How the grid method works

You lay a grid of equal cells over the reference, and you draw the same grid on the canvas. Then you are not drawing a face or a house. You are drawing whatever sits inside one cell.

Learning at a computer

Breaking something complex into something simple

A cell of 3 by 3 centimeters almost always holds something simple: a couple of lines, the edge of a shape, an arc. Anyone can copy that accurately, even if they have never drawn.

Why the brain stops filling in

When you see an eye, the brain hands you a memorized eye template instead of what is actually in the photo. It is the same principle as drawing negative space, covered in the guide to learning to see like an artist.

Step by step

The method only works if the setup is careful. A skewed grid ruins the whole transfer.

A grid on the reference and on the canvas

  1. Crop the reference so its proportions match the canvas.
  2. Divide the reference into cells: 4 by 5 for a simple subject, 8 by 10 for a portrait.
  3. Draw the same number of cells on the canvas. They will be larger, and the count stays the same.
  4. Letter the columns and number the rows, so you do not lose your place.
  5. Mark the center cell. It is a convenient place to start.

Transfer one cell at a time

Work one cell at a time, and look only at that cell. Cover the rest. In each cell, mark where a line enters the edge and where it leaves. Those points keep the drawing accurate.

Refine, then remove the grid

When every cell is filled, remove the grid and look at the drawing as a whole. Connect those marks in one stroke from the shoulder, and only then move on to value.

Check

Before you remove the grid, shrink the image to icon size and compare it with the reference. Big proportion errors only show up at a small size, and they are still easy to fix while the grid is there.

A grid in an online editor

A digital grid is easier than a drawn one. You do not have to erase it, and it does not crease the paper. One separate layer is enough, and one button turns it off.

A layer for the grid

Make a top layer, draw the grid lines on it, then lock it so a stray stroke cannot land there. How that kind of work with several layers is set up is covered in the guide to layers in a graphics editor.

Opacity and snapping

When the method helps

A grid is not a universal tool. It is a way to solve one job: an accurate transfer. Two cases are where it saves hours.

Large formats and murals

On an A1 sheet, or on a wall, the eye is wrong more often. What looks right up close drifts when you step back. A grid holds the proportions across the whole surface, which is why people use it for wall paintings.

Difficult angles

Strong foreshortening is something the brain refuses to accept. The hand keeps drawing the arm at a normal length. A grid ends the argument, because you are copying shapes, not the object.

The downside

The method has a cost. It gives you accuracy, and it does not train the thing you are studying for. If you draw on a grid all the time, the skill stalls at copying.

It does not train your eye

Your eye improves when you compare sizes by sight and get them wrong. The grid does that work for you, so ordinary measuring and comparing, covered in the guide to proportions in a drawing, stay mandatory. Keep a rule: one piece on a grid for every four pieces without one.

How to phase it out

Reduce the number of cells. Start at 8 by 10, move to 4 by 5 after a month, then keep only two lines through the center. The last stage is working from memory, with no prompts on the canvas at all.

Time to draw, not just read

Theory fades in a couple of days if you never use it. Open the editor and spend 15 minutes on what you just read. That is enough for the skill to stick.

Try it now

A free online editor: brushes, layers, a palette, and a shared canvas. No install and no sign-up.

Start drawing

Topics in this guide

← Speedpaints and timelapses: how to learn from recordings and speed up Drawing from memory: how to learn to draw without a reference →