apparatus in use.

Sign painting, scenic painting and mural work have used it continuously since, and for the same reason: it is the only method that requires no equipment and produces an exact enlargement. A projector is faster and needs a projector, a dark room and a surface you can stand in front of — or a phone, which is what the art projector uses in place of a lamp.

For large work the grid has an advantage a projector does not: it works in daylight, on a surface you may only be able to reach in sections, and it survives the drawing being interrupted. A chalked grid is still there tomorrow.

The usual practical sequence is to draw the grid on the target first, at the reported target cell size, using a chalk line for anything above about a metre. Then work cell by cell, and resist the temptation to draw whole features across several cells at once — the accuracy comes from the constraint.

Numbering both grids identically is the other half of the discipline. A cell reference on the source and the same reference on the target means you can stop, walk away, and resume without losing your place, which on a large piece matters more than any efficiency.

One thing this overlay deliberately does not do is compute for you. It reports what a cell will be at each end and leaves the decision alone, because the right cell count depends on the subject, the medium, the reach of your arm and how long you have — none of which a tool can know.

If you want the arithmetic done as a calculation rather than as an overlay, the proportion calculator handles the same relationship as numbers: source dimensions in, target dimensions out, with the scale factor and cell sizes computed. The two are complementary — the calculator plans, this draws.

For anything above about two metres it is worth also considering the mural scaling grid generator and the mural scaling tool, which state the wall dimensions directly and reports the cell size in metres or feet rather than millimetres. Millimetres stop being a useful unit somewhere around the point where you need a ladder.

The grid draws over your own reference rather than exporting a fixed template because you need it on the source image to read from. The target sheet gets the same grid drawn out physically, at the size the readout gives.

The export sizes itself from the paper format and the DPI you choose rather than from your screen, so a grid printed at A4 and 600 DPI has genuinely fine lines. SVG scales without loss, which matters when the target is large.

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.